ML20054F507
| ML20054F507 | |
| Person / Time | |
|---|---|
| Issue date: | 05/31/1982 |
| From: | Long J NRC OFFICE OF STATE PROGRAMS (OSP) |
| To: | |
| References | |
| FRN-47FR27371, FRN-49FR44645 AA47-2-054, AA47-2-54, NUREG-0891, NUREG-891, NUDOCS 8206170048 | |
| Download: ML20054F507 (115) | |
Text
NUREG-0891 Nuclear Property Insurance:
Status and Outlook U.S. Nuclear Regulatory Commission Office of State Programs J. D. Long, Consultant
, ~ ~,,
k bibi )
,s.....
8864381!!*82 =
0891 R PDR
n
,a NOTICE
~
Availability of Reference Materials Cited in NRC Publications E
Most documents cited in NRC publications will be available from one of the following sources:
1.
The NRC Public Document Room,1717 H Street, N.W.
Washington, DC 20555 2.
The N RC/GPO Sales Program, U.S. Nuclear Regulatory Commission, Washington, DC 20555 I
- 3. The National Technical informa' ion Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications,
,\\
it is not intended to be exhaustive.
Referenced documents available for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; i
Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and b
.c liceno.: documents and correspondence.
The following documents in the NUREG series are available for purchase from the NRC/GPO Sales Program formal NRC staff and contractor reports, NRC sponsored conference proceedings, and m.
NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of y
Federal Regulations, and Nuclear Regulatory Commission issuances.
l f Documents available fro m the National Technical Information Service include NUREG series
$.~
reports and technical reports prepared by other federal agencies and reports prepared by the Atomic l
l Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
.'. g l
.~
l Documents avmlable f rom public and special technical libraries include all open literature items, 7,,,
such as books, lournal and periodical articles, and transactions. Federal Register notices, federal and state impslation, and congressional reports can usually be obtained from these libraries.
~
Documents such as theses, dissertations, foreign reports and translations, and non N RC conference x
proceedings are ava lable for purchase from the organita Jon sponsoring the pubhcation cite <1.
Single copies of N RC dratt reports are available free upon written request to the Division of Tech mcal information and Document Control, U S. Nuclear Regulatory Commission, Washington, DC I
20555.
I i
+
l Copies of industry codes and standards used in a substantive manner in the NRC regulotory process are maintmned at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available there for ref erence use by the public. Codes and standards are usually copyrighted and may be purchased f rom the oriipnating orrpraation or, if they are American National Standards, f rom the l American National Standards Institute,1430 Broadway, New York, NY 10018.
- . " J ovoPoma,onyvne 56.00 C.
- N l}
}.
.J 4d 47 ;. '
NUREG-0891 Nuclear Property Insurance:
Status and Outlook Manuscript Completed: April 1982 Date Published: May 1982 J. D. Long, Consultant *
- Chairperson and Professor of Insurance Indiana University Office of State Programs U.S. Nuclear Regulatory Commission Washington, D.C. 20555
,, ~ s i
D,
'...... f
FOREWORD BY NRC STAFF This report was prepared by Dr. John Long for the U.S. Nuclear Regulatory Comission under a Personal Services Consultant Agreement. The views expressed in this report are not necessarily those of the Comission.
5 i
4 PREFACE This study was undertaken by the author in his capacity as a consultant to the United States Nuclear Regulatory Commission.
His specific assignment was to
]/
ascertain the nature of the market for nuclear property insurance as purchased by b
utilities, identify the distinguishing characteristics of each type of insurance,
^
verify the amount of insurance of each type available per nuclear site, 1y.e speculate about the likely changes in such availability, define the major problems that contribute to the inadequacy of the insurance, and recommend the
~
e,a '
]~0 remedial action that NRC should take.
4.-l;,-
4 The author is pleased to acknowledge that information was generously provided by officials of American Nuclear Insurers, Marsh & McLennan Incorporated,
.s M & M Nuclear Consultants, Mutual Atomic Energy Reinsurance Pool, Nuclear Electric Insurance Limited, Nuclear Mutual Limited, several nuclear utility risk managers, several utility financial experts, and a few professorial colleagues as well as Jerome Saltzman, Assistant Director, State and Licensee Relations, Office of State Programs of the Nuclear Regulatory Commission. Without access to this information the author would not have been able to prepare this report. The author assumes full responsibility for any errors he might have made in the use F
of the information. He also points out that the recommendations are his own and that his acknowledgnent of the sources of information is not intended to
~
imply that any other individuals necessarily concur in any of his recommendations.
~
Finally, the author would like to thank Mrs. B. V. (Rose) Fitzgerald for faithful and expert secretarial assistance.
John D. Long e
e Bloomington, Indiana April, 1982 e
ii 1~
+
+
'.*f
~
TABLE OF CONTENTS Page 1
I.
NA TUR E OF T HE S TU DY...................................................
2 A.
Questions addressed in this Report................................
B.
Definitions.......................................................
3 II.
DESCRIPTION OF PRINCIPAL SOURCES OF NUCLEAR PROPERTY INSURANCE........
4 4
A.
ANI-MAERP.........................................................
4 1.
Origin and Purpose............................................
2.
Method of Operation--ANI......................................
5 3.
Method of Operation--MAERP....................................
8 4
Coordination between ANI and MAERP............................
9 5
Lnsuring Activities and Operating Results.....................
10 B.
Nuclear Mutual Limited............................................
17 1.
Origin and Purpose............................................
17 2
Method of Operation...........................................
20 3
Insuring Activities and Operating Results..................... 25
~i j
C.
Nuclear Electric Insuranc e Limited................................
1.
Origin and Purpose............................................
32 2
Method of Operation...........................................
33 3
Insuring Activities and Operating Results.....................
34 III.
DISTINGUISHING FEATURES OF EACH TYPE OF IlGURANCE..................... 37 A.
A N I-MA ERP Ins uranc e............................................... 37 1.
All-Risk Approach.............................................
38 2.
Measurement of Loss...........................................
39 3
Coinsurance...................................................
40 42 4.
O ther Polic y Feature s.........................................
44 5
Ratemaking....................................................
B.
NML Insurance.....................................................
46 1.
Minor Differences as Compared to ANI-MAERP Insurance.......... 48 2
Major Differences as Compared to ANI-MAERP Insurance.......... 49 a.
Assessability.............................................
49 b.
Limitation of Liability................................... 52 C.
NE IL In s uranc e..........................,......................... 53 1.
Features that Distinguish NEIL Insurance from NML Insurance... 53 2
Further Distinguishing Features of NEIL-I Insurance...........
55 3
Further Distinguishing Features of NEIL-II Insurance.......... 57 IV.
AMOUNT OF NUCLEAR PROPERTY INSURANCE AVAILABLE FROM EACH PR I NC IPA L S 0UR C E.....................................................
60 A.
ANI-MAERP........................................................
60 B.
NML..............................................................
60 C.
NEIL-I...........................................................
62 D.
NEIL-II..........................................................
63 111
Page V.
ANTICIPATED CHANGES DURING 1982 IN AMOUNTS AVAILABLE.................. 63 A.
NML av.d NEIL-I....................................................
63 B.
ANI-MAERP.........................................................
64 C.
NEIL-II...........................................................
65 D.
NEIL and ANI-MAERP C ooperative Efforts............................ 68 VI.
LIKELIHOOD OF EMERGENCE DURING 1982 0F NEW PRINCIPAL SOURCE /S IN PRIVATE SECT 0R........................................................
69 A.
Limited Interest of Private Sector in Nuclear Property Insurance.. 69 B.
Efficient Organization of Private Sector..........................
70 C.
A P IC ' s Slow Gr owth................................................
73 D.
Isolated Purc hases Have B een Made................................. 75 VII. PROBLEMS AND RECOMMENDED NRC RESPONSES................................
76 A.
Historical Unreliability of Assessment Tec hnique.................. 76 1.
Reason for Use................................................
77 2
Resistance of Insureds........................................
79 3
Possible Confluence of Calls..................................
81 4
Recommended NRC Action........................................
84 B.
Waste Inherent in Dual Supplies................................... 85 1.
Capacity of NML and ANI-MAERP Each to Serve All Nuclear Utilities.....................................................
86 2
Need for Each to Serve All Nuclear Utilities.................. 87 3
Reasons Why Each Not Already Serving All Nuclear Utilities....
91 a.
Coinsurance Problem.......................................
92 b.
Speculation about NML's Posture........................... 94 4.
R ec ommended NRC Ac tion........................................ 95 C.
Disinclination of Utilities to Purchase Available Insurance....... 96 1.
Ways in which the Disinclination Manifests Itself............. 96 2
NRC's New Regulation..........................................
97 3
Suggested Alternative Approach for NRC........................
99 D.
Lack of Statistical Stability in Annual Operatin6 Data...........
101 1.
Artificiality of Annual Accounting...........................
102 2
Industry Credit Rating Plan..................................
103 3
R ec ommend ed NRC Ac tion....................................... 104 E.
Prohibitions on Purchase of Mutual Insurance.....................
106 1.
The Problem..................................................
106 2
R ecommended NRC A ction.......................................
107 F.
Avoidanc e o f Accidents...........................................
108 1.
Fragility of Insuranc e Mec hanism.............................
108 2
NRC ' s Critical Func tion......................................
109 VIII. S UMMAR Y OF R ECOMMENDATIONS..........................................
109 B IBLIOGRAPHY OF SOURCE MATER IA LS....................................
112 iv
LIST OF EXHIBITS Exhibit Page Number Number 1.
Domestic Nuclear Property Insurance Underwriting Results by Year and Cumulative, ANI-MAERP, 1958-1981............
12 2
Underwriting Gain or Loss Plus Net Investment Income as a Percentage of Earned Premium on Domestic Nuclear Property Insurance Operations by Year and Cumulative, ANI-MAERP, 1958-1981............................................
15 3
NML's Reinsurance Plans, 1982...................................
25 4
Underwriting Results by Year and Cumulative, NML, 1973-1981.....
27 5
Underwriting Cain or Loss Plus Net Investment Income as a Percentage of Earned Premium by Year and Cumulative, NML, 1973-1981......................................
28 6.
Underwriting and Investment Income for Extra Expense Insurance Program, NEIL, July 4, 1978-December 31, 19E !.......... 36 7
Rating Modification Factors, Nuclear Property Insurance Industry Experience Guide, ANI-MAERP, 1972-1982.................
47 8.
Maximum Amount of Insurance Per belear Site Offered by ANI-MAERP, 1957-1981.................................
61 9
Maximum Amount of Insurance per Nuclear Site Offered by NML, 1973-1981.......................................
62 10.
Three Alternative Ways to Design the $500 Million Excess of $500 Million Package..................................
67 v
EXECUTIVE
SUMMARY
T.
The accident at Three Mile Island demonstrated a need on the part of nuclear utilities for more nuclear property insurance than is available. This report, consisting of six questions and their respective answers, addresses this shortage.
7
=
The questions and a brief summary of each answer are presented in this abstract of the full report.
1.
What has been the development of each principal source of nuclear property insurance used as of early 1982 by nuclear utilities in E
51 the United States?
The oldest principal source of nuclear property insurance in the United States is a commercial insurance consortium of about 140 stock companies that comprise American Nuclear Insurers (ANI) and about 120 mutual companies that constitute the Mutual Atomic Energy Reinsurance Fool (MAERP). This consortium
=_
is abbreviated in this report an ANI-MAERP. The insurance and reinsurance in ANT-MAE9P have been supplied since 1957 through an intricate system that results in more than half the exposure coming to rest abroad.
In 1981 ANI-MAERP insured
[
reactors at thirty-four sites.
Without any allowance for the possibility of serious future losses, the cumulative underwriting results of ANI-MAERP prior to the Three Mile Island (TMI) accident were highly favorable to the participat-ing insurers. With payment of the $300 million TMI claim, ANI-MAERP, ignoring investment incone, found itself in a loss position on domestic nuclear property insurance operations for its total operating life.
Another principal source is Nuclear Mutual Limited (NML), a mutual insurance company domiciled in Bermuda.
It was created by several investor-owned utilities.
NML issued its first policy in 1973 In 1981 it insured reactors at twenty-seven sites. !!eing an " offshore" corporation and not conducting business in the United States, NML escapes not only U.S. federal income tax but also state insurance T
regulation. Utilities patronizing NML are subject to a 4 percent U.S. federal Jl vii
-i 5
excise tax.
Not having suffered to date any significant claim, NML's operating data are distinctly profitable.
Had the TMI claim fallen upon NML, however, NML's financial situation would have been drastically altered.
Still another Bermuda-based mutual insurer Nuclear Electric Insurance Limited (NEIL), was created in 1978 and issued its first policy in 1980.
It provides two types of insurance, as described below.
As of early 1982 it insured thirty-six sites with one type and thirty-two with the other.
2 What are some of the distinguishing features of nuclear property insurance as offered by the principal sources?
ANI-MAERP and NML issue " primary" insurance that, except for a mandatory deductible, applies to each insured loss up to an annual limit mentioned below.
The minimum ANI-MAERP deductible is $50,000 per loss; for NML it is $100,000.
Optional deductibles range much higher. The insurance is "all-risk," that is, it applies to any peril that is not explicitly excluded.
As the result, the insurance applies to radioactive contamination as well as to the conventional perils associated with electricity generating stations.
Insurance proceeds are payable not only for repair or replacement of damaged property but also--and more importantly from society's point of view--for costs of debris removal and on-site decontamination of utility property. Claims for damaged property are settled on an " actual cash value" or " replacement cost new" basis, depending on the option chosen by the utility at the time of purchase.
In this study " primary" insurance means the insurance that responds initi-ally to an insured loss.
It is distinct from " excess" insurance.
The latter responds only after and to the extent that the loss exceeds a specified substan-tial amount, all or part of which may be but is not necessarily covered by pri-mary insurance. As brought out below, primary nuclear property insurance was available as of the time of this writing up to $450 million per site. Exce ss insurance applied, subject to a limit, to that portion of the loss in excess of
$500 million.
viii
The two types of NEIL's insurance are " extra expense" and " property damage excess." The extra expense insurance (called the NBIL-I program) applies to part of the cost of purchasing replacement power when a reactor has suffered outage from a peril covered in the AN!-MAERP or NML primary insurance (or equiva-lent). Weekly payments are provided, after a twenty-six week waiting period, for up to two years but for no longer in any case than neaded with the " exercise of due diligence and dispatch" to restore the damaged reactor to use. The property damage excess insurance (called the NEIL-II program) tracks the primary coverage but pays only to the extent that the loss is in excess of a stipulated amount but not above the NEIIcII limit.
Two major differences exist between ANI-MAERP insurance on the one hand and the insurancm of NML and NEIL on the other. One difference is that ANI-MAERP insurance is financed by premiums payable at the beginning of each year at a specified rate (i.e., guaranteed cost insurance but insurance that is subject to a possible discount if loss experience has been favorable). By contrast, NML and NEIL insurance depends not only on annual premiums but also on assessment.
The premiums are lower than they would need to be without the assessment provi-sion in the policies. This provision allows NML and NEIL to call for additional premiums (called " retrospective premium adjustments") from the utilities, if and as needed to pay losses incurred during the policy year. Assessments can be called during the policy year (or anytime during the next six years).
In NML assessments for any policy year can be called up to fourteen times the annual premium for that year.
For NEIL-I the multiple is fives for NEIL-II it is seven and one-half.
.For some years the multiple may be lower.
ix
The other major difference between the two sets of insurance is that NML and NEIL undertake to limit their liability for insured losses to the assets available to pay claims. The NML policy states that the ability of NML to pay claims is limited to the assets of NML. The NEIL-II policy goes further in pro-viding that, if two or more insureds suffer losses that in the aggregate exceed NEIL's ability to pay, NEIL discharges its liability to each by paying each the pro rata share of NEIL's assets as identified in the policy.
By contrast ANI-MAERP provides a distinct and separate limit of coverage for each site insured.
A loss at one site insured by ANI-MAERP does not reduce the coverage at any other site also insured by ANI-MAERP. As brought out in the report, the significance P
of these provisions in the NML and NEIL policies is problematical. Because the ability of any corporation to pay its debts is limited to its actual and poten-tial assets, the NML and NEIL provisions may be redundant and therefore innocu-ous to the insureds.
Yet, the author regards as odd what seem to be explicit disclaimers by these insurers of any liability for claims they cannot pay.
3.
How much nuclear property insurance was offered by each of these sources as of January 1,1982?
ANI-MAERP and NML each offered up to $450 million of primary insurance per insured site.
NEIL-II, as of March 15, 1982, offered up to $290 million per insured site excess of $500 million.
All of these limits refer to the aggregate insured losses per insured site during any one policy year, irrespective of the number of reactors at the site. As of early 1982, the NEIL-I maximum benefit, payable after the waiting period, was $2 3 million per week for the first year and $1.15 million per week for the second, during continuing outage.
The NEIL-II policy also specifies that NEIL-II insureds have no claim on NEIL-I assets. Presumably the NEIL-I policies that have been renewed after the NEIL-II program was activated contain a comparable provision.
X
4 Assuming t.wt present plans come to fruition, how much nuclear property insurance is likely to be offered by each of these sources as of January 1, 1983?
Both NML and ANI-MAERP expect to raise their maximum limits per insured site to $500 million during 1982 NEIL may increase the NEIL-I weekly maximum by a small amount yet to be determined. NEIL and ANI-MAERP in a cooperative effort expect to increase the limits of available excess insurance on a bimonthly schedule of revision during 1982 at a pace to reach $500 million (excess of $500 million) by the end of 1982 (Purther subsequent increases are hoped for beyond 1982.)
The plans of all these insurers are subject to alteration, especially to slow-downs.
5 What, if any, new principal sources of nuclear property insurance are likely to emerge in the private sector by January 1, 1983?
No emerging sources comparable in magnitude or potential to the sources described above were identified by the author. Given the long lead time required for source development, the author regards the answer to be "none."
As explained in the full report, the American Power Insurance Company has not yet begun the growth that was envisioned by its founders.
6.
What problems with regard to nuclear property insurance warrant action by the United States Nuclear Regulatory Commission (NRC) and what action should NRC take in response to each problem?
Six problems are treated in the report and a recommendation about NRC's response is offered for each.
Assessment Insurance.
In the author's judgment assessment insurance is not necessarily dependable.
Its record with regard to non-nuclear insurance extends back many years and is mixed at best. Most insurers that have used assessments have grown large enough to discontinue the practice or have gone out of business.
xi
Assessments, being levied in times of crises, tend to create animosity and to cause some insureds to resist payment. The peculiar aspect about the dependence of NML, NEIL-I, and NEIL-II on assessments is that an event that would create the need for an assessment in one program would be likely also to create the need in one or both of the others. Worse yet, the event could also lead to an assessment of the same utilities under the Price-Anderson Act.
No certainty exists that state public service commissions would allow utilities to pass large assessments along to their customers in the form of increases in the price of electricity.
The possible confluence of calls and the uncertainty about pass-through of assess-ments are disturbing and have lead some critics (with some justification) to refer to assessment insurance as " funny insurance." Even so, no option exists, in the author's view, to restructure any program already in place. The author recommends, however, that NRC not encourage the creation of any new types of nuclear-related insurance that rely on assessments.
Dualism in Primary Capacity. Another problem is inherent in what the author l
calls " dualism" in primary capacity. Demonstrably, much more nuclear property l
insurance is needed than is offered in the aggregate by the principal sources.
Yet, no utility buys primary coverage simultaneously both from ANI-MAERP and NML.
The author thinks all utilities should do so.
The practice not only would increase the primary maximum limit per insured site from $450 million to some figure approaching $900 millions it also would improve the underwriting base for each insurer. Furthermore, it would be likely to reduce the underinsurance risk factor present in the rate of interest nuclear utilities pay on borrowed money.
While nothing technically prevents a utility from buying insurance from both sources, a practical obstacle apparently exists. This obstacle is the pricing practice of NML. The practice under review is the apparent unwillingness of NML to discount its rate (the price per $100 of insurance) for its insureds who would also buy the maximum insurance available from ANI-MAERP.
xii
An elaboration about the discount may be helpful.
Its justification rests on the recognition that most property losses subject to insurance are small In the absence relative to the value of the property subject to the insurance.
of any provisions limiting recovery, insureds who buy insurance at a low ratio of insurance to value have as much protection for small losses as do insureds with a high ratio of insurance to value. As observed, most losses are relatively small. For example, nuclear utilities with, say, 45 percent of insurance to value are likely as a group to collect about as much in claims over the years as nuclear utilities with a ratio, say, of 90 percent. Continuing the example, the reason is that, given the historic pattern of property losses generally, most of the losses are unlikely to exceed 45 percent of the value of the property insured. With the price for the insurance being stated per $100 of insurance bought, equity suggests that the price per unit should decrease as the quantity bought, relative to the value of the property exposed, increases.
Thus, a utility with a high ratio of insurance to value, other things the should expect to pay a lower rate for its insurance than would a utility
- same, with a low ratio of insurance to value. This expectation is justified even if the utility buys part of its insurance from one source and part from another.
The author's understanding is (1) that ANI-MAERP is willing to lower its rate to recognize the existence of insurance bought from UML by a utility that is also buying insurance from ANI-MAB7P, but (2) that NML is not willing to lower its rate for a utility buying insurance from NML in recognition of the insurance the utility also buys from ANI-MAERP.
If this impression is accurate, the author can only speculate about NML's seeming intransigence. One explanation is that NML does not wish to encourage The fact that NML joint patronage because of possible antitrust complications.
does not operate in the United States causes the author to be skeptical about the xiii
substance of this explanation. Another explanation--and one perhaps deserving of more attention--in that a significant segment of the nuclear utility industry prefers assessment insurance in a utility industry captive over advance premium insurance in the conventional insurance market.
If so, these utilities may pre-for growth of NML and NEIL at the expense of growth of ANI-MAERP.
If such a preference realb exists, the author would regard it as inconsistent with social well-being.
In the author's view social well-being calls for maximizing the amount of nuclear property insurance t hat can be developed on reasonable terms by all sources.
NRC could require nuclear utility licensees to purchase and maintain property insurance that is " acceptable" to NRC.
It could specify that, for the insurance to be acceptable, the insurer must not practice the type of discrimination alluded to above, namely, failure for pricing purposes to recognize the insurance bought by the insured from another insurer.
The author recommends that the NRC establish the acceptability requirements as articulated in the full report to cause utili-ties to patronize both primary sources.
The author concludes, however, that, given the nature of NRC's mission, NRC should concern itself only with insurance to pay the costs of on-site decontamina-tion and debris removal necessitated by accidental radioactive contamination of insured property. Whether the insurance would also pay, as it does now, for re-pair or replacement of damaged property could be immaterial, so long as the costs of decontamination and debris removal had first claim on the insurance proceeds.
As spelled out in the report, the author senses that a trustee of a bond-holders' indenture could be under pressure to preserve insurance proceeds for use in restoring damaged property.
While a trustee doubtless would be keenly inter-ested in the decontamination, such trustee might want the activity to be financed from other sources so as to leave the insurance proceeds intact for use in restoring the property. Trustees might even seek to have the proceeds placed in xiv
escrow to make sure the proceeds are not diverted. The author's view is that this problem could be formidable but that NRC's primary duty is to protect the public's interest in decontamination and removal of contaminated debris.
Utility Disinclination. Another problem is found in the disinclination of utilities to purchase all the insurance that is presently available. Aside from the failure of the utilities to use both primary sources, a few utilities do not purchase from either one as much as is availabic. Further, only thirty-one util-ities had made commitnents to NEIL-II as of this writing.
NPC's regulation to require nuclear utilities "to have and maintain" the maxinum amount of nuclear property insurance available is a recognition of the disinclination problem."
The individuals who commented to the NRC on this regulation as proposed in August,1981 identified several problems that could arise. Responding to these comments, NRC staff members suggested to the Commis-clon alterations in the proposed regulation so as to overcome the problems iden-tified by the commentators. The revised regulation was approved by the Commis-sion on March 11, 1982 It remains, however, an interim measuto.
(Theportion particularly relevant to nuclear property insurance is 10 CFR 50 54(w).]
looking beyond this regulation, the author offers the approach that he views as being in the best long-range interest of the public. This approach is con-sistent with--but goes beyond--the recently promulgated regulation.
' Incidentally, URC should require that the insurance explicitly give precedence to debrin removal as well as to decontamination.
In conventional insurance terminology the latter is not necessarily subsumed under the former.
'This regulation was proposed on August 18,1981 (159 FR 41790).
xv
'L The recommended approach, which is presented in detail in the full report, is simply that NRC require (a) that each nuclear utility obtain all the decontam-ination and debris removal insurance available to it on reasonable terms from each principal source, (b) that each utility agree in advance that in case of any dispute between the utility and an insurer about reasonableness of terms to submit to arbitration the question of whether or not the terms in dispute are reasonable, (c) that each insurer agree to do the same as a condition for NRC to deem the insurer's policies " acceptable," and (d) that a procedure similar to the one already described in the policies under " Appraisal" (or " Arbitration")
be used to resolve any dispute as mentioned above.
Attention is called to the fact tnat the proposed arbitration would be confined to the question of whether or not any disputed term was reasonable.
A positive answer would oblige the utility to purchase the coverage. A negative answer would merely relieve the utility from any duty to do so; it would not force the insurer to alter its terms.
Lack of Statistical Reliability of Data. The report stresses the lack of statistical stability of the operating data used by ANI-KAERP, NML, and NEIL in predicting insured future losses. The number of reactors is too small to give full play to the law of large numbers, as normally used in insurance ratemaking.
As the result, nuclear property insurance ratemaking is something less than an exercise in precision. A long period of favorable underwriting results is no reliable indication of similarly favorable results in the future. Amid such uncertainty, the year-by-year operating statements of these insurers are not necessarily meaningful.
As TMI demonstrated with respect to ANI-MAERP, what appears to be underwrit-ing profit can prove to be a nira6e. Prom 1957 through the mid-1970's the member companies of ANI-MAERP found themselves paying federal income taxes on amounts reported as underwriting gain but that the TMI accident showed to be only deferred loss and not gain at all. Because NML and NEIL are subject to neither XVi
f U.S. federal income taxation nor comparable taxation in Be muda, this problem may be of little concern to them. For ANI-MAERP, however, it is major.
The author recommends that, in the interest of strengthening and stabiliz-ing the nuclear property insurance market, 18C support the establishment of a device by means of which the Internal Revenue Service would allow the member companies of ANI-MAERP to reduce the variability in the annual amounts they report for tax purposes as gain or loss from their share of ANI-MAERP operations.
The device could be something similar to the Industry Credit Eating Plan used by ANI-MAELU in the liability insurance context or could be a simpler device described in the full report.
It would not necessarily change the amount of taxes payable over the long run; it could affect the timing of such payments.
Frohibitions on Mutual Insurance. Texas (as well as Idaho and Louisiana) has a constitutional provision that appears to prohibit certain publicly owned utili-ties from buying mutual insurance.
The author recommends that IRC issue one or more regulations as drafted by counsel to preempt such prohibitions in regard to nuclear property insurance.
The author further recommends, pending the issuance of such regulation /s, that the NRC in the public interest countenance the common practice of " fronting," by means of which a mutual insurer arranges to provide the insurance through a stock insurer. Tha latter may or may not be a subsidiary of the former.
Avoidance of Accident.
Finally, the author recognizes that the insurance mechanism is fragile and highly vulnerable to increases in insured loss frequency and/orseverity. Another serious accident or a run of smaller accidents in the near future could do irreparable harm to the nuclear property insurance market.
He recommenda, therefore, that the NRC remain diligent in its calling, namely, to promote public health and safety in the use of nuclear energy.
'Ferhaps any federal regulation on the subject necessarily would preempt con-flicting state and local law. The author leaves this matter to legal experts.
xvii
I.
NA'IURE OF THE S'IUDY The General Public Utilities Corporation (CPU) faces extra expenditures of at least $1.6 billion dollars as the result of the accident that occurred on March 28, 1979 to its Three Mile Island Nuclear Unit No. 2 Decontamination of the plant is the major cost, amounting perhaps to close to $1 billion. Other elements of the estimated $1.6 billion include damage to the plant itself, fuel replacement, maintenance of the plant during the decontaminating process, and restart activity. The $1.6 billion figure does not include the cost of purchas-ing electricity from substitute sources in order to service customers or the increase in the financing cost that is attributable to GPU's damaged credit rating. Some of these expenditures, for example, the extra costs of borrowing money, are not considered to be insurable. Others, such as decontamination and physical damage to the plant, are insurable.
GPU had only $300 million of insurance applicable to the accident.
This figure was the maximum amount of insurance generally available at the time to respond to cleanup of a nuclear utility's plant and to repair or replacement of damaged property belonging to the utility. Conventional wisdom in some segments of the utility and insurance industries was that $300 millien was perhaps about enough. No previous accident had occurred to test such a figure. Apparently, the magnitude of the decontamination costs associated with the Three Mile Island accident has been an unpleasant surprise to many utility and insurance experts.
Since 1978, the amount of insurance generally available to cover nuclear plant
' Statement by the Treasurer of General Public Utilities Corporation, to A.omic Industrial Forum Insurance Conference, New Orleans (September 21,1981).
This figure does not include the insurance CPU also carried to cover its liability to members of the public for bodily injury or property loss arising out of the accident. This report, however, does not pertain to liability insurance.
1
' 7.
a s
1 2
damage and on-site decontamination has increased; however, it is still substan-tiallybelowtheamountthathindsightrevealsahhavingbeenneededatThree i
Mile Island. Presumably, an equal or larger amount could be needed by other i
i nuclear utilities.
i This report is addressed to the topic of nuclear property insurance. This 4
insurance, distinct from liability insurance, is applicable-to. decontamination of nuclear utility plants, to repair or replacement of damaged or destroyed utility property, to debris removal, and to otber relat'ed' activities undertaken in response to accidents. The insured perils include hot are not confined to radioactive contamination at plant sites of nuclear utt,11 ties.
& Questions A_ddressed M this Report More specifically, the research underlying this report was undertaken to
.I
+
answer the following questions:
1.
What has been the development of each principal source of nuclear property insurance used as of early 1982 by nuclear utilities in the United States?
2 What are some of the distinguishing features of nuclear property s.
insurance as offered by the principal sourced?
l 3
How much nuclear property insurance was offered by each of these j
sources as of January 1, 1982?
4 Assuming that present plans come te fruition, how much nuclear property l
insurance is likely to be offered by each of;these sources as.f
/,
January 1. 1983?
r 5
What, if any, new principal sources of nr21 ear property insurance are likely to emerge in the private sector b'y January 1, 1983?
6.
What problems with regard to nuclear property insurance warrant action by the United States Nuclear Regulatory Commission (NRC) and what action should NRC take in response to each problem?
e-
j 3
1 I
i B.
Definitions i
In this report the following respective definitions are used in the interest l
I of minimizing Ambiguity:
1.
Ibclear utility is taken to mean an organization licensed by the U.S. Nuclear Regulatory Commission to construct and/or operate one or more nuclear-powered units for generating electricity for sale and any other organization having an ownership interest in such unit /s.*
2 Nuclesr property insurance is defined as any system of collecting premium money from two or more nuclear utilities and of using the moneytopayfordamageto,destructionof,and/ordecontaminationof property of any such utility or utilities and expenses related thereto caused by nongradual radioactive contamination (not intended by an insured) from a source on~an insured's premises. The system allows also for expenses, taxes,ard profit incidental to the collection of premiumsandpaymentof,Insuredlosses. The system may but is not required to include payment for damage to, destruction of, and/or decontamination of nuclear utility property arising out of one or more perils other than radioactive contamination. Nuclear property insurance is distinct from nuclear liability insurance. The latter covers the legal liability of insured nuclear utilities.
3 Principal source refers to (a) the joint enterprise operated by American Nuclear Insurers and the Mutual Atomic Energy Reinsurance Fool (ANI-MAERP) as a part of the conventional insurance market, (b) Nuclear Mutual Limited (NML), a mutual insurance company domiciled in Bermuda, whose members Nuclear property insurance may also be sold to suppliers and transporters of nuclear material or equipment, to nuclear fuel fabricators and reprocessors, and to waste disposal and storage facilities. This report pertains only to insuring of nuclear utilities..
4 are some but not all of the nuclear utilities as defined above, and (c) Nuclear Electric Insurance Limited (NEIL), another mutual insurance company domiciled in Bermuda, whose members are some but not all of the nuclear utilities as defined above. Some are members of NML: some are not.
For purposes of question 5 principal source refers also to any other program comparable in magnitude or potential to any one of these that could come into being.
4 Private sector means that portion of the insurance market in which insur-ance is offered for sale by an insurer that is not a part of any govern-ment and that is not owned or operated by any government.
II.
DESCRIPTION OF PRINCIPAL SOURCES OF NUCLEAR PROPERTY INSURANCE A brief account is presented of each principal source of nuclear property insurance for nuclear utilities in the United States.
A.
ANI-MAERP Attention is turned first to American Nuclear Insurers and the Mutual Atomic Dnergy Reinsurance Pool. These two organizations cooperate to provide nuclear 4
property insurance.
Origin and Purpose.
In anticipation of the passage in 1957 of the Price-Anderson Act that amended the Atomic Energy Act of 1954, the private sector insurance industry prepared to offer nuclear liability insurance.
(ThePrice-l Anderson Act requires nuclear utilities to provide spccified financial protection.
l Liability insurance is an acceptable and commonly used method of meeting this requirement.) While the insurance industry was so enga6ed, it also created a mechanism for offering nuclear property insurance to nuclear utilities. With respect to nuclear property insurance the Nuclear Energy Property Insurance Association (NEPIA) was created to channel to nuclear utilities the insurance 42 U.S.C. 22210 (1970).
- 42 u.S.C. 3g2011-2281 (1970).
l l
t
5 to be provided by stockholder-owned insurers.
Similarly, the Mutual Atomic Energy Reinsurance Pool (MAERP) was created to marshall the resources of the mutual insurance companies interested in nuclear property insurance." In 1974 NEPIA became a part of the Nuclear Energy Liability and Property Insurance Association (NEL-PIA), which was renamed as American Nuclear Insurers (ANI) early in 1978. Over the years MAERP has retained its original name.
ANI and MAERP constitute the insurance industry's coordinated response to the need for nuclear property insurance.
The stated joint aim of these two organizations is to provide fixed-premium insurance "at a fair price" to the users and "at a reasonable profit" to the participating insurers.
Insurance companies that participate in ANI or MAERP wrote about 70 percent of the premiums for property-liability insurance sold in the United States in 1981.
Method of Operation--ANI. ANI is a voluntary, unincorporated association of insurance companies.
It operates four separate pools. One pool provides nuclear liability insurance in the United States another provides nuclear liability insur-ance and nuclear property insurance on Canadian installations: another provides nuclear property insurance abroad and nuclear liability abroad except for Canadas and still another provides nuclear property insurance in the United States. Only the last-mentioned pool is subject to detailed attention '.n this report.
The stockholder-owned insurers also created the Nuclear Energy Liability Insurance Association (NELIA) to provide nuclear liability insurance to nuclear utilities and others.
.*The mutual sector also created Mutual Atomic Energy Liability Underwriters (MAELU) to provide nuclear liability insurance.
ANI-MAERP policies are the only substantial source of nuclear property insur-ance as provided by what is commonly thought of as "the insurance industry." Other property insurance policies normally are subject to a broad exclusion of loss caused by radioactive contamination.
- ss.See Insurance for the Nuclear Industry (New York:
Insurance Information Institute, 1982), p. 1.
..... Estimate as made by an officer of one of the ANI subscribing companies.
6 Technically, ANI is not a nuclear property 11 surer but is rather a joint underwriting ancociation.
Its nethod of operation is comewhat complex. The insuring is done by member companies of ANI.
To understand the insuring process used, one must first appreciate a char-acteristic about state regulation of insurance.
Each state has its own set of laws to specify the conditions that must be met before agents, employees, or other representatives of an insurer may operate an office, visit prospects, or otherwise maintain an official presence in the state.
Insurers that have met the conditions set forth in these laws are eligible to receive a license. The act of licensing is said to constitute " admission" of the insurer into the state.
The state into which an insurer has been admitted thereby has the authority to regulate many of the activities of that insurer in that state.
If an insurer has not been adm3.tted into a given state, it can sell insurance in that state only by use of the mail, telephone, or other means that involve no physical presence of the insurer in the state. The insurer is said to be "non-admitted." With respect to such an insurer, the given state has little or no control, except that it may exercise control over a resident insurance broker (representative of an actual or prospective insured of that insurer) who serves as the liaison between that insurer and its customer or prospect.
ANI has chosen to use only " admitted" insurers to sell nuclear property insurance. As of this writing, about forty of its member companies had been admitted into (that is, had been licensed to operate) in all states.
These A given state may issue licenses to insurers domiciled in that state and to insurers domiciled in other states.
..Some states go further, and, as the price of admission, attempt to regulate the activity of the company outside as well as inside the state.
State regulation extends to ratenaking, underwriting, investing, accounting, financial reporting, auditing, and numerous other activities.
... Over time the number of ANI members admitted into all states has changed and in likely to change again.
7 forty or so companies literally do the insuring by subscribing in respective percentages to the insurance that is provided in a common policy with each of them as an insurer. These companies are liable (in their respective proportions) up to the policy limit to the insured nuclear utility for any insured loss. The liability is several and not joint.
Each menber company of ANI (including the forty or so as above), upon being accepted into membership, signs a declaration of participation that obligates the company to pay to ANI, up to a specified maximum amount, its share of any insured loss the admitted companies are committed to pay to a nuclear utility that purchased the ANI insurance. ANI, in turn, agrees to collect the premium and, on behalf of the forty or so companies, to pay the insured losses. Upon collecting the premium on behalf of the forty companies, ANI distributes part of it to the members (the 140 or so) that signed the declarations of participation.
uses another part of it to pay ANI operating expenses, and uses still another part to buy reinsurance.
The reinsurers, in consideration of the premiums paid by ANI, agree to pay ANI for the portion of insured losses that is reinsured. The money paid by the reinsurers in the event of a loss has the effect of reducing the amount of money that has to be paid for the loss by each ANI member company out of that company's own resources.
In the process ANI serves as the collector and disburser of funds, the record keeper, and the general manager of the nuclear insurance enterpriso.
ANI buys some of the reinsurance from nuclear property insurance pools (similar to ANI) located abroad, some from several syndicates of Lloyd's of London,
- That is, no company is liable for the defaults of another.
- Each member receives a proportionate share equal to that member's share in the exposure that is insured.
8 and some from individual foreign reinsurers.
ANI also buys reinsuranco from MAERP, and vice versa.
Method of Operation--MAERP. MAERP, as its name indicates, is a pool of reinsurers. The reinsurance applies to each nuclear property insurance policy that is issued to a nuclear utility by a member company of the MAERP pool. ***
The company that issues the policy becomes directly liable to the insured up to the applicable limit of the policy for any insured loss'.
The policy is then reinsured by the members of the pool in the respective proportions agreed upon in advance by the companies participating in the reinsurance pool. Each rein-surer is severally liable to the primary insurer for the reinsurer's share of the loss.
A portion of each MAERP exposure is reinsured abroad, with the same rein-surers taking the same percentage of each exposure as would occur had the expo-sure been handled in ANI. MAERP also routinely reinsures a standard percentage of each exposure with ANI. As observed above, MAERP serves as a reinsurer in accepting a standard percentage of each ANI exposure as ceded by ANI. At the outset in 1957 102 mutual companies were members of MAERP. As of January, 1982 MAERP reported having 121 members. As of January 1, 1982, administration of MAERP was assumed by the MAERP Reinsurance Association. The latter organization was created to facilitate day-to-day operations.
l l
l Foreign" in this context means domiciled outside the United States. More information about ANI's purchase of foreign reinsurance is presented below.
See the following discussion for the percentages involved.
The bulk of the insurance was wri,t+en by eight large members of MAERP.
An exception is that six insurers operate as a group with each of the six acting as a subscriber for a percentage of the insurance issued by the six in a common policy. This subscription technique is similar to that used by ANI.
9 Coordination between ANI and MAERP. Activities of ANI and MAERP are highly integrated. The result is that for practical purposes the nuclear utility using ANI-MAERP insurance can regard ANI-MAERP as the insurer. The integration men-tioned above is manifested in the following ways, among otherst 1.
Every nuclear utility that is covered through ANI is also covered through MAERP and vice versa. The mechanism by means of which this outcome is accomplished is reinsurance. With respect to exposures located in the United States, MAERP participants provide reinsurance for every policy initiated at ANI and ANI participants provide rein-surance for every policy initiated at MAERP.
As of January 1, 1982 ANI took 75 7 percent of each MAERP policy exposure and MAERP took 24.3% of each ANI policy exposure.
When foreign participation, as discussed above, was taken into account, the net capacity remaining with ANI members was 34.7 percent and with MAERP members was 11.2 percent.
The remaining 54.1 percent was indirectly supplied by foreign nuclear pools (37.6 percent), Iloyd's syndicates (10 percent),
and individual foreign reinsurers (6 5 percent).
2 ANI and MAERP use policies that except for ideratification of insurer /s are almost identical in essential provisions.
3 For any given nuclear utility the same price per $100 of insurance applies to insurance issued through MAERP as through ANI. The Nuclear Insurance Rating Bureau (NIRB) establishes the rates for nuclear property insurance. Managed by the Insurance Services Office, NIRB draws its members and actuarial expertise from member companies of both ANI and MAERP.
- Information provided in a January 5,1982 letter to the author from ANI.
" Ibid.
10 4
ANI-MAERP operates a Joint Engineering Committee to exercise loss control.
One aim of the committee is to bring to bear on nuclear utility exposures the best loss control knowledge within the insurance industry. Some of it is peculiar to nuclear technology much of it is not. ANI has a staff of eight regional engineers to perform the " Fire and All-Risk" portion of the insurance inspection function, that is the portion that excludes vessels under pressure and machinery with moving parts.
MAERP calls upon the staff of the factory mutual insurers and the extensive support services of these insurers for similar inspections. The " Pressurized Systems Machinery and Code Inservice Inspections" (inspectiors of vessels under pressure and machinery with moving parts) are conducted by any one or more of caveral insurers, including the Hartford Steam Boiler Inspection and Insurance Company.
5 Consistency as between MAERP and ANI is also maintained in underwrit-ing and claims adjusting. MAERP and ANI have underwriting committees that work closely with each other. A joint loss adjustment committee establishes procedures for dispost ng of claims. The Industrial Risk Insurers adjust the claims for ANI and MAERP. The practical effect of this cooperation is that with regard to the two activities the two organizations almost operate as one.
Insuring Activities and Operating Results. As of January 1,1982 ANI-MADlP insurance in the United States extended to fifty nuclear units at thirty-four power plant sites. Of the fifty units, thirty-nine were in operation, six were In insurance parlance " loss control" refers to efforts by the insurer and/
or insured to minimize the probability of the occurrence of an insured loss and to minimize the severity of any such loss that does occur.
l l
11 under construction, two were being decommissioned, and three were completed reactors that contained nuclear fuel awaiting activation.
Exhibit i summarizes the gross written premium and the gross amounts incurred for losses, loss expenses, and other expenses (usually called under-writingexpenses). This exhibit shows that, without attention to investment earnings, the cumulative result of ANI-MAERP's United States nuclear property insurance operations from 1957 through 1981 has been a substantial deficit.
One common measure of underwriting performance used in the insurance industry is the " underwriting ratio," which for a given period is the dollars I
of incurred losses, loss adjustment expenses, and underwriting expenses expressed as a percentage of the dollars of earned premium.
An underwriting ratio equal to 100 indicates that the insurer has " broken even" on its insuring activities, apart from any investment gain or loss it experienced in its custody
- Additionally, ANI-MAERP provided insurance for three fuel storage facilities, thirteen testing and research reactors, fifteen fuel fabricators, one fuel repro-cessor, and eight laboratories or other test facilities. Additionally, builders risk policies covered thirty-three units under construction at nineteen sites.
American Nuclear Insurers Reports, Report #1 (Revised March 3,1982), p. 4
- ANI-MAERP uses the term " loss expenses" in a manner roughly comparable to the use of " loss adjustment expense" in customary insurance terminology to mean expenses directly allocable to adjusting an insured loss. An example would be the incurred compensation for a person engaged in ascertaining the dollar amount of the insured loss. Loss expenses are in contrast to " underwriting expenses" that include administrative and selling expenses as well as unallocated loss adjustment expenses. An example of underwriting expenses would be rent or depreciation expenses for an office used by ANI or MAERP.
- The figures in the exhibit do reflect the reduction in premiums arising out of the Nuclear Property Insurance IndustryExperience Guide, as described in Setbo III,A.
- A popular variation of the underwriting ratio is the " combined ratio." For practical reasons not germane to this discussion the combined ratio is stated as f1 ws:
Incurred Losses + Loss Adjustment Expenses + Incurred Expenses Earned Premiums Written Premiums The reasons warranting use of this ratio in other circumstances do not apply to nuclear property insurance.
12 EXHIBIT 1 DOMESTIC NUCIEAR PROPERTY IN3URANCE UNDERWRITING RESULTS BY YEAR AND CUMULATIVE ANI-MAERP 1958-1981 Incurred Total Earned Year gg33,,a loss Unde m h "
f Premium "
By Year c
b Cumulative
- Expenses, Expenses 2, 3, and 4a Since 1957 a
Col. 1 Col. 2 Col. 3 Col. 4 Col. 5 Col. 6 Col. 7 Col. 8 1958 0.1 0.1 0.2 03 66.7%
66.7%
1959 0.1 0.1 0.2 05 40.0 50.0 1960 1.1 0.4 1.5 1.6 93 8 79A~
d 1961
- 0.i 03 0.2 1.'6 12 5 52 5 1962 0.2 03 05 1 9-26.3 44.1 1963 0.4 03 0.7 2.4 29 2 39.8 1964 0.2 03 05 2.6 19 2 34 9
~
~~
1965 0.2 0.4 0.6 2.8 21.4 32.1 1966 2.7 0.4 31 29 106.9 45 2 1967 19 0.6 25 3.4 73 5 50.0 1968 1.0 0.6 1.6 37 43.2 48.9 0.7 0.8 15 4.7 3E9 46.1
_1969_
1970 13 1.1 2.4 6.7 35 8 44.2 971 58 0.2*
1.8 78 11.4 68.4 50.1 1
1972 3.4 2.7 6.1 15 8 38.6'~
47 2 1973 351 o'.I 1.8 50 13.4~'
- ~-
37 3 45.4 1974 1.7 2.0 37 15 A ~ ~l5470 41.8
~
1975 4.7 0.1 2.2 70 20.4 34 3
~
40.4 1976 54 0.2 2.3 79 25 5 31.0 38 7 1977 8.8 0.2 2.3 11.3 26 5 42.6 39 3 1978 30.6 0.4 19 32.9 29 3 112 3 50.4 1979 248.2 1.9 2.8 i'52.9 36.5 692 9 152.6 1980 116.4 1.4 4.2 122.0 51.6 236.4 168.1 1981 27 9 0.8 54 34.1 70.0 48.7 144.2 Total me for 465 8 53 35.1 506.2 350 9 V
144.2
//
Periodi a In millions of U.S. dollars, with some inaccuracies because of rounding b Col. 5 divided by Col. 6 e Curaulative total of Col. 5 divided by cumulative total of Col. 6 d Negative apparently because of correction of figures for earlier year /s e For 1958 through 1971 Source: Data supplied to author by ANI and MAERP.
l t
13 of funds provided by policyholders, stockholders, and/or creditors. An under-writing ratio less than 100 indicates an underwriting profit. The lower the ratio, the higher the rate of profitability. Conversely, an underwriting ratio larger than 100 indicates an underwriting loss.
A ratio above 100 is normally the cause of insurer concern.
In recent years the extraordinarily high levels of interest rates have emboldened some insurers to allow underwriting ratios to creep somewhat above 100. These insurers depend on heavy investment income to compensate for expected underwriting losses and to leave an ample margin for what insurers call " operating profit." Even the most emboldened insurer, however, would be likely to become apprehensive about an underwriting ratio that approached, say, 110. The ANI-MAERP cumulative under-writing ratio of about 144 for the 1958-1981 period for domestic operations is distinctly atypical in terms of insurance industry normative conduct.
Interest-ingly, even the ANI-MAERP cumulative loss ratio (incurred losses and loss expenses over earned premiums) exceeded 100 for the 1958-1981 period, for domestic operations, a condition normally regarded as alarming by insurance company financial analysts.
Exhibit i shows that the cumulative loss ratio was drastically influenced by the Three Mile Island accident. The impact of this accident on the cumulative underwriting ratio will require several years of favorable experience to over-Come.
Analysts point out, however, that the deterioration in the underwriting ratio was not entirely attributable to the Three Mile Island episode. Exhibit 1 indi-cates an increase in incurred losses that manifested itself sharply in 1978, the year before the large loss at Three Mile Island.
In 1978 the loss ratio was higher than 90 percent. This deterioration was attributable largely to conven-tional perils and not to radioactive contamination.
Two general types of insured losses have been particularly troublesome in both frequency and severity.
14 One type it lossinthemovingofmaterialsorequipment(includingthematerials handlingequipment). The other is loss to transformers.
Although these losses have not arisen out of nuclear accidents, they have created problems for ANI-MAERP.
Giving consideration to investment income brightens the ANI-MAERP financial results somewhat. Column 4 of Exhibit 2 shows the effect of adding net invest-ment income to underwriting gain or loss.
This operating gain or loss is per-haps the closest approximation to conventional year-by-year operating results in a noninsurance business enterprise that available data permit one to make.
Again, the exhibit pertains only to domestic operations.
At least two serious difficulties, however, are associated with placing undue credence on the operating gain or loss results reported in Exhibit 2 One such difficulty--treated at more length in Section VII, D of this report--is the lack of statistical stability to be found in the year-by-year underwriting results.
This condition is attributable to the possibly high year-to-year volatility in loss frequency and severity. This possibility of volatility stems at least in part from the small number of nuclear units exposed to insured loss. Predictions of insured losses necessarily are loose and subjer.t to large error. Loss results for any one year are not necessarily indicative of loss results to be expected in any other year. Consequently, the rates for the insurance may prove to be inadequate.
The other serious difficulty to be mentioned is that neither ANI nor MAERP was designed to invest funds and accumulate assets. Both were originally per-ceived as essentially pass-through mechanisms with respect to funds not immedi-ately needed to pay losses or expenses. Thus, whatever investment income might
- As explained in Section III, A, the coverage provided by ANI-MAEHP includes conventional perils as well as the peril of radioactive contamination. Without the stabilizing influence of the flow of premiums from non-nuclear perils, the insurers might not be willing to insure against radioactive contamination at all.
l l
15 EXHIBIT 2 UNDEPWRITING CAIN OR IDSS PIDS NET INVESTMENT INCOME AS A PERCENTAGE OF EARNED PREMIUM ON DOMESTIC NUCLEAR PROPERTY INSURARANCE OPERATIONS BY YEAR AND CUMULATIVE ANI-MAERP 1958-1981 Operating Gain or Loss TM Net as a Pert:entage of Underwriting Operating Earned a ed h emium Year a
Investment Cain or Loss" Premium Cain or Loss Incomea D
"""1" (Col. 2 + Col. 3)
By Year Since 1957 o.91. 1 Col. 2 Col. 3 Col. 4 Col. 5 Col. 6 Col. 7 1958 0.1 0.1 03 33 3%
33 35 03 05 60.0 50.0 1959 03 0.1 1.6 6.2 20.8 1960 0.1 1961 1.4 1.4 1.6 87 5 47 5 1962 1.4 1.4 1.9 73.7 55 9 1.7 2.4 70.8 60.2 1963 1.7 2.1 2.1 2.6 80.8 65 1 1964___
,_ j 1965 2.2 0.1 2.3 2.8 82.1 68.6 1966
- 0.2 0.1
- 0.1 2.9
- 3.4 56.0 3.4 29 4 51.5 1967 09 0.1 1.0 1968 2.1 0.2 2.3 37 62.2 53.2 0.2 34 4.7 72.3 56.3 1969
. -. _3 2 1970 4.3 03 4.6 6.7 68.7 58 7 1971 3.6 0.2 3.8 11.4 33 3 52 5 1972 97 0.4 10.1 15.8 63 9 55 4 1973 8.4 0.6 90 13 4 67.2 57 5 1974 11.7 1.0 12.7 l
15.4 82 5 61 7 1975 13.4 1.1 14 5 l
20.4 71.1 63 4 1976 17.6 1.4 19.0 f~25.f 74.5 65 5
~~
1977 15.2 1.7 16.9 26.5 63.8 65 2 1978
-36 30
- 0.6 29 3
- 2.0 55 0 1979
-216.4 4.5
-211 9 i
36 5
-580 5
-46.2 1981
~ ~j5.i) 70 42.f -
~p0.6 ~ ' ~ 6F.3 ~'
-36 5~
~ ~ ~
' ~ ~ -
~~
Total 1
Z 4
for
-155 3 27.2
-128.1
! 350.9
?
-36 5 Period a
In millions of U.S. dollars, with some inaccuracies because of rounding b
Col. 4 divided by Col. 5 Cumulative total of Col. 4 divided by cumulative total of Col. 5 c
Source: Data supplied to author by ANI and MAERP.
16 be earned on premium monies was visualized as being earned by the participating companies and not by ANI or MAERP. Such investment earnings by the participat-ing insurers were not identifiable by the author. The extent to which they are included in Exhibit 2 is problematical.
In 1972 in an action reflecting a signi-
~
ficant change 1n operating procedure, NEPIA began investing certain funds. ANI has continued this practice.
MAERP, however, has continued to pass its share of investable funds back to its participating companies.
Exhibit 2 represents an effort to recognize the investment income attributable to funds held by or for ANI and MAERP. The figures, however, are only estimates and may be understated, particularly as to ANI.
In Columns 6 and 7 operating gain or loss is expressed as a percentage of earned premiums for each year and for the cumulative operations since the incep-tion of ANI-MAERP. This ratio is roughly comparable to the rate of return on sales in a noninsurance enterprise. One's assessment of the operating results of ANI-MAERP depends on the base against which one draws a comparison.
In this author's view the profitability performance of domestic operations of ANI-MAERP for the cumulative period of its operations--even with the most generous allow-ance for the understatement of investment income is Exhibit 2--would have to be judged as less than enviable by any reasonable standard of profitability.
1 4
ANI remits nuclear property insurance premiums on a quarterly basis to member companies. The author does not understand precisely what funds as derived from nuclear property insurance are invested by ANI.
No effort was made by the author to estimate realized or unrealized capital gains or losses experienced by ANI or experienced by the ANI or MAERP participat-ing companies on investments related to funds generated by the sale of nuclear property insurance.
17 Ubatever the reason may have been for the financial performance to date, the nuclear property insurance of donestic exposures by ANI-KAERP thus far has not been a substantial source of profit to the participating members of ANI and MAE9P.
One night try to make a case to the contrary by arguing that the results could have been significantly different had the Three Mile Island accident not occurred or had loss control for losses from non-nuclear perils been more effective. Such an argument, however, would be merely wishful thinking.
The Three Mile Island accident did occur and the frequency of non-nuclear losses has been substantial.
B.
Nuclear Mutual Linited Another principal source of nuclear property insurance in the United States is Nuclear Mutual Limited (NML) to which several descriptive comments are addressed.
Origin gd Purpose.
IE!L is a mutual insurance corporation established under Bermuda law with its home office in Hamilton, Bemuda.
NML began its insuring activities as of January 1,1973 m a sells only nuclear property insurance and serves only its nenbers who are nuclear utilities in the United States. The tiML insurance applies to construction and operation of nuclear generating plants and to nuclear property in transit to and from such plants.
NML's stated purpose as a mutual company is:
- By contrast, reinsuring of foreign nuclear property exposures has to date been profitable, assuning no offsetting allowance for possibly catastrophic future lonces.
- The insured costs of the mishap on January 25, 1982 at the cinna plant of the Rochester Gas and Electric Corporation also will fall upon ANI-MAERP. At this writing the magnitude of the insured loss was not known. " Rochester G & E to Begin Repair of Nuclear Plant," The Wall Street Journal (January 27, 1982),
- p. 7.
Doubtless substantial decontanination expenses will be incurred.
- See "The Nuclear Mutual Company Act,1971," as amended on May 18, 1972, by the Legislature of Bermuda.
18 to obtain for its members the economic benefits of insurance coverage at a cost equal to their share of the Company's losses and administrative expenses.
iML is on occasion referred to as a nuclear utility industry "self-insurance captive."
In this author's opinion the " captive" portion of the label fits-reasonably well but the "self-insurance" portion is inaccurate.
NML is a captive in the sense that it was created for the express purpose of serving only nuclear utilities. Even though it serves only some and not a11' nuclear utilities, it is an instrument for use by at least a segnent of an industry and in that sense is properly labeled as an industry captive.
UML, however, does not engage in what is commonly thought of in the insur-ance community as self-insurance.
Self-insurance is a vague concept with numerous shades of neaning, but it invariably has to do with retention by some economic entity of its financial exposure to accidental loss. T;11s retention is in con-trast to the transfer of this financial exposure to some other economic entity, as takes place when insurance is bought.
By its nature, insurance involves a sharing of losses among two or more economic units. By serving more than one nuclear utility, NML brings about a change in the incidence of losses among the utilities.
The losses subject to NHL policies do not rest totally on the economic entities on which they fall but are redistributed, at least in part.
The utilities that enter contracts with UML are distinct economic entities, largely independent from each other. The product of NHL's efforts, therefore, l
1s not self-insurance but insurance.
If UML had been created by and served only I
" Summary o_f Operations, Nuclear Mutual Linited (August,1981), p.1.
For a discussion of whether self-insurance is necessarily any different fron the simple absence of insurance, see John D. Long, " Soft Spots in Insurance l
Theory," in John D. Long, Bd., Issues in Insurance (2d. ed. ; Malvern, PA:
l American Institute for Property and Liability Underwriters, Inc., 1981),
l pr. 372-380.
1 l
19 one utility, it might more appropriately be labeled a self-insurance captive.
By serving numerous utilities, NML meets the necessary and sufficient criteria to be classified as an insurer.
The fomation of NML may have reflected the interest of numerous nuclear utilities in pursuing two principal objectives. One objective was to reditee outlays for nuclear property insurance.
The excellent loss experience of ANI-MAERP in the early years of operation caused some nuclear utility decision makers to think that the price of nuclear property insurance could and should be reduced. The formation of NML in 1973 apparently was viewed as a way to achieve such a reduction.
The other objective, entertained by some nuclear utilities, was to retain a substantial portion of the monies ultimately to be paid as premiums to the insurer until the monies were needed by the insurer to pay the insured losses. This objective perhaps led to the adoption of the assessment feature of NML insurance, a feature that is discussed elsewhere in this report.
- "Five Utilities Mull Forming a Nuclear-Plant Insurer," The Wall Street Journal (February 4,1971), p. 30.
- Perhaps the introduction by ANI-MAERP in 1972 of the Nuclear Property Insurance Industry Experience Guide that allows premium discounts for favorable loss experience was motivated in part by the unrest among insureds. This rating system is discussed in Section III of this report.
- According to NML's legal counsel, another and perhaps the overriding moti-vation for establishment of NML grew out of a philosophical conviction that was held by numerous officials in the nuclear utility industry. This conviction was that, whatever might be the nature of the nuclear property insurance market, nuclear utilities would have to bear their own insured losses. They should not expect, so the conviction went, that accidents at nuclear sites would be subsi-dized from premiums from other lines of insurance. Given this expected absence of subsidy from other lines, a sentiment developed among various nuclear utili-ties that the nuclear utility industry night as well accept the responsibility directly and devise an efficient mechanism for spreading its insured losses strictly among the industry's own members.
20 Method of_ Operation. Being an insurer, NML aust perform the usual insurer functions of marketing, underwriting, ratemaking, loss adjusting, reinsuring, and investing in addition to exercising loss control. The company has entered a con-tract with Marsh & McLennan (Benuda) Ltd. for the latter to perform as a con-sultant various ratemaking, loss adjusting, and other services, including the inspecting and evaluating of nuclear utility properties under consideration for or already subject to NML insurance.
To be eligible for membership in NL an applicants must have an insurable interest in such nuclear or other risks as the company may from time to time offer to insure and must i
meet such reasonable underwriting standards as are established by the Directors of the Company.*
Because NHL insures only exposures of nuclear utility plants plus the interests of such utilities in nuclear material in transit to and from plant sites, all eligible members are nuclear utilities.
NML requires that an applicant be " capable of meeting its financial obliga-tions to the Company."
NML also reserves the option to require as a condi-tion for approval of the application that the applicant submit " guarantees or such other appropriate undertakings" as NML may request.
Being an " offshore," nonadmitted insurer, NML is not in a position to main-tain an office, hold meetings, or otherwise conduct its business in the United States. An applicant for insurance is free to conlact the NML office in Bemuda directly or through any intermediary (broker) of its choice. NML has no agents.
The activities pertinent to the insuring process that have to be done at the Ibid., p. 2 As of August,1981 NML also insured four " conventional power facilities" of utilities whose nuclear power facilities were also insured by NML.
- Summary of Operations, p. 2 Ibid.
l l
l
21 plant sites or elsewhere in the United States are done by Marsh & Mclennan (Bemuda) Ltd. or possibly by some other NML consultant. NML is operated by a general manager and a small staff in Bermuda.
NML's interest in a foreign domicile is apparently twofold:
(1)toescape the regulation imposed by state insurance departments on admitted insurers, and (2) to avoid the federal and state taxes imposed upon domestic insurers."
For purposes of the latter, NML assiduously avoids activity that might cause it to be deemed to be conducting a " trade or business" in the United States.***
The governance of NML is by a board of directors elected by the members.
Each member is entitled to one vote for each $10,000 of contingent liability for assessments (as discussed in Section III, B of this report) that could be levied by NML. The additional proviso is imposed that no member is to be entitled to exercise more than 9 percent of the total number of votes.
- This regulation as observed earlier, extends to rates, investments, accounting, annual statements, audits, trade practices, and numerous other operational subjects.
"The Bermuda Exempted Undertakings Tax Protection Act of 1966 (as amended in 1973) provides that no Bermuda income tax (including capital gains tax) can be levied upon exempted companies before March, 2006.
Section 18 of The Nuclear Mutual Company Act,1971 of the Bermuda legislature (the act establishing NML) specifies that "the Company (that is, NML) shall be deemed to be an exempted company within the meaning of the Exempted Companies Act of 1950, and the Exempted Undertakings Tax Protection Act,1966, and the provisions of those Acts shall apply to the Company as if it were within the definition of ' exempted company' contained in subsection (1) of section 1 of the Exempted Companies Act, 1950."
... Internal Revenue Code, Section 842 specifies that if a foreign corporation meets certain conditions and conducts an insurance business in the United States, the income derived from such activity is subject tc being taxed.
Section882(a) is more general and imposes a tax on the income of a foreign corporation to the extent that the income is " effectively connected with the conduct of a trade or business within the United States."
If the foreign corporation does not conduct a trade or business within the United States, its United States federal income tax liability may apply only to income or capital gains, if any, from the cor-poration's investments, patents, copyrights, or certain other properties in the United States.
22 I
i The reason for this percentage limitation is found in the portion of the Internal Revenue Code that pertains to " controlled foreign corporations."*
The concept of this portion of the tax law is that, if a foreign corporation is owned primarily by one or a few United States taxpayers, the undistributed income of that corporation is treated as an imputed dividend (and, therefore, as includable at least in part in the gross income) of the taxpayers in the United States.
A foreign mutual insurance company meeting certain other conditions is a
" controlled foreign corporation" in case more than 25 Percent of its voting power rests with " United States shareholders."
With regard to a foreign mutual insurer, a " United States shareholder" includes a " United States person" with ability to exercise 10 percent or more of the total combined voting power of the foreign mutual insurer."**
A " United States person" in this context includes a public utility domestic corporation.
To make sure that the nuclear utilities buying insurance from NML do not run afoul of the " controlled foreign corporation" income tax pitfalls, the designers of NML placed a limit of 9} percent on the relative voting power of any nuclear utility, no matter how many units the utility might insure in NML. With this limitation, no NML insured need face the risk of becoming liable as a " United States shareholder" of a
" controlled foreign corporation" for payment of federal income tax on any undis-tributed NML profits.
- The Revenue Act of 1962, by adding Subpart F of Subchapter N to the Internal Revenue Code, made several types of undistributed income of a " controlled foreign corporation" taxable to the entities in the United States who controlled the corporation by stock ownership, or by voting rights in the case of a mutual cor-poration. IRC, Sections 951-964 Provisions specifically applicable to income of foreign mutual insurers of " United States risks" are Sections 952(a)(1), 953, and958(a)(3).
- IRC, Section 957(b).
- IRC, Sections 951(b) and 958(a)(3).
- IRC, Sections 957(d) and 7701(a)(30).
l
23 This method of operation is not totally effective, however, in avoidance of taxes levied by taxing authorities in the United States. A federal excise tax becomes payable when an exposure located in the United States is insured by a foreign corporation that does not conduct an insurance business in the United States.
With respect to insurance bought from ?ML the tax falls on the insured nuclear utility and amounts to 4 percent of the premium paid.
In its Summary of Operations, fML mentions that this federal excise tax and any state tax or other state charge " levied on premiums paid to non-admitted insurers will...be the responsibility of the insured members."
!ML uses reinsurance. As of January, 1982, its reinsurance was of the excess-of-loss type and consisted of two plans. One plan applied to the first
$75 million of loss above the first $100 million of loss for any one policy The other, when ccmpleted, will provide another $50 million of reinsurance year.
above $200 million.
This one will pay 100 percent of one or more losses that are incurred in a given policy year and that in the aggregate are in excess of
$200 million but not in excess of $250 million. As of the date of the letter cited above, fNL had succeeded in securing commitments from reinsurers for about 53 percent of the desired $50 million plan.
The following examples show the protection the two plans will afford IML, once they both are complete.
Suppose that in a given policy year a loss occurs for which IML agrees to pay the insured $80 millio'..
No part of that amount will
- IM,Section4371(1).
"Op. Cit., p. 7.
to author from IML.
Letter of January 18, 1982
24 be subject to recovery by NHL under either plan." Suppose that in the same policy year, a utility suffers a loss for which NML agrees to pay $25 million.
NML could look to the reinsurers for $5 million under the first plan because NML would have borne its first $100 million of retained, unreinsured loss.
If NML suffers during the same policy year a third loss of, say, $90 million, NML can recover $70 million from its reinsurers under the first plan, $5 million of the
$75 million having already been paid. Suppose still another loss occurs during the same policy year and that it is in the amount of $100 million. With $195 million of earlier losses having already occurred, the first $5 million of the 100 million loss would be borne by NML because NML's first plan of reinsurance would be exhausted for the year and its second plan would apply only above
$200 million of losses. Under the second plan, the next 50 million of the $100 million loss would be recoverable from the reinsurers, leaving $45 million as i
being above the upper limit of the second plan.
If any further loss occurs dur-ing the policy year, no reinsurance is applicable to it.
Exhibit 3 summarizes the two reinsurance plans.
It shows that only $125 million of the aggregate losses of $250 million or more are subject to reinsur-The reinsurance recoverable would be the same, were one loss of $250 million ance.
or more to occur instead of two or more lossec aggregating $250 million or more.
Incidentally, all NML policies have the common expiration date of March 31.
For any insured the initial policy may be for a period shorter than one year.
All other policies are issued for one year's intended duration.
Whether or not this utility is the same one that suffers the first loss is irrelevant to the examples, i
I
25 EXHIBIT 3 NML'S REINSi1RANCE PLANS 1%2 Aggregate of Incurred Losses in Given Policy Year (Dollars in Millions)
$250 Reinsured in Second Plan 200 Not asinsured 175 Reinsured in First Plan 100 Not Reinsured 0
Source: Summary of Operations, NML (August, 1981), p. 5 and January 18, 1982 letter to author from NML Insuring Activities and Operating Results.
NML began its insuring activi-ties in 1973 with thirty nuclear units at sixteen sites subject to its coverages.
As of March 26, 1982 it insured fifty-two nuclear units (plus four conven-tional units) at twenty-seven sites.
As of the time of this writing, NML had not been subjected to any substantial claim or to any undue frequency of small claims.
As the rssult, its operating results were exceedingly favorable. Again, however, the observation is relevant that operating results based on such a small number of units of exposure are not necessarily indicative of future loss experience.
In fact, the exposure base is so small and the history of NML is so short that few meaningful statements can be made about NML's financial operations to date.
Subject to this caveat, the same types of operating data about NML are summarized in Exhibits 4 and 5 as are su==arized about ANI-MAERP in Exhibits 1 and 2 respectively. Exhibit 4 shows a remarkably low cumulative underwriting
- Letter of March 27, 1992 to author from NML.
26 ratio of about 26 percent." Aside from the first year, this ratio has remained below twenty. The table also indicates that the sum of the incurred losses and loss adjustment expenses for the period was even lower than the underwrit-ing expenses, an unusual situation but perhaps actuarially insignificant in view of the very small number of exposure units insured by NML. Comparison i
of the figures in Exhibit 4 with the figures in Exhibit i does suggest that to date NML's loss ratio (incurred losses and loss adjustment expenses divided by l
earned premiums) has been distinctly better than was ANI-MAERP's loss ratio for the first ten years of ANI-MAERP's operations." Similarly, NML's underwriting expense ratio for the first ten years of its operations was lower than the underwriting expense ratio of ANI-MAERP for the first ten years of ANI-MAERP's operations.
Exhibit 5 summarizes hHL's operating results when net investment income is added to the underwriting data."*
As is done in Exhibit 2, the operating gain or loss is expressed for each year as a percentage of earned premiums. As of the end of each year the cumulative operating ratio (from 1973 to the end of that year) is also presented.
As emphasized already, the past is not a reliable indicator of the future with respect to NML's insured losses.
Had Unit #2 at Three Mile Island been insured for $300 million with NML instead of with ANI-MAERP and had other NML loss experience remained unchanged, the appearance of Exhibits 4 and 5 would For consistency one should add to the NML underwriting ratio the 4 percent of preniums to reflect the federal excise tax paid by nuclear utilities on NML premiums.
Insureds of ANI-MAERP have no comparable charge because the state premium taxes are paid by the insurer and are included in underwriting expenses.
"The discrepancy might be attributable in part to the fact that NML enjoyed a much larger premium volume during its early years than did ANI-MAERP. This larger premium volume caused start-up costs to be relatively less burdensome for NML than was the case with ANI-MAERP.
"* No data were available to the author about NML realized capital gains and losses, if any, except that they have been negligible. Letter of March 26, 1982 to author from NML.
I 27 EXHIBIT 4 UNDERWRITING REULTS BY YEAR AND CUMUIATIVE "
NML 1973-1981 m
Incurred Underwriting Ratio gg sses and Earned I
Year Underwriting of Col.s Cumulative Expenses 2 and 3b hemium By Ye d b
Since 1973d Ex en s I
Col. 1 Col. 2 Col. 3 Col. 4 Col. 5 Col. 6 Col. 7 I
1973
$ 2.4
$09
$ 33
$ 7.1 46.4%
46.4%
1.1 1.1 9.1 12.1 27.2 1974 1975 0.7 1.2 19 11.7 16.2 22.6 1976 0.1 1.2 1.3 14.2 92 18.0 j
1977 0.7 13 2.0 18.1 11.0 15 9 1978
- 0 3' 13 09 19 5 4.6 13 2 l
1979 6.6 1.6 8.2 24.1 34.0 18.0 l
1980 1.8 1.6 3.4 29.6 11.5 16.6 1981 20 5 15 22.0 38.0 57 9 25 7 l
Total
.W +'
for
$ 32 5
$ 11.6
$ 44.1
$171.4 25 7 Period i
/4 a Net of reinsurance transactions b In millions of U.S. Dollars, with some inaccuracies because of rounding c Col. 4 divided by Col. 5 d Cumulative total of Col. 4 divided by cumulative total of Col. 5 Negative apparently because of correction of figures for earlier year /s e
Source: Figures provided in letter of March 26, 1982 to author from ANI i
L
2'8 EXHIBIT 5 UNDERWRITING CAIN OR IDSS PLUS NET INVES'INENT INCOME AS PERCENTAGE OF EARNED PREMIUM BY YEAR AND CUMULATIVE" NML 1973-1981 ra n6 Gain or Mas I'"#
Underwriting Net Total b
operat ng g.ned as a Percentage of Cain or Loss Investagnt b
Earried Premium g,@
By Ye e d
Income c
Cumulative Since 1973 Col. 1 Col. 2 6
Col. 3 Col. 4 Col. 5 Col. 6 Col. 7 1973
$38 l
$ o.5
$ 4.3
$ 7.1 60.6%
60.6%
1974 8.0 i
1.0 90 9.1 98 9 82.1 1975 9.8 1.1 10 9 11.7 93 2 86.7 1976 12 9 09 13 8 14.2 97.2 90 3 1977 16.1 15 17.6 18.1 97.2 92.4 1978 18.6 3.2 21.8 19 5 111.8 97.1 1979 15 9 l
58 21.7 24.1 90.0 95 5 1980 26.2 l
10 5 36.7 29.6 124.0 101.8 1981 16.0 20.1 36.1 38.0 95 0 100 3 Total for
$127.3
$ 44.6
$171 9
$171.4
?
?
100,3 Period a Net of reinsurance transactions l
b In millions of dollars with some inaccuracies because of rounding Col. 4 divided by Col. 5 c
d Cumulative total of Col. 4 divided by cumulative total of Col. 5 f
Source:
Same as for Exhibit 4 i
[
l
29 have been dramatically different. The precise difference is a matter of con-jecture. Several assumptions are needed for indulgence in conjecture. Suppose the following conditions had prevailed:
1.
That NML for purposes of its own accounting would have recognized (i.e., considered as " incurred") $100 million of the Three Mile Island loss in 1979 and the remainder in 1980. Net of reinsurance, the 1980 figure would have been about $125 million.
2 That, as implied above, the reinsurers would have responded by paying
$75 million to<ard the $300 million.
3 That NML's incurred loss adjustment expenses would have increased by 1 percent of $100 million in 1979 and by 1 percent of $125 million in 1980.
4 That NML's premiums (as distinct from assessments) would not have increased for the policy year beginning April 1,1979 and for subse-quent policy years as the result of the Three Mile Island accident.
With these changes, the NML underwriting ratio for 1979 could have been
.n*
about 453 This figure is in stark contrast to the underwriting ratio of 34.0 as reported for 1979 in column 6 of Exhibit 4 For 1980 the revised under-writing ratio would also have been about 438 instead of the 11 5 figure reported
- That is, $300 million, which was the amount of insurance, less the sum of (a) the $100 million recognized in 1979 and (b) the $75 million of minsurance.
NML did not have its second reinsurance plan in 1979 "The ANI-MAERP ratio, as indicated by Exhibit 2, was approximately 1 percent.
"'An increase in 1980 and/or 1981 could have occurred because of one or more rate increases.
"' Incurred losses, loss adjustment expenses, and underwriting expenses of about $109.2 million divided by earned premium of about.$24.1 equals about 4.53 or about 453 percent. The underwriting ratio is a percentage.
j' 30 p,
s in Exhibit 4 The cumulative underwriting ratio as'of the end of 1981 would i
have been about 158 instead of 25 7 as reported in Exhibit 4.
Had the Three Mile Island claim fallen on NML, the operating results, as-reported in Exhibit 5, would have been similarly altered.
For 1979 the operat-ing gain or loss as a percentage of earned premiums in Column 6 of Exhibit 5 would have been about a negative 329 percent for 1980 the figure would have been about a negative 302 percent. The cumulative operating loss as a percent-age of earned premiums as of the end of 1981 would have bee labout a negative 32 instead of the positive 100 3 as reported in, Exhibit 5
~
NML would have had sufficient assets (aside from retrospective premium adjustment) to pay perhaps at'l$ast hsif of the 1979 portion of the Three Mile Islandclaimundertheassunptionss(atedabove.
For 1980, however, NML's c.
l
.?f^
4 assets (again ignoring retrospuctive premium adjustments) would have enabled
. ~,,.
'I NML to pay no mors than perhaps one-fifth or so of the obligation, net of reinsurance, that would, have been incurred for this claim in 1980, given the
)
,s n
. /
e assumptions above. To have remained in business and honored the remainder cf I
t.
this claim, NML doubtlcss would have had to levy an assessment on'i',s memberts,
\\'
- Operating loss would have been the $21.7 million gain reported in Exhibit 5 less the $101 million increase in incurred losses and loss adjustment expenses for 1979 as the result of the new claim, or a negative figure of $79 3 million.
This figure divided by $24.1 million of earned premiums equals about 3.29 or, as a percentage, a negative 329 In a real sense, however, NML's option to call for a retrospective pre-mium adjustment is an asset, albeit one that is not capitalized.
r
[
31 Pending collection of the assessment, sit could have availed itself of its line of credit that stood at about $50 million as of the beginning of 1979 Had the Three Mile Island claim fallen on IEL, it likely would have produced in due course a substantial increase in the reinsurance rates payable by NML.
Reinsurance--at least as to " working cover" losses--normally attempt over the long run to make a profit on'each account.
The $75 million probably would have been viewed as a " working cover" loss, that is, a loss of a size that i
reinsurers would expect to pay during a span of ten years or so.
The above conjecture is useful in showing the extreme sensitivity of NML's underwriting ratio to even one significant claim. The exercise is nothing more than conjecture, however, because the Three Mile Island loss fell upon ANI-MAERP and not upon NML. NML's underwriting and operating ratios, in fact, were excel-
~$entthrough1981asmeasuredbytheusualinsuranceindustrystandards.
C.
Nuclear Electric Insurance Limited The third principal source of nuclear property is Ibclear Electric Insurance Linited (NEIL). Like NML, NEIL is a Bermuda-domiciled mutual insurance company.
- Any resulting loan would have acy ued interest "on an annual basis at a rate equal to two percent over the London Euro-dollar deposit rate."
Summary of Operations (January,1979), Annex III, p. 5 The loan also would have been collateralized by all of the assessments that NML could have levied as the result of the claim. Maintenance of the line of credit requires payment by NML of an annual commitment fee of one-half of 1 percent per annum on the daily unused balance. The fee ran about $253.000 in 1977 This letter of credit was can-celled as of June 30, 1981.
..This practice stands in contrast to insurance in that an insurer may be content to mak2 a profit on its book of business without expecting that each account would prove to be profitable.
.*. The information about NEIL presented in this report was drawn largely from Description of_ Excess Property Insurance Program, Nuclear Electric Insurance Limited (Au6ust 28, 1981). This document is cited elsewhere in this report as Description...
S.
\\
32 l
Origin and Purpose.
NEIL utilized a Bemuda charter issued to the Energy Mutual Liability, Limited by an act of the Bemuda legislature on February 3,1978.
l The charter became effective on July 4,1978.- On January 9,1980 the name was i
officially changed to Nuclear Electric Insurance Limited.
NEIL issued its first l
l insurance policy as of September 15, 1980.
i 1
NEIL is a companion company to NML, having been established by representa-tives of some of the same nuclear utilities that are members of NML. The reasons for establishing NEIL as an offshore industry captive perhaps bore close resem-l blance to the reasons for establishing NML. Some overlap exists between the i
composition of the NML board of directors and the composition of the NEIL board.
As of January 1,1982, nine persons were serving on both besards.
At this time j
NEIL and NML used the same legal counsel but used different local legal firms in Bermuda and different chartered public accountants.
i NEIL operates two types of nuclear property insurance programs and main-tains a segregated set of accounts for each program. The members who participate in only one program have no claim on the assets identified with the other pro-gram. One program, which became effective on September 15, 1980, provides extra j
expense insurance that indemnifies an insured nuclear utility for part of the cost of purchasing substitute electric power. The insurance applies where the i
purchase is necessitated by interruption of power production at the insured J
nuclear utility's plant site because of an accident causing property dama6e and arisingoutofradioactivecontaminationand/oroneormoreotherinsuredperils.
The other program, the first policy of which was issued November 15, 1981,
)
provides nuclear property insurance that is excess over the insurance that is i
I h
This company had been formed by the Edison Electric Institute for an unre-l lated purpose but had not been utilized.
i
33 provided by ANI-MAERP or NML. " Excess" as an insurance term means that for any insured loss this coverage applies only if and after the underlying insurance limit has been exhausted.
Excess insurance usually is at least as broad as the underlying coverage. NEIL's excess program is typical in this regard.
NEIL's excess insurance program was established as the result of recommen-dations of the Task Force on Nuclear Property Insurance." This task force was formed early in 1981 by the Edison Electric Institute.
Doubtless, the impetus for the formation arose in part from the work done by the Institute's Tusk Force on Nuclear Institutional Issues.
The mission of the task force was to explore the promising alternatives for increasing the amount of nuclear property insur-ance available to nuclear utilities and to recommend a course of action. As observed elsewhere in this report, the cleanup costs at the Three Mile Island plant site had demonstrated the gross inadequacy of available insurance.
Where appropriate in the comments that follow, NEIL's extra expense insur-ance program is abbreviated as NEIL-I and NEIL's excess insurance program as NEIL-II.
Method of Operation.
NEIL conducts its business in a manner similar to that of NML.
Like NML and for the same reasons, NEIL transacts all of its business outside the United States.
It uses committees composed of experts from its member utilities.
One such committee is the " Insurance Advisory Committee" to provide advice on ratemaking, coverages, and other insurance-related subjects.
The other is the "Enginearing Advisory Committee" to provide loss control expertise.
As of this writing a small gap or corridor could exist between the upper limit of the underlying coverage and the activation point of this excess insur-ance.
Details are offered in Section IV, D of this report.
- See Description..., p. 2
- See Cleanup of Three Mile Island No. 2, Report of the Task Force on Nuclear Institutional Issues to Edison Electric Institute'c Board of Directors (March, 1981).
1 On December 18, 1981 NEIL established a wholly-owned subsidiary incorporated in the state of New York. This subsidiary is a service organization for NEIL in 1
the United States and serves as a liaison between NEIL and its members' committees.
NEIL's Bermuda manager, Marsh & McLennan (Bermuda) Limited, providea office facil-ities. An employee of Marsh & McLennan (Bermuda) Limited serves as general man-ager of NEIL.
As of February, 1982, NEIL-I was in the market for reinsurance that would provide an aggregate limit during a policy year of $89 7 million excess over
$179.4 million. The reinsurance would become available for a loss or losses incurred by NEIL-I in a given policy year in which NEIL-I had already incurred
$179.4 million in one or more losses. By February 17, 1982, NEIL-I had succeeded in purchasing only about $19 million of this reinsurance.
NEIL-II is also dependent in part on reinsurance, which at this writing was of the quota share type and was in place in the amount of $61 million excess of $500 million."
Insuring Activities and Operating Results. As of March 15, 1982 NEIL-I provided coverage on fifty-five nuclear units at thirty-six sites. Thirty-six utilities held insurable interests in one or more of these sites.
Of these fifty-five nuclear units twenty-eight were also insured in NML.""
Fourteen of
- nu the nuclear utilities purchasing NEIL-I insurance also patronized NML.
The other twenty-one patronized ANI-MAERP.
- Description..., p. A-9 plus February 17, 1982 letter to author from NEIL.
n See Section IV, D for further comment.
- " Letter of March 26, 1982 to author from NEIL.
Ibid.
u*n i
l
35 Having begun its NEIL-II insuring activities only on November 15, 1981, NEIL was in the process of establishing its excess insurance program at the time of the drafting of this report. As of March 15, 1982 NRIL-II provided coverage on forty-five nuclear units at thirty-two sites, with' thirty-one util-ities having insurable interests in one or more of the units.
Of the thirty-one utilities, nineteen had purchased the underlying coverage in ANI-MAERP; eleven were insured by NML and one utility had one unit insured by NML and another unit insured by ANI-MAERP.
With NEIL being such a young organization, the brief history of its finan-cial operations is not particularly revealing for NBIL-I and even less so for R IL-II.
Little can be said about the operating results of NEIL-II because its first policies were issued only on November 15, 1981.
NEIL-I does have a short period of operating experience.
Exhibit 6 summarizes the underwriting and investment income results for NEIL-I from July 4,1978 (the effective date of NEIL's formation) through December 31, 1981. The summary indicates that NEIL-I had suffered only one claim and that NEIL had paid or " reserved" (that is, had established a current liability for claim payable on its balance sheet of) $33,200,000. The amount was only an estimate that might or might not prove to have been accurate. Even with this substantial claim, however, NEIL-I showed a cumulative underwriting ratio through Ibcember 31, 1981 of only about 39 percent. When net investment income up to December 31, 1981 was added to its underwriting gain to the same date, its cunulative amount of operating gain relative to cumulative earned pre-niums was about 80 percent.
Ibid.
Ibid.
36 EXHIBIT 6 UNDERWRITING AND INVESTMEfff INCOME FOR EXTRA EXPENSE INSURANCE PROGRAM NEIL July 4,1978-December 31, 1981
($ in millions) i Earned Premiums
$ 89.6 i
Less:
a Insured Losses and loss Adjustment Expenses
$ 33.2 Underwriting Expenses 1.8 35.0 i
Underwriting Gain 54.6.
Plus Net Investment Income 17.1
]
Operating Gain 71.7 a.
This figure includes amount paid plus amount reserved as of December 31, 1981 for a loss that occurred on April 21, 1981 b.
Includes all expenses Source: Letter of March 26, 1982 to author from NEIL j
k l
l l
37 As is the case with regard to the other principal sources of nuclear property insurance, caution should be exercised in imputing significance to these profitability figures. The exposure base is too small and the operational his-tory much too short to lend statistical significance to the data with regard to predicting future losses. One large insured loss could change the results drastically for either or both of NEIL's nuclear insurance program.
III.
DISTINGUISHING FEATURES OF EACH TYPE OF INSURANCE In this section of this report a brief description is offered of the distinguishing features of the nuclear property insurance provided by each principal source. Attention is given first to some of the features of the insurance issued by ANI-MAERP. Then note is taken of the major respects in which NML insurance differs from insurance of ANI-MAERP.
Next, the main features that distinguish NEIL insurance from NML insurance are examined. Finally, the prin-cipal difference between NEIL-I and NEIL-II insurance are recognized.
A.
ANI-MAERP Insurance ANI-MAERP issues five different types of policies. One covers damage to a nuclear unit under construction.
Another applies to property damage to an oper-ating facility. Still another applies to loss of income of a nuclear fuel fabri-cation or processing facility (but not to a spent fuel reprocessing plant or a re-actor) because of interruption of operations as the result of property damage from an insured peril. Two other types of policies pertain to the coverage of nuclear materials in transit.
Further attention to ANJ-MAERP insurance in this report is confined to the property damage type of policy applicable to an operat-ing facility of a nuclear utility, w In insurance teminology a property damage policy applicable to construc-tion in progress is called a " builders risk" policy.
38 All-Risk Approach. To someone familiar with property-liability insurance policies the most conspicuous feature of ANI-MAERP insurance is the close resem-blance it bears to "all risk" insurance sold by commercial insurers to cover large industrial properties. The major distinction of ANI-MAERP insurance is that it does not exclude--rather, it specifically includes--radioactive contam-ination as an insured peril. Other policies are almost sure to exclude radio-active contamination from the insured perils.
In designing the policy ANI-MAERP followed the usual technique employed with respect to "all-risk" policies.
It specified that the insurance would apply to all risks of direct physical loss (including radioactive contamination) except as specifically excluded in the policy.
One then has to study the exclusions in juxtaposition to the insuring clause to ascertain what coverage remains. The exclusions are mainly of three types. One type has to do with perils that are generally regarded as uninsurable.
Examples are war, wear and tear, fraudulent acts by an insured, and gradual accumulation of radioactive contamination. Another type has to do with perils that, if covered at all, are subject to additional premium.
Examples are earth movement, such as earthquake and volcanic eruption.
In some policies where these exclusions can be deleted by the payment of extra premium the insurer is unwilling to issue as much cover-l age for these perils as for others and specifies a lower maximum limit if lia-bility is to apply to them.
For the proper charge ANI-MAERP will delete the earth movement exclusion without imposing any sublimit on the maximum loss pay-able from that cause. The third type relates to property that normally is not A peril is a cause of loss, such as fire, wind, explosion, or radioactive contamination.
This technique is in contrast to the " named perils" approach in which coverage is provided only for the perils specifically listed by the insurer as being applicable.
P l
l
39 insured at all or is the subject of specialized insurance. Examples are money, records, trees, vehicles licensed for highway use, and land.
Measurement of Loss.
Insurance policies applicable to property nonnally specify one of two bases for measuring the extent of any insured loss. Some policies call for payment of the loss on the basis of " actual cash value;"
others provide for payment on the. basis of " replacement cost new."
ANI-MAERP insurance can be written on either basis at the option of the insured. The policy itself is couched in terms of actual cash values an endorsement can be used to convert the coverage to replacement cost new.
The actual cash value type of insuring clause specifies that in case of destruction of the property the insurer will indemnify the insured to the extent of "the actual cash value of the property at the time of loss, but not exceeding the amount that it would cost to repair or replace the property with material of like kind and quality within a reasonable time after such loss...."
In case of partial loss the measure is the difference between the actual cash value of the property immediately before the loss and its actual cash value immediately after the loss, subject to the repair or replacement option asabove. Actual cash value is usually regarded as replacement cost new less the decline in value attributable to use, passage of time, obsolescence, and other relevant variables.
The replacement cost new value of an object of property is the amount required to replace that object with a new object of the same type and quality.
In measur-ing the loss no consideration is given to the fact that the object damaged or
- Coverage is available by endorsement for vehicles licensed for highway use but damaged on the site and for extension of the decontamination and debris removal coverage to the land that is a part of the insured site.
=_
i 40 4
I l
destroyed was not necessarily new and, in fact, might have been old, worn, or i
]
otherwise less valuable than a new object of thir come type and quality.
l Coinsurance. The ANI-MAERP policy, like most other insurance ~on indus-3
]
trial properties, contains a coinsurance clause. The reason for the inclusion of this clause can be understood by recognition of the fact that most insured losses to property--even to nuclear utility units--are likely to be small rela-
]
tive to the value of the property exposed. An insured is free to buy much or i
little insurance. Unless the rate for the insurance (that is, the price for a 4
l standard amount, say, $100 of insurance) is to vary proportionately and inversely with the relative amount of insurance bought, inequities can arise between an l
insured who buys a large amount of coverage relative to the value of the property covered and an insured who buys a small amount. The latter would pay a lower i
premium but would have just as much coverage as the fomer had for any loss that did not exceed the amount of insurance purchased. With most losses being small, the insureds with a high ratio of insurance to value wotld bear an unduly large i
share of the prenium burden.
4
[
The purpose of the coinsurance clause is to introduce equity into the rela-tionship between these two insureds. The insurer offers the insurance to both at the same relatively low rate. By.means of the coinsurance clause the insurer then As of the date of this report, ANI-MAERP insureds were somewhat evenly l
divided as between actual cash value and replacement cost new types of insuring clauses. Whichever type is used, the policy remains subject to the maximum amount
{
of insurance applicable to a given type of property, as indicated by the limit stated in the policy. To the extent the insurance is applicable to decontamina-tion or debris removal costs (as discussed below), the distinction between measur-ing loss by an actual cash value standard and measuring it by replacement cost new l
standard loses its significance.
i For example, an insured with only 10 percent of insurance to value would pay only about one-tenth of the premium paid by a similarly-exposed insured with 100 percent of insurance to value exposed.
Yet, for any loss not in excess of 10 per-cent of the value exposed, the former insured would have just as much protection i
as the latter. Loss history demonstrates that for losses from traditional perils of fire, wind, explosion, and the like the size of the loss for the heavy prepond-y erance of losses is under 10 percent of the value exposed. Whether this skewness will characterize losses befalling nuclear properties is a matter Wt to be deter-i mined. Early experience is affirmative.
41 stipulates, in effect, that, if any insured fails to buy and maintain at least the agreed amount of insurance, the insurer will hold back to the same degree in paying for any loss, even a small one, that the insured might suffer. The required amount of insurc,ce, as mentioned above, is usually stated as a percentage of the value orthepropertyinsuredI A 90 percent coinsurance figure is stated in the ANI-MAERP policy. This per-centage can be changed by endorsement to any percentage that is a positive integer from 1 through 89 Other things the same, the lower the percentage the higher the rate. A table is available from ANI-MAERP to show the relative rate for the in-surance when the selected coinsurance percentage is a figure other than 90. For example, when 80 percent is selected, the rate is about 1.0918 of the 90 percent coinsurance rate.
At 10 percent the factor is 51462 A problem obviously arises if 90 percent of the value of the property exceeds the amount of insurance available in the market. No insured should be penalized in a loss settlement for failing to purchase insurance that is not available.
A difficulty of this type is normally circumvented by the parties agreeing (by use of an agreed amount endorsement) that a specified dollar amount of insurance will satisfy the obligation imposed upon the insured by the coinsurance clause.
ANI-MAERP uses this endorsement to indicate tnat purchase and maintenance of $450 million (or other agreed amount) satisfies the requirements imposed by the coinsurance clause.
- In a policy written on an actual cash value basis the percentage applies to the actual cash value.
If the policy is on a replacement cost new basis, the percentage applies to the replacement cost new.
An example may be useful to illus-trate the operation of a coinsurance clause. Suppose an insured buys only $300 million of insurance in an actual cash value type policy on property with an actual cash value of $400 million and that the policy contains a 90 percent coinsurance clause.
If the insured suffered a $600,000 actual cash value loss and if no deduct-ible applied, the insurer would pay no more than $500,000. The insured was obliged to have $360 million (90 percent of $400 million) but had only 3300 million, or five-sixths of the required amount.
Hence, the in surer would pay only five-sixths of the loss.
- In fac t, ANI-MAERP uses the agreed amount endorsement to deal with the problem of underinsurance at the time the policy is issued. For example, if an insured would buy, say, only 45 percent of insurance to value, ANI-MAEEP would use the agreed amount endorsement to indicate that such amount would satisfy the coinsurance clause.
ANI-MAERP, however, would also charge about 1.6767 of the 90 percent coinsurance rate for the insurance. See Section VII, B for further discussion of this point.
42 Other Policy Features. Other significant features of ANI-MAERP policies include the following:
1.
As observed in Section II of this report, the policy responds, within the limit of coverage, not only to direct physical damage to or destruc-tion of property but also to the cost of decontamination and debris removal.
4The policy contains a section entitled " Debris Pemoval and Decontamination."
The wording in this section indicates that, subject to other stipulations, the insurance applies to expenses "in removing debris of and in decontaminating the property covered by this policy following direct physical damage to such property caused by any peril not excluded hereunder." This section clearly adds coverage with regard to debris removal. Without this section, the cost of removing debris would not be covered.
One could argue, however, that the policy would cover at least some expenses of decontarinating radioactive property, even if this section were not in the policy. Tnis argument rests on the fact that radioactive contamination is a peril that is explicitly included in the main insuring clause of the policy. For this reason the policy presumably responds, subject to the amount of insurance pur-chased, to loss caused by the radioactive contamination. The insuring clause specifies that the insurer will " indemnify" the insured for loss from an insured peril but not in excess of "the amount which it would cost to repair or replace the property with matarial of like kind and quality within a reasonable time after such loss...."
To the extent that decontamination is the most economical way "to repair" the property, decontamination expenses would seem to be encompassed within the insured's promise to indemnify even in the absence of the debris removal and decontamination section.
The added section, however, renders the argument moot by specifically includ-ing decontamination expenses. The section contains no sublimit as to the amount of insurance applicable to expenses for decontamination or debris removal.
Even if extraordinarily large expenses were incurred to decontaminate property or to dispose of contaminated debris, the coverage could apply up to the amount of insurance bought that had not already been used to pay for the damage to the property or to pay for prior losses during the policy year.
Incidentally, if an insured has purchased "all-risk" insurance equal to at least 90 percent of the insurable value but has chosen not to purchase "all-risk" insurance up to the limit ANI-MAERP offers for sale, ANI-MAERP is willing to offer " decontamination only" insurance in the amount of the difference between the maximum amount of "all-risk" coverage ANI-MAERP offers and any smaller amount of "all-risk" coverage an insured wants.
See Section IV A, for the maximum amount of "all-risk" coverage offered by ANI-MAERP.
s
43 2
The intent inherent in ANI-MAERP insurance is to avoid using the insur--
ance for nonserious claims. The policy contains a deductible feature, the purpose of which is to preclude payment by the insurer for small losses. The minimum deductible is $50,000 per loss with larger deductibler, and commensurate premium credits available for use at the option of the insured.
A larger mandatory deductible is applicable to various turbines and transformers, with the size of the deductible dependent on the capacity of the turbine or transformer. The highest such manda-tory deductible is $2.25 million.
In a carryover fron ocean marine insurance the policy provides that the deductible is waived in any case where the loss exceeds 50 percent of the amount of the insurance. The rationale is that no reason exists to penalize an insured who suffers a serious loss.
3 The ANI-MAERP policy is cancellable either by the insurer or insured but is atypical with respect to early cancellation. Conventionally, if the insured cancels the policy, the insurer will refund the unearned premium at "the customary short rate."
This policy does not call for short-rate refund of the premium upon cancellation of the policy by the insured.
Rather, it provides for a pro rata refund--just as it does for cancella-tion by the insurer.
It contains the additional stipulation that, no matter how early in the policy period the insured may cancel the policy, the insurer is entitled to a pro rata premium for at least six months of coverage.
The most commonly selected deductible in early 1982 was reported to be
'250,000.
The largest optional deductible was $10,000,000.
..A "short rate" is a less-than proportionate refund of premium.
For example, canec11ation of an annual policy after three months would allow a refund of not 75 percent but only 65 percent of the premium.
.** This provision is waived for an insured who cancels early in order to purchase a new policy at a higher limit, e.g., an insured who goes from 0375 million to
$450 million.
44 4
ANI-MAERP insurance is atypical in another important respect. Most property insurance policies contain a loss clause. This clause normally provides that a claim does not reduce the amount of insurance applicable to the next Joss that might occur. ANI-MAERP insurance by contrast stip-ulates that the limit in the policy is the maximum amount the insurer will pay to the insured in any one policy year, even if the insured suffers mul-tiple losses. Thus, after a loss has occurred, the amount of insurance available for a subsequent loss in the same policy year is the initial amount of insurance less the amount already payable for the loss that occurred.
Upon request by the insured, the insurer might but is not obliged by the policy to restore the amount in consideration of additional prenium.
Ratemaking.
In establishing rates for ANI-MAERP insurance the Nuclear Insur-ance Rating Bureau (NIRB) uses a p ocess that bears a close resemblance to that used in fire and allied insurance ratemaking. The rate consists of fcur elements as follows:
1.
The rate for insurance that provides protection against the conventional fire and allied perils.
This element of the rate is based on informa-tion provided by the Insurance Services Offices, the principal ratemaking organization for conventional property and liability lines of insurance.
This element gives cognizance to the location of the property, construc-tion (assuming the property is a structure), proximity to other proper-ties, use, protective devices, security, maintenance, housekeeping, and other variables.
Consequently, the resultant figure is not necessarily the same for any two nuclear utility reactors. The rates supplied by the More precisely, these perils include the ones in the standard fire policy, the extended coverage endorsement, and the vandalism and malicious mischief endorsement.
The system used to produce this element is called " scheduled rating."
It con-sists of adding charges and subtracting credits from a residual or basis rate to arrive at a rate that reflects the physical and certain other characteristics of the property that presumably have a bearing on the frequency and severity of loss.
45 Insurance Services Office, based on its " Electric Generating Station" schedule, are merely " advisory" for NIRB. That is, NIEB is free to accept the element as received or adjust it before adding to it the other elements mentioned below.
2 A loading for the radioactive contamination peril. This loading is judg-mental rather than being the product of a calculation based on empirical data. The judgment, however, is made with attention to the type of reactor (such as the pressurized water type), use of reactor (with use related to the anticipated schedule of operation), power level of reactor (thatis,thelevelofmaximumlicensedpoweroutput),typeofreactor containment (such as thickness of walls), design of the reactor (its configuration among other points), and age of the reactor.
3 To the rate accumulated by the addition of elements one and two is added still another element.
It is the conventional rate for direct loss from a peril covered in boiler and machinery insurance. This charge is needed because the boiler and machinery perils are not excluded from the policy.
This line of insurance pertairs to accidental breakdown of a vessel under pressure (such as a boiler) or of a machine with moving parts (such as a turbine), with neither type of breakdown involving radioactive contamination.
This element is comparable to the charge that would be made for boiler and machinery insurance were the reactor powered by conventional fuel.
4 Finally, because the ANI-MAERP policy is "all risk," another element is added to provide a charge for all the other perils that have not been excluded. An example of such a peril is damage done by the collapse of a crane during the time materials are being moved.
- The charge for the radioactive peril is in element two.
46 ANI-MAERP in 1972 introduced its " Nuclear Property Insurance Industry Experi-ence Guide" to allow loss experience of the insured nuclear utilities as a group to influence the rates. ANI-MAERP's loss experience for some preceding period (now twenty years) is used to compute a " rating aodification factor" to be applied against each insured's premium for the following year. To the extent the loss experience in the relevant period was good, the rate for the following year is reduced. To the extent that experience was bad, the rate is increased. Whereas the unadjusted rate payable by each nuclear utility is individualized and may be different from the rate payable by any other, the same modification factor for a given year is applied to all insureds.
Exhibit 7 shows the rating modification factors that have been used. For example, the -7 7% factor for 1972 was applied against the next premium otherwise due for each policy being renewed after the effective date of the newly-announced modification factor. The sharp change in the modification factor for 1980 com-pared to 1979 and the further deterioration for 1981 indicate the powerful effect on ANI-MAERP loss experience of the accident at Three Mile Island.
B.
NML Insurance NML issues a builders risk policy, an operating facility policy, a combined operating facility and builders risk policy, and a transit policy. Purther atten-tion in this report is confined to the operating facility policy, except for tb9 observation that the policies are quite similar to each other in underwriting approach.
The NML operating facility policy tracks fairly closely the ANI-MAERP poli-cies on the fundamental aspects of coverage. The NHL policy embodies the "all-risk" approach and includes radioactive contamination.
It uses actual cash value as l
l the basis for measurement of loss, but an optional endorsement permits the insured to use replacement cost new as the neasurement of loss. The coverages extend to l
i l
47 EXHIBIT 7 RATING MODIFICATION FACIT)RS ~
NUCLEAR PROPERTY INSURANCE INDUSTRY EXPERIENCE GUIDE ANI-MAERP 1972-1982 Approximate Percentage by Which Next Year's Premium to Be Altereda 1972
- 7 7%
1973
-12.2 1974
-17 5 1975
-24.1 1976
-29.5 1977
-34.0 1978
-36.0 '
1979-
-29 5 1980
- 2.4 1981
+38-1982
- 0.8 The application of these factors throt gh 1980 led to a reduction in premiums a
of about $75 million, which amount is about 20 percent of the approximately
$378 million in premiums otherwise payable Source: " Nuclear Insurance: Facts and Figures," American Nuclear Insurers Reports,-#1, (Revised March,1981), pp. 5-6 and (Revised March,1982),
pp. 8-9 l
48 the costs of decontamination and debris removal. The amount of insurance is the annual aggregate maximum payable, whatever may be the number of losses at.the insured site during the policy year.
NML's rates are a function of a set of variables that is similar to the set used to produce ANI-MAERP rates.
Because of the possibly heavy use of retrospective premium adjustments, as discussed below, NML advance premiums, other things the same, may be somewhat lower than ANI-MAERP premiums that are not subject to retrospective increase.
Minor Differences as Compared to ANI-MAERP Insurance. Several minor differ-ences exist between the ANI-MAERP and NML policies, some of which are as follows:
1.
The exclusions, while similar, are not identical. For example, whereas the ANI-MAERP policy excludes loss covered by a manufacturer's warranty, the NHL policy does not, except by endorsement. The author understands, however, that the NML endorsement is mandatory so as to render the NML coverage equivalent to the ANI-MAERP coverage.
2 The NML policy offers the insured a choice from 90 percent down to 10 percent of coinsurance clause percentages.
In ANI-MAERP policies, as noted above, the choice goes to 1 percent and is offered by endorsement.
3.
The minimum "overall" deductible in the NML policy is larger than in the i
ANI-MAERP policy--$100,000 as compared to 350,000. The minimum required deductible applicable to any " turbine generator unit," while related to the size of the unit, does not exceed $750,000, irrespective of the
[
capacity of the unit.
As with ANI-MAERP, larger optional deductibles l
are availabic.
4 With regard to cancellation by the insured early in the policy year, NML does not impose the stipulation--as does ANI-MAERP--that at the minimum l
Whereas ANI-MAERP rates are set by the Nuclear Insurance Rating Bureau, NML established its own rates upon the advice of its consultant.
49 a pro rata premium for six months is deemed to have been eamed. Rather, NML uses the customary short rates to compute the premium to be refunded.
With regard to cancellation by the insured in the second half of the policy year NML provides only the short-rate refund of premium, whereas ANI-MAERP provides the larger pro rata refund.
5 Whereas the ANI-MAERP policy is silent as to the jurisdiction the law of which shall govern the settlement of any dispute, the NHL policy specifien that the law of the State of New York shall apply to the constniction and interpretation of the policy provisions.
6.
While the NML policy is silent with respect to policy dividends payable to insureds, authorization for such distributions to members is found in Sections 46 (c) and 49 of the "by-laws" of NML. The practical effect of these provisions is to render the policy " participating." The ANI-MAERP policies are not. A participating policy is one that provides for possible but not mandatory distribution to policyholders--usually annually--of redundant premiums, investment earnings not needed for other purposes, or assets from other sources. Policion of mutual com-panies are normally participating.
Major Differences as Compared M ANI-MAERP Insurance.
Individually or even in the aggregate the differences enumerated above may not be significant. Two other differences, however, between NML's insurance and ANI-MAERP's insurance are, in this author's judgment, of profound significance.
Each is described briefly.
Assessability. One of these two differences is that the NML policy is assess-able.*
The policy provides for what are called " retrospective premium adjustments" and stipulates that:
'Under the terms of an assessable policy the insurer at its option can levy against the insured additional charges up to some specified maximum. The charges so levied become a debt legally owed by the insured to the insurer.
I t
50 The Insurer may make demand for the Retrospective Premium Adjustment in whole or in part or parts from time to time, but only to the extent necessary, in the sole discretion of the Board of Directors of the Insurer, to cover losses and the costs thereof incurred by the Insurer with respect to the policy period stated in the Declaration of this Policy under any and all insurance agreements, including this Policy, entered into by the Insurer. The costs of losses shall include with-out limitation all costs attributable to paying, financing, litigating and settling such losses.
This wording indicates that the insured could become liable to the insurer for an amount above and beyond the " regular" premium. Several characteristics of this contingent liability, as specified in the policy, are summarized as follows:
1.
If the insurer's board of directors should decide to call for one or more retrospective premium adjustments in response to one or more losses occurring during a given policy year, any insured of a policy issued to apply to all or any part of that policy year shall be liable fortheretrospectivepremiumadjustment/s.
The liability, up to the limit specified below, holds even though such insured may have l
cancelled the policy after the occurrence of any such loss or losses.
2 Each insured is obliged to pay any retrospective premium adjustment in full within twenty days after it has been demanded by NML.
3 Failure on the part of an insured to pay a retrospective premium adjust-ment within twenty days of the call allows the insurer to demand from i
Under " Cancellation of Policy," the specimen NML operating facility policy available to the author contains this provision: "Neither the cancellation of the Policy on the part of Member Insured (s) nor the Insurer shall affect the obliga-tion of the Member Insured (s) to pay the Retrospective Premium Adjustment callable under this Policy for losses and the costs thereof incurred by the Insurer with respect to the policy period specified in the Declarations." (Dmphasis added by the author.) A literal interpretation could be that an insured who cancelled i
could be liable for a retrospective premium adjustment for a loss that occurred during the policy period specified in the declarations but after cancellation.
No loss occuring before a policy is issued creates any contingent assessment lia-bility because for any given insured the " policy period" begins at the inception date of the policy. Thus, an insured buying an NML policy as of, say, April 1, 1982, would not be liable for any call made with respect to any loss occurring before that date.
l l
51 all insure ls in the aggregate an additional retrospective premium adjustment in the amount of the earlier demand that was defaulted, cubject to the maximum mentioned below.
4 For any policy year the maximum amount of retrospective premium adjust-ment for which an insured can become liable is the product of that innured's annual premium times the multiple specified in Item 4, C of the Declarations that are a part of the policy.
According to Section V, B of NML's Summary of Operations the multiple in Item 4, C cannot cxceed fourteen.
5 A retrospective premium adjustment demanded by NML in response to one or more lonnes occurring in a given policy year can be called only from an incured whose policy was in force for all or any part of that policy year.
6.
The insurer may assign to one or more banks that make credit available to the insurer the insurer's interest in one or more retrospective premium adjustments, which assignment allows the bank /s to demand pay-ment of the retrospective premium adjustment /s in the event the insurer defaults on its obligation /s to such bank /s.
(As noted above, NML has no letter of credit as of thic writing.)
- (August,1981),p.4
- A recent insurance trade prenn piece indicated that the multiple in current NML policies was six in early M1rch,1982 and that it might drop to five and six-tenths. John W. Milligan, "TVA Movec Nuclear Coverage to Mutual," Buninens Insurance (March 8,1982), pp. 3-4.
Peadern should notice, however, that assess-nents that are still callable under older policies are for the multiples speci-fled for those policico.
Purther, NML apparently can reinctitute any higher multiple up to fourteen in renewal policien, chould it see fit to do so.
Readers should also be reminded that the caption in the article cited in this footnote is misleading. TVA did not move its nuclear property insurance. Rather, not having any, it bought cuch incurance from NML.
52 7
No insured shall be liable for any retrospective premium adjustment that is not demanded by the insurer (or a bank to which assignment has been made) within six years after the expiration of the policy year with respect to which the call is made.-
This enumeration indicates that for each policy year an NML insured can become liable for one or more assessment /s.
Any such assessment may be levied during that policy year or during the following six-year period. Any such assess-ment must be in response to an insured loss or losses occurring during th policy year.
The "long tail" on this contingent liability means that the liability could accumulate. At the worst, an insured would be contingently liable for assessment /s for each of seven years, with the contingent liability for each year being at the maximum of fourteen times the premium for that year.
For a given insurer the total possible accumulation of contingent assessments, assuming that the multiple of fourteen was used in each policy and assuring that the pre-mium had not changed over the seven year period, would be ninety-eight (that is, fourteen times seven) times the annual premium.
Limitation o{ Liability. The second of the two major differences between NML and ANI-MAERP insurance is attributable to another provision in the NML policy as follows: "The ability of the insurer to pay the amount of insurance to the Insured (s) in case of loss is limited to the assets of the Insurer." The legal significance of this statement is beyond the expertise of this author, a lay persen to judge. Perhaps the statement, being self-evident, is not limiting in any way with respect to the insurer's liability.
- This situation could arise if each year's policy specified a multiple of fourteen, if serious losses occurred each year for seven years, if no retrospective premiums had been called before the seventh year, and if the insured had been issued a policy in each of those seven years.
Admittedly, this situation is far-fetched.
Yet, it is possible.
It is posited here merely to call attention to the possible accumulation, even though the probability of such a development may be extremely low.
If such a series of losses occurred doubtless NML (or any other insurer on which the losses fell) would be drastically restructured before the end of the seven-year period.
l
53 Aside from possible legal significance, however, its appearance as a provi-sion in the policy could be a revealing indication of insurer philosophy. The statement could be construed as an assertion that if the insurer does not have the ability to pay a claim the insurer is not to be deemed as owing what it can-not pay. According to NML's legal counsel, such construction was not intended in the design of the policy. Rather, the precision was deemed by counsel to be needed in the interest of full disclosure to emphasize to prospective insureds the fact that NML'n resources for paying claims were limited.
In any case, the pro-vision is unusual. Not having encountered such phraseology previously, the author regards it as a significant and distinguishing feature of NML insurance.
NEIL Insurance As explained above, NEIL has two separately-operated insurance programs. The insurance in each program is structured in a fashion similar to that of NML.
- Yet, each NEIL program differs in important respects from its NML counterpart and from the other NEIL program. With NML's insurance having already been contrasted with that of ANI-MAERP, the technique used in this part of the report is first to comment on the characteristics that distinguish both of the NEIL programs from NML and, next, to explain how the two NEIL programs differ from each other and/or how each differ's further from NML.
Features that Distinguish NEIL Insurance from NML Insurance. Both NEIL pro-grams can be distinguished from NML insurance in the following ways:
- As explained to the author by NML's counsel, an invitation to prospective insureds to participate in NML could be considered as an " offering" under the Securities Act of 1933 and the Securities and Exchange Act of 1934 Through its counsel NML obtained a "no action letter" from the Securities and Exchange Com-mission that obviated the need for NML to file a registration statement with the SEC. For prudence's sake, however, counsel felt that explicit attention should be drawn in the policy to the fact that NML's ability to pay claims was restricted to its assets, its reinsurance, and its ability to call for retrospective premium ad justnents. These same comments also apply to NEIL.
As mentioned below, NEIL policies seem to go even further in specifying possible limitations of liability.
54 1.
Each insured is required at the time of initial purchase of a NEIL-I policy to pay in addition to the' initial premium an amount equal to 13 percent of the annual premium. The same requirement applies to a NEIL-II initial purchase. This amount is called the " reserve premium" and is sub-ject to being refunded at the time the member withdraws from such pro-gram, provided the amount has not been and is not deemed to be needed by NEIL for other purposes. The reserve premium, somewhat like a condi-tionally refundable membership fee, is intended to satisfy a statutory requirement.
3 2
NEIL policies stipulate that if two or more insureds suffer losses that in the aggregate exceed NEIL's ability to pay, NEIL discharges its lia-i
~
bility by paying each claimant a pro rata share of NEIL's assets. The l
policy provides information on precisely how the prorating is to be undertaken.
3 The members of NEIL expressly acknowledge that a claim is payable only out of the assets of the program under which the claim arose. That is, NEIL-II assets are not available to pay NEIL-I claims and vice versa.
- Under the Bermuda Insurance Act of 1978, which became effective January 1, 1980, NEIL is required to maintain a surplus equal at least to an amount related to NEIL's premium income. The reserve premium was designed to meet this require-4 ment.
l As alluded to above, a disclaimer by the insurer for any continuing respon-sibility for a claim that it cannot pay strikes the author as odd.
NML may handle multiple losses in the same way. The specimen policy at the author's disposal, however, was not explicit on this point.
)
I
- ~..
55 4
While a NEIL policy is cancellable by NEIL, it is not cancellable by the insured. The stated reason for this prohibition was to maximize financial stability.
Also, NEIL may have thought this feature might help it avoid adverse selection.
Further Distinguishing Features of, NEIL-I Insurance. NEIL-I can be further distinguished by the following observations:
1.
As observed in Section II, C NEIL-I pertains to indirect loss rather than to " direct physical loss" as covered in the UML policy. NEIL-I is an example of " consequential loss" insurance, that is, coverage of a loss that grows out of a previous insured loss. Specifically, NEIL-I provides payment for a part of the extra cost of purchasing re-pincement power during a prolonged period of outage at a nuclear utili-ty's insured site. The NEIL-I policy pays only if the outage is attribut-able to interruption of the operation of a nuclear unit because of direct damage from an insured peril. The NEIL-I perils are similar--but not necessarily identical--to the perils in the underlying NML or ANI-MAERp insurance.
2 The policy provides for weekly payments by the insurer to the insured that begin after a twenty-six week waiting period following the onset of the outage. The waiting period is tantamount to a deductible. The
- Upon 60 days notice and a pro rata refund of premiums.
- NEIL initially considered the use of a three-year, noncancellable policy but decided that a one-year conmitment "would be more broadly accepted." Letter of March 26, 1982 to author from UEIL.
- Adverse selection in an insurance context is a manifestation of the tendency of a decision maker to obtain insurance when the expectation of insurable loss increases and to discontinue the insurance when the expectation lessens.
- Also called " time element" insurance.
- 1. !
56 payments can be made, during the continuing outage and after the 'e_nd of the waiting period, for the maximum of 104 weeks to apply to the reasonable cost of necessary expenditures for replacement power for-41 the disabled utility. " Reasonableness" is examined by NEIL-I under-writers at the time application is made for the covenge. Once the coverage has been issued, the specified weekly indemnity becomes auto-matica11y payable after the waiting period in case an outage results from an insured peril. The insured can purchase a weekly indemnity up to 90 percent of the " Reported Weekly Values" (that is, the estimated cost of obtaining replacement power), or $2.3 million, whichever is less.
For the second fifty-two week period the maximum-is only 50 percent of what it is for the first such period.
3 The maximum multiple of the annual premium subject to a IEIL-I re,tro-spective premium adjustment is five. With the liability for a call remaining contingent for six years beyond the expiration of the policy, the maximum contingent liability--given the assumptions stated in the NML " worst case" example--would be thirty-five T,imes the annual premium.
4 To date no distributions (dividends) have been paid to NEIL members.{
5 If an outage of more than one unit results from the same incident, the weekly indemnity per unit payable by the insurer is reduced. 'For example, if two units are simultaneously out of service, the weekly indemnity per
- The same assumptions and caveats hold here as with the NML " worst case"'
example. The point is worth repeating that, before IEIL-I suffered serious losses for six consecutive years (without making any retrospective premium adjustments),
an insured utility would be likely to have dropped the coverage.- 'In such,a case the utility would not be liable for a call for losses that occurred during any policy year after the policy expired.
If an unduly large number of insureds failed to renew their NEIL-f policies, however, NEIL-I could experience diffi-culty in paying its claims.
l l
t
1' 57 s
unit is only 80 percent of what would be paid if only one unit were disabled. For three units the comparable figure is 60 percent; for four units it is 50 percent. The use of this sliding scale for,the recovery per unit is intended to reflect the interdependency that may exist between or anong the units in the multiple-unit site. This adjustment is in addition to the increase in the rate as discussed below when multiple units are insured.
6.
NEIL-I uses an annual base premium of $1,514,000 for each reactor that is insured at the level of $2 million of weekly indemnity.
This figure was arrived at after empirical studies were made of the lose experience l
of all United States reactors suffering outages of at least twenty-six weeks during the years &om 1973 ,m;h 1979 The aim of NEIL-I was to develop a premium base adequate
.6IL-I to respond to two full-limit losses. The base premium can te ahjusted to take into account inter-dependency of multiple units sharing common facilities (such as a control room). The premium for two interdependent units is larger than the total premium that would be charged for the two units, other things the same, were they not interdependent. The premium is adjusted also to recognize compliance by the insured with loss control standards. NEIL-I is in the process of building further refinements into its ratemaking.
Further Distinguishing Features of NEIL-II Insurance. NEIL-II bears a closer resemblance to NML than to NEIL-I but is unlike either in the following respectsi 1.
NEIL-II, as observed in Section II of this report, provides excdss insurance rather than primary insurance." NEIL-II cioverage is actjvated Letter of January 29, 1982 to author from NEIL. No additional charge is made for the extra $300,000 now' includable in the maximum weekly benefit.
Whereas NML and ANI-MAERP issue policies to cover property in transit'as' well as policies to cover property at the plant site, NEIL-II issues excess 3 insurance only for on-site property.
- v.
L
_!(
+
y.s o
. C
, ym,
-t
-r 58 r
only if and to the.cxtent th'at the insured's loss exceeds $500 million.
NEIL-II then pays, subject td r$umerous qualifications and the specified maximum, the amount of loss in exusa of the $500 million deductible.
Unless the underlying coverage is in place to the full amount available, NEIL-II coverage.ts not normally offered.
2 UEIL-II provides coverage on the "followin6 form" basis. On this basis, NEIL-II covers losses that would have been,/-covered in the underlying i
insurance provided by ANI-MAERP or NML except for the fact that the s) underlyin6 insurance was not issued or that its limit was exhausted.
One modification of this general proposition is' that NEIL-II insurance c
applies to the perils of windstorm, flood, and earthquake, whether or
+-
D not the underlying insurance by en'dorsement or otherwise extends to these perils'. For these perils as for others, NEIL-II covers up to the limit of the NEIL-II policy only the' portion'of the loss above $500 m
i million.
3.
Prior to the date when NEIL-II can iscu at. least $500 million of insur-ance per insured per policy year, NE113s maximum liability per insured f
per policy year is a lower amount known as the " effective recovery t
limit."
For example, if an insured loss of $312 million occurred 4
- The basis for measuring the loss (e.g., actual cash value, or replacement cost now) is the same as in the underlyin6 insurance.
v The provision about underlying insurance not having been' bought was intended primarily to accommodate TVA.
As noted eldewhere in this report, TVA recently pur-chased underlying insurance from NML.
ns The NEIL-II policy formerly included a provision that in case the amount of insurance was inadequate to respond in full to all losses NEIL-II would pay.first for the decontamination costs. Because of a potential conflict between this pro-vision and certain provisions in the mortga6e indentures of the insureds, this pr ion was deleted. See Section VII. E for further discussion of this point.
Pending the completion of the NEIL-II program, the effective recovery limit is the sum of funds that would be available from retrospective premium assessments and from reinsurance. See Section IV, D for additional comment.
59 at a time when th,e effective recovery limit was $200 million, no more than $200.million would be payable.
?
4 Instead of relying on a coinsurance clause, the NEIL-II policy embodies a related but different technique. Any insured a6reeing to purchase as much insurance as is available up to $500 million, can collect in full for any loss up to the effective recovery limit. In terms of the preceding example, $200 million of the $312 million loss would be recov-erable, provided the insured covenanted to buy up to $500 million of excess insurance, as it became available.
II, by contrast, the insured agreed to purchase less than $500 million, say only $300 million, the effective recovery limit would be reduced proportionately.
In this example the insured would collect only three fifths of $200 million
($120 million) for the $312 million loss.
5 The maximum multiple of annual premium subject to the retrospective pre'ium adjustment is seven and one-half. The " worst case" fi6ure m
comparable to the ones cited above is fifty-two and one-half.
6.
The rate (price per 0100 of insurance) is the same for all insureds, with the result being that the premium is a function of the amount of coveragebought.
The charge per site is $1,700 per million of cover-age for the decontamination.
For everything else it is $500 per million dollars of'value in excess of 3500 million as reported most recently to the primary insurers.
If these rates hold when the full J500 million
- Again, the assumptions and caveats as observed above apply.
" Description...
- p. 12
- n To the extent the available amount is less than 3500 million, the charges are reduced proportionately. Letters of Ianuary 29, 1982 and February 17, 1982 to author fron ISIL.
m.
60 1
i of ICIL-l! insurance becomes available, the premium per site will be Q850,000 for decontamination and ^,,250,000 for the remainder of the i
coverage, producing a total of $1,100,000.
NEIL expects to introduce the use of classification criteria (such as constniction, exposure, occupancy, loss control, and the like) into NEIL-II ratemaking at an I
early future date.
I IV. AMOUNT OF NUCLEAR PPOPERTY INSURANCE AVAILABLE FROM EACH PRINCIPAL SOURCE The amount of nuclear property insurance offered for sale by each principal i
f source has undergone change. The changes are summarized in this section.
A A NI-MAERp 2
1 The changen in the amount of ANI-MAERP insurance are reported in Exhibit 8.
I The dollar figures, in millions, refer to the maximum aggregate amount payable i
by ANI-MAERP for one or more losser to any one nuclear site that occur during any one policy year, irrespective cf the r.cr 0er of reactors at the site.
Effective 'anuary 1,1982 ANf-MAERP increased the amount of its coverage to i
$450 million with approximately $340.6 million (75 7 percent) supplied by ANI l
cources and approximately $109 4 million (24 3 percent) supplied by MAERP.
Of the $450 million about 3156.2 million (34.7 percent) is supplied by domestic i
I insurers in ANI. about 350.4 million (11.2 percent) by MAERP domestic insurers, i
and about 0243 4 million (54.1 percent) by foreigr. nuclear pools, Lloyd's syndi-f cates, and individual foreign reinsurers through the arrangements described in i
I Section II.
B. NML 4
i n
The maximum amount of coverage initially offered by NML was $100 million.
Several changes have occurred, usually as of August 1 of the respective years.
- Letter of January 5,1982 to author from ANI.
..As with ANI-MAERP this amount is the maximum aggregate payable for one or more losses that occur at any one nuclear site during any one policy year.
61 EXHIBIT 8 MAXIMUM AMOUNT OF I!GUPANCE PER NUCLEAR SITE OFFERED BY ANI-MAERP 1957-1981 Year /s During Maximum (Dollars in Millions)
Uhich !!aximum Applied" ANI MAERP TOTAL D
b D
1957-1960 0 54.2
- 97 0 63 9 D
b 1961 53 7 97 63 4 1962 51.4 1g,)b 61 7 b
1963 50.1 10 60.1 1995-1965 50 to 60 1966-1968 60 14 74 1969 66.5 15 5 82 1970-1971 68 16 84 1972-1973 81 19 100 1974 102.7 27.3 130 D
1975-1976 134.8 40.2 175 1977 174 9 45 1 220 1978 173.6 46.4 220 1979-1980 232 5 67 5 300 83 8 375 1981 291.2 Change in maximum did not necessarily take place at beginning of year a
b Rounded c $75 million of this amount technically was excess insurance Source:
American Nuclear Insurer Reports, No.1 (Revised March,1981), p. 4, with 1981 figures supplied to author by ANI
62 in the maximum offered. These enlarged amounts and the respective years in which new amounts became available are summarized in Exhibit 9 As observed earlier, about 375 million of reinsurance excess of the first $100 million of loss per year and 050 million of reinsurance excess of $200 million have been (or are being) purchased by UML. This reinsurance is included in and is not in addition to the figures in Exhibit 9 EXHIBrr 9 l'AXIMUM AMOUNT OF INSURANCE PER NUCLEAR SITE OFFERED BY NML 1973-1981 Years During g,1 ch (Dollars in Iullions) a A
e 19 73
$ 100 1975 150 1977 175 19 78 225 1979 300 1980 375 1981 450 3
a Usually as of August 1 Source:
Letter of January 6,1982 to author from INL C.
NEIL-I The insurance offered by NEIL-I is quantified by the maximum weekly benefit that is payable as the result of one or more outages of one or more units at any given nuclear site beginning during the policy period. 'Jhen NEIL-I policies were initially offered, the naximum benefit available was $2 million per week during the fifty-two weeks beginning with the end of the 26-week waiting period and 31 million per week for the next fifty-tuo weeks, assuming continuing outage. As of September 15, 1981, those dollar amounts were increased respectively to $2.3 million and $1.15 million with regard to outages beginning on or after that date.
Description..., p. A-2
63 These limits apply to a nuclear utility site, whether one or more units suffer an outage. The less the interdependency between or among the units, the lower the cost, other things the same, of NEIL-I insurance applicable to a given multiple of units at any one plant site.
D.
NEIL-II NEIL-II launched its program with a maximum effective recovery limit of
$118 million as of November 15, 1981. The commitments UEIL had secured from reinsurers and members as of March 15, 1982 enabled it to set its effective recovery limit at the maximum at about $290 million of insurance excess over
$500 million. Again, this amount represented the maximum aggregate payable for one or more losses to any one nuclear site that occur during any one policy year.
Of the $290 million about $229 million was accounted for by NEIL's option to levy retrospective premium adjustments. The balance of the l290 million--about l61 million--was represented by reinsurance.
V.
ANTICII ATSD CHANG 33 DURING 1982 IN AMOUNTS AVAllABLS Ceveral increases are anticipated among the principal sources in the maximum anounts of nuclear property insurance that can be offered during 1982 A.
NML and NEIL-I An informed surmise of an MIL official is that NHL's Executive Committee will increase NML's maximum offering to $500 million when the committee next reviews the situation.
Apparently no plan exists to raise the naximum above $500 million.
According to NEIL, a decision is likely to be made at NEIL's annual meeting in June, 1982 about an increase in the maximum amount of insurance to be offered in NEIL-I.
Letter of February 17, 1982 to author from NEIL.
+. Letter of January 6,1982 to author from MIL.
... Letter of January 29, 1982 to author from NEIL. The author has no infor-mation about the size of the increase being considered.
64 h AllI-MAERP ANI-liAERP expects to increase its primary insurance capacity during 1982.
and also to develop an excess insurance capacity.
It plans to increase "in steps" the maximum primary limit to $500 million by the end of 1982 The initial incre-ment of 010 million (to produce a limit of $460 million) was scheduled to become availa' ale April 1,1982, with other increments expected to be announced during the year.
Iio plans are known to the author for ANI-14AERP to enlarge the primary insurance limit above 0500 million. The emphasis, rather, is to be on the devel-opment of excess insurance capacity.
AllI-!!AERP's current undertaking to develop excess insurance followed efforts in 1981 that failed to produce the desired results. The insurance trade press contained reports about ANI-MAERP's interest in using a Bermuda-domiciled subsidi-ary of several ANI member companies for the purpose of creating a layer of cover-age excess of 3500 million. The announced intent was to design a system of con-tingent liability of participating nuclear utilities for retrospective premium adjustments that would be similar to the system used for nuclear liability insur-ance.
The size visualized for the average assessment per nuclear unit per loss was 35 million.
If seventy nuclear units would have been represented among the participants, this layer would have provided the maximum of $350 million of cover-age.
The further plan was then to provide at the third level of coverage another
$150 million or so of fixed-premium insurance that would be excess over the X350 million or other amount available in the underlying layers." These efforts were unsuccessful.
The apparent reasons were a lack of sufficient interest among
' Letter of February 18, 1982 to author from MAERP.
See as examples of the news accounts of these efforts: Mary Jane Fisher,
"'4ashington Roundup: Congress Turns Attention to Insurance," The National Under-writer (Property & Casualty Insurance Ed.), Vol. 85 (May 15,1981), pp. 2, 58 and
" Nuclear Insurance Pools Hike Available Property Coverage," Business Insurance (riay 25,1981), pp. 2, 30.
65 nuclear utilities incommitting themselves to ANI-MAERF's proposed second layer and among insurers and reinsurers in subscribing to the third layer.
ANI-MAERP then worked out an arrangement with NEIL.
During this period UEIL was reported to have been encountering problems of its own in building the HEIL-II program as fast as it wished to do.
The cooperative efforts of ANI-MAERP and NEIL are described below. The description follows brief comments about the NEIL-II program.
C.
NEIL-II In 1981 NEIL launched a major program to enlarge the maximum limits of NEIL-II at least to 0500 million excess of $500 million. This objective was set forth as follows in August,1981:
This proposal [of ths Edison Electric Institute's Thsk Force on Nuclear Property Insurance, as mentioned earlier in this report) contemplates a new 3500 million layer of property insurance cover-age to meet losses in excess of $500 million. Of this new layer
$300 million would be provided by the participating utilities, through a combination of current premium and retrospective pre-mium adjustments. The remaining $200 million would be sought from commercial sources in the form of excess insurance or rein-surance of the program.
Increases in this coverage are contem-plated in the future.**
Apparently, NEIL's preference was to provide $300 million of the $500 million of excess insurance through retrospective premium commitments and the remaining 3200 million through reinsurance.
This arrangement would have been the one depicted as Alternative 1 in Exhibit 10.
Under this plan NEIL would have been
. Information provided to the author by ANI. Additionally, several nuclear utility risk managers voiced to the author their opinions that nuclear utilities generally would prefer to fund their retrospective premium adjustraents through a utility industry captive rather than through an insurance industry captive.
- Desc ription..., p. 2.
- In nuclear utility reinsurance as with the underlying primary insurance, the policy limit is the maximum aggregate amount payable for one or more losses to any one nuclear site occurring during any one policy year.
66 liable to the insured for up to $500 million of excess insurance and would have collected the appropriate premium for the coverage.
NEIL would then have entered a 40 percent quota chare reinsurance contract under which the reinsurer/s, in consideration of reinsurance premiums paid to the reinsurer/s by NEIL, would have agreed to pay NdIL 40 percent of each loss that NEIL paid to an insured nuclear utility. With NEIL liable for no more than $500 million to any insured, the reinsurer/c would be liable to NEIL for no more than $200 million.
In arriving at the premium to charge NEIL, the reinsurer/s night allow NEIL a " commission" of 15 percent or so against the premiums NEIL otherwise would owe.
Hearray suggests that NEIL was not able to find sufficient interest among reinsurers to implement this plan.
For whatever reason, this plan had not been completed as of this writing.
Purther hearsay evidence indicated that, upon encountering difficulty with Alternative 1, NEIL officials gave some attention to Alternative 2 as depicted in Exhibit 10.
Under this alternative NEIL would have been liable to the insured /s for 0300 millior. of insurance excess over $500 million in box two and would have collected premiums accordingly, probably relying heavily on the assessment tech-nique. The third level of 0200 million excess over )R00 million in box three would have been insurance as distinct from reinsurance and would have been supplied by the commercial market. The $200 million excess of $800 million apparently was not forthcoming either from ANI-MAERP or any other source. One reason could have been that no insurers were interested in exposing themselves up to $200 million per site per policy year for the small amount of premium envisaged for the third-level excess insurance.
l Other things the same the premium for a dollar of excess insurance tends to decrease as the level of the attachment point for the coverage rises. The reason is simply that a large loss is less probable than a smaller one.
67 EXH1 BIT 10 TFREE ALTERNATIVE WAYS TO DESIGN THE $500 MILLION EXCESS OF $500 MILLION PACKAGE Alternative i Use of Quota Share Reinsurance 1500 Million Excess Incuiance in NEIL
$300 Million
$200 Million Unreinsured Portion Reinsured Portion i
(Box 2)
(Box 3)
$500 Million Trimary Insurance in ANI-MAERP or NML (Box 1)
Alternative 2 Use of Two Layers of Excess Insurance
$256 Million Excess Insurance over $800 Million in ANI-MAERP or other Commercial Market (Box 3)
$300 Million Excess Insurance over $500 Million in NEIL (Box 2) i I
$500 Million Priiaary Insurance in ANI-MAERP or NML (Box 1)
Alternative ]
Use of Reinsurance and Quota Share Excess Insurance
$400 Million Excess Insurance in NEIL
$100 Million Excess
$300 Million l' einsured Portion
$100 Million Insurance (Box 2)
Reinsured in Portion A NI-MAERP l
(Box 3 (Box 4) 3500 Million Primary coverage in A!!I-l'AERP or I.74L (Box 1)
68 In any case NSIL apparently found itself left with Alternative 3 in Exhibit
- 10. This alternative is discussed in the following paragraphs.
A NEIL and ANI-MAERP Cooperative Efforts Perhaps because of the difficulties each encountered in attempting to gen-erate excess capacity or perhaps for some other reason, NEIL and ANI-MAERP were engaged in a cooperative program at the time of this writing.
Tne plan, as explained to the author by ANI-MAERP officials and a NEIL consultant, is depicted by Alternative 3 in Exhibit 10.
In this arrangement NEIL would fill boxes two and three. Box two would represent the capacity provided by retrospective pre-mium assesc.ments; box three would represent reinsurance. As already observed, NEIL had commitments of $229 million in box two and $61 million in box three as of March 15, 1982 The total was $290 million.
ANI-MAERP had undertaken to provide advance premium excess insurance to com-plement the NEIL-II offering.
This coverage would fit in box four of Alternative 3 in Exhibit 10. ANI-MAERP would issue its own policy but the coverage would be consistent with that in the NEIL-II policy. ANI-MAERP is on record as hoping to generate $100 million in such capacity before the end of 1982 and to have $50 million of this capacity available by April 15, 1982 As the author understands, the insurance to be provided by ANI-MAERP is to be of the quota share type. An example will show how the Alternative 3 pattern would function. Suppose that ANI-MAERP succeeds in developing $50 million of excess insurance capacity at a time when the NEIL-II capacity remains at $290 million.
Suppose further that a utility purchases excess insurance in the amount
- Thus the NEIL-II effective recovery limit as of March 15,1982 was $290 million.
- Letter of February 18, 1982 to author from MAERP.
69 of $290 million from NEIL-II and $50 million from ANI-MAERP. For any insured loss suffered by the utility in excess of $500 million, NEIL-II would pay 290/340 and ANI-MAERP would pay 50/340 up to the point where NEIL-II would have paid $290 million and ANI-MAERP $50 million.
Each would be directly liable to the insured utility for its portion of the loss, but NEIL would collect one-fourth of its portion from its reinc t er/s, assuming it bought quota share reinsurance.
In this example only $340 million of excess insurance would be available at the maximum.
If and as NEIL-II and ANI-MAERP develop additional capacity, the ratio will change. Officers of the respective organizations apparently expect that by the end of 1982 ANI-MAERP will have generated a capacity of $100 million excess insurance and that NEIL-II will have provided $400 million.
If these expectations come to fulfillment, a total of $1 billion of primary and excess nuclear property insurance will be available (assuming a utility buys the primary insurance from ANI-MAERP or NML but not from both). Conceivably, the total could grow even larger over the longer-term future.
VI.
LIKELIHOOD OF EMERGENCE DURING 1982 OF NE'l IRINCIPAL SOURCE /S IN FRIVATE SDCTOR In this section attention is directed to what, if any, new principal source /s of nuclear property insurance is/are likely to emerge in the private sector in 1982 In the author's considered judgment the answer is none. The reasoning supporting this judgment is briefly reviewed.
A.
Limited Interest of Private Sector in Nuclear Property Insurance The chief reason the supply of nuclear property insurance has not grown faster doubtless has been the perception of the market potential by actual and potential suppliers.
This perception has not been as enthusiastic as some per-sons would have preferred.
70 The grounds for absence of greater enthusiasm are numerous.
Except for the pools' staffs and the insurers' representatives, who serve on various pool commit-l tees, insurers have little knowledge about the contingencies subject to the insur-ance, especially about accidental discharge of radiation at nuclear utility sites.
The exposed sites are small'in number, large in value, and not necessay'ly homo-geneous. As the result, predictions that insurers have to make about I ature losses in order to set dependable rates for the insurance are subject to large error. The catastrophe potential is severe. Because of pressures from state and federal regulatory agencies to hold down electric utility rates and because of the economic clout provided by the magnitude of utility purchases, nuclear utilities are notorious within the insurance community as being "hard bargainers" in insurance transactions. Beyond these considerations is the realization by insurers that nuclear power installations are vulnerable to being damaged through cabotage, outright armed attack, or other means by individuals or groups opposed i
to the operation of such facilities. Finally, the effect of the Three Mile Island accident on the profitability of participation in ANI-MAERP is viewed as having been depressing.
Whether or not this perception by insurers of the nuclear property insurance i
market is accurate may be beside the point.
If it is sufficiently widely held, as l
this author believes to be the case, it will act as a serious impediment to growth of supply.
B.
Efficient Organization of Private Sector Another reason for the speculation that no new principal source of nuclear property insurance is likely soon to emerge is that the present market organiza-tion is reasonably efficient. The author is not aware of any reorganization of I
the private-sector market being proposed within or outside of the insurance community with the aim of increasing the supply. This point warrants elaboration.
l l
t
71 Observers of the nuclear property insurance market cannot help but be curi-ous as to whether the establishment of the ANI and MAEPP pools, while structur-ing the supply of this insurance, might also have constrained its growth. This author was subject to such curiosity.
The possibility may exist that, had each insurer operated independently, the aggregate supply would have been larger than, albeit not so standardized as, is now the case.
Having completed the study, this author does not regard the probability that the ANI and MAERP pools to date have hindered the growth of nuclear property insurance in the private sector as being significantly above zero.
As already discussed, the insurance is relatively new.
It is also complex and unique. An extraordinary degree of expertise is needed to underwrite and price the insurance as well as to exercise loss control.
Individual insurers are not likely to have such expertise or be interested in paying the high costs of developing it independently.
'Jithout its availability from a convenient source, an innurer may be less prone to write nuclear property insurance than if the exper-tise were available on reasonable terms.
The establishment of pools is a time-honored response by insurers to provide insurance that does not fit marketing channels existing when the need for the insurance arises.
lools, for example, are used to service exposures related to shipping, railroading, motion picture producing, oil drilling and refining, cotton processing, and other complex activities. Iools also are used to cover aircraf t, bridges, granaries, factories, and numerous other large objects. Given the decree of competition in the property-liability insurance industry, these pools would not have been likely to survive had they impeded competing insurers from offering their products in the varieties and quantities and at the prices favored by these competing insurers.
72 4
i The use of pools tends to maximize supply of insurance by maximizing the participation of reinsurers. Much reinsuring is done by use of treaties.
j Because multiple and intricate treaties are used and because the reinsurance, a
itself, can be reinsured, prudent reinsurers must exercise caution in the extent to which they accept cedings, some of which, in fact, may be retro-cessions of the first or higher degree.
For example, Company A may cede business under a treaty to Company B and under another treaty to Company C.
Company B may have a treaty with Company C as well as treaties with Companies D and E, each of which also may have a treaty to cede a portion of each transaction to Company C.
In the absence of precautions to the contrary, Company C could accept under its treaties four
" pieces" (but not necessarily the same size) of the single exposure insured by Company A.
Pour pieces of this exposure may be more than Company C would want.
]
Company C's natural defense in an. unstructured market would be to hold down its treaty commitments with all companies, lest it unwittingly become responsible for an inordinate proportion of any single exposure.
Had each of these companies been participating in a pool in which each agreed to retain the exposure to which it committed itself, each company might have been better informed about the nature of its exposures.
- In "facultative" reinsurance the reinsurer and the company wanting to buy the reinsurance (called the " ceding" or primary company) negotiate each trans-action.
In " treaty" reinsurance the reinsurer and the primary company agree l
in advance that stipulated amounts or percentages of various types of coverages underwritten by the primary company will be ceded to the reinsurer.
In either type of reinsurance, however, the primary company remains liable to the insured as though no reinsurance had been bought and the reinsurer is liable only to the l
primary company according to the terms of the reinsurance.
t i
The process of a reinsurer buying reinsurance is called "retrocession" and I
can occur more than once to produce reinsurance to the " nth degree." Reinsurance of reinsurance is called reinsurance and not re-reinsurance.
l
- Much reinsurance is done by companies that sell only reinsurance.
Some reinsurance, however, is sold by companies that also sell insurance.
73 In its simplest tenas with regard to nuclear property insurance, the argu-ment about reinsurance is that a structured market is likely to evoke more participation by companies that sell reinsurance than would an unstructured market. To the extent the argument is valid, removal of the structuring pro-vided by ANI and MAERP could reduce rather than increase the supply of insurance.
Finally, the social pressure on reluctant insurers to cooperate in supply-ing insurance regarded at least by some persons as serving unusually important national interests could be stronger in a structured market than in an unstruc-tured one.
More publicity is likely to attend the project in the structured than in the unstructured market.
If nuclear property insurance serves such an interest, the pressure on insurere may be stronger coming from ANI and MAERP than would otherwise be the case.
In summary, no compelling reason exists to think that insurers in the private sector would make more nuclear - operty insurance available, were the ANI &nd MAURP pools to cease operations.
R.d the interest in supplying insurance to the nuclear utilities on mutually acceptable terms been sufficiently strong, nothing would have prevented existing insurers from independently offering coverage or from forming new coalitions to do so.
Similarly, new specialized insurers (aside from the utility captives) could have been established in the private sector to serve the nuclear utility market, had entrepreneurs been convinced that profit potential warranted such action. The failure of these developments yet to occur is evidence that supply of nuclear property reinsurance in the private sector is not susceptible to major short-run growth.
C.
APIC's Slow Growth A source of insurance not mentioned heretofore in this report is the American Power Insurance Corporation ( APIC). This organisation, a nonprofit corporation established under the auspices of the American Fublic Power Association, provides
?4 a time-element coverage called Continuing Expenses Insurance. The intent of the insurer in to reimburse an insured utility for at least a portion of the expenses that would continue during the outage of a nuclear reactor. For the insurance to apply the outage must have occurred from a loss that would evoke payment under an ANI-MAERP policy as described above.
The " continuing expenses" include debt payment (principal as well as interest), payroll expenses that necessarily con-tinue, ordinary maintenance and security costs, and "other costs which are not diminished as a result of a power outage."
The insurer promises to pay a daily indemnity of no more than $150,000 for up to no more than 730 days of indemnity, except that the total indemnity cannot exceed 350 million.
Furthermore, the daily indemnity payments are not made for the initial 120-day period of any outage. Provision is made for smaller payments in case of a partial shutdown.
No provision is made in the specimen policy for other contributing insurance, but permission is given for the insured to purchase similar insurance that is excess over the limits of this insurance.
An important exclusion in the APIC policy excuses the insurer from paying the daily indemnity because of NRC delay. The exclusion pertains to any suspen-sion of oporations of the facility or any delay in the resumption of operations when the suspension or delay in resumption is "due to an order or other action "No mention is made in the specimen policy about accommodating underlying insurance in NML.
This wording is from the introduction to the specimen policy. Because the cost of replacement power is one that is newly incurred in the event of an out-age rather than one that is "not dininished" by the outage, this insurance apparently does not extend to such cost. The APIC policy covers " extraordinary expenses as are necessarily incurred...for... reducing any loss under this Policy...
not exceeding...the amount by which the loss under this policy is reduced thereby."
The purchase of replacement power would not necessarily reduce the duration of the outage.
It would, therefore, not necessarily reduce the amount payable under the APIC policy.
For that reason the purchase apparently would not be covered by APIC.
l
75 or failure to act by the Nuclear Regulatory Commission (or any other regulatory or governmental authority)" that extends the outage beyond the time required with due diligence and dispatch to restore operations.
APIC is managed by a large brokerage firm (Fred S. James & Company) and is served by an Underwriting Advisory Committee comprised of insurance industry officials familiar with nuclear reactor insurance.
Details about APIC's financial structure were not available to the author. Presumably, virtually all of the exposure to which AFIC commits itself is reinsured in the United States or abroad.
As of April 30, 1981 APIC had issued only one policy.
The fact that APIC's clientele is so limited is the reason that APIC insur-ance was not discussed in Sections II, III, IV, and V.
On the basis of APIC insurance in force APIC can hardly be regarded as a principal source of nuclear property insurance. The author is not aware of any reason to anticipate any major surge in demand for this coverage by nuclear utilities.
D.
Icolated Purchases Have Been Iade The foregoing discussion is not intended to suggest that no additional cover-age will be forthcoming from new sources of supply.
Indeed, come utilities have been able to fill all or part of the $50 million cap between the maximum of ANI-MAE'l or I.ML underlying insura'nce and the touchpoint of NEIL-II excess insurance by reliance upon their own brokers who have found insurers willing to supply the Doubtless, other nuclear utilities not yet patronizing HEIL-II would coverage.
be able to add a few million dollars of coverage to their portfolios by aggressively combing the world markets.
The thrust of this section has been nerely to report the absence of evidence that any principal source of supply, conparable to ANI-f:A.0F, ITL, or NEIL-II, is Creater Connitment 1:eeded to Solve Continuing Froblens at Three Mile Island, Staff of U.S. General Accounting Office (August,1981), Chapter 6, p. 81. Accord-ing to hearsay evidence, linitations on the available reinsurance resulted in the anount of coverage for this one policy being reduced somewhat below 350 million at the renewal of the policy.
76 in the process of being established in the private sector.
hindful that such development takes time, the author feels confident in speculating that no new principal private sector source is likely to emerge in 1982.
VII. FROBL~mu AND '$0 MEND 3D EC PIl0NSE3 The system of nuclear property insurance as it operates in the United States is beset with several major problems.
In this section six of these problems are discussed. For each one the autaor recommends a course of action to be pursued by URC as a response to the problem. The order in which the problems are dis-cussed is not necessarily the order of increasing or decreasing urgency as seen by the author.
3ather, it is merely a convenient order in which to discuss the problens and offer the recommendations.
A.
Historical Unreliability of Assessment Technique One problem with the nuclear property insurance systen is the heavy reliance of NML and UEIL on the assessment technique. The option of these insurers to call for " retrospective premium adjustments" is discussed in Section III. The history of use of assessable policies in traditional lines of insurance is not inspiring. Too often insureds, upon being assessed, have been unable to pay the assessments or have resisted them in the belief that they were unfair, unnecessary, or otherwise unwarranted.
The result has been that some insurers, depending on assessments, have not been able to pay the insured losses.
The danger of insurer default is perhaps nost pronounced with respect to cataetrophic loss that occurs early before the insurer has had time to build its net worth through retention of earnings.
Critics of assessment insurance refer to it derisively, but with some accuracy, as " funny insurance." Attention is called in Section III to the peculiar provi-sion in NEIL policies that in the event of multiple claims that exceed the insurer's resources, NEIL's liability is discharged by paying the competing l
??
claims on a pro rata basis. This provision emphasizes the difficulty faced by a new cmall mutual in responding to early catastrophes. The explicit statement in NML policies that IML's liability is limited to its resources is perhaps of similar import.
peason for Use. Historically, the. assessment technique has been used by newly forned mutual companies.
An insurer formed as a stock company can develop an initial not worth position by selling shares of ownership in an amount at least as large as the minimum capital requirements imposed by the state of the company's domicile. An insurer formed as a mutual does not usually have the option of sell-ing stock.
Consequently, its net worth is likely to be small.
A mutual company may be required, according to the law of the state of its incorporation, to obtain commitments from at least a specified minimum number of potential in-sureds for at least a specified minimum amount of insurance before it issues any
, A few insurers in certain lines have relied exclusively on assessments.
These insurers are said to use " pure assessments." Insurance paid for exclusively by assessments is also called " post-loss assessment" insurance. Examples in life insurance could be found among burial societies (often called mutual benefit societies).
In fire insurance a few cmall county mutuals used the pure assessment technique.
Virtually all of these organizations abandoned the technique or went out of business. More commonly, the assessment feature is used in conjunction with a modest " advance premium" structure. The assessment option provides the insurer the means of generating additional funds, should they be needed, to pay insuredlosses, expenses,and/ortaxes. The UML and NEIL policies combine assessability with relatively small advance premiums.
- Oddly enough, the statutes of some states allow a newly forned mutual in-surer to issue " temporary stock" to its incorporators in exchange for " temporary capital." The stock is retired" as soon as retained earnings reach the level d
prescribed in the statute.
See, for example, Minn. Stat. Ann., Vol. 6, Chap.
60A.07, 3ubd. 10(1), p. 83 (Uest's, 1968).
- n In insurance terminology net worth, whether for a stock or mutual company, is often referred to in its entirety as " policyholders surplus."
78 policies.
To afford the fledgling mutual company that has no base of contributed capital at least a limited ability to respond to substantial early losses, the ntate in likely also to require that the company's insurance policies be assess-able. The statute also nay ntipulate:
1.
That the aggregate accensmente levied against any insured cannot exceed a specified maximum, which maximum may be some multiple of the " regular premium."
2 That the accecament can be levied against any insured to whom a policy was icsued during the year for which the assessment was made.
3.
That the calling of an accessment, cubject to the constraints in 1 and 2 above, is at the cole option of the insurer's board of directors.
4 That the accessment in payable, whether or not the insured on whom it is levied chooses to cancel the insurance.
The potential accessmente (retrospective premium adjustments) that can be levied by NHL and NEIL are concistent with the pattern described above.
The assessments are authorized because of the absence in UML and NEIL of a cubstan-tial base of contributed capital.
If each of the two compan3es continues to
'As an example, an Indiana statute pertaining to " farmers mutual incurance companies" writing limited coverages specifies that no policy of insurance shall be issued by a given company until that company "shall hold" at least 250 appli-cations for innurance aggregating at least 0500,000. EUDNS IND. STAT., Sec. 27 1-4 (1975 Ed.).
Formation in Indiana of other mutuals is subject to more stringent requirements. An Alabama statute specifies that formation in Alabama of a mutual insurer to issue property insurance requires at least 100 applications covering at least 250 properties with no less than $1,000 of insurance on each one at rates custonarily charged by other companies for each coverage. AIA. CODE, Sec. 27-27-15 (1975 Ed.). As a general proponition, state laws in recent years have been enacted to render mutual insurers increasingly difficult to organize in the United States.
- The stipulation in UEIL policies of a " reserve premium" of 13 percent of the initial premium is a small effort to compensate for lack of substantial contributed capital.
l
79 operate profitably and to build its not worth, each will ultimately arrive at the position in which it can dispense with the assessable feature as to newly issued policies.
Resistance of Insureds. The inherent difficulty observed in the use of assecs-nents in non-nuclear lines of insurance has been that those upon whom they are levied may not pay then and that the insurer, therefore, may default on its obli-gations.
Assessments, beinc essentially unuelcome by those upon whon they are levied, tend to foster an adversarial relationship.
Insurers may be under heavy pressure to make the assessments.
Insureds who have not suffered a loes but upon whom assessnonts are levied have an incentive to look for excuses to resist the assessments. The result is that asseconents breed litigation.
In an effort to avoid favoring one party at the expense of the other, courts tend to look pains-takingly at the statutes and at the policy relevant to a given dispute and to interpret the wording rigorously.
- Because the assescable feature in a dense is an admission of financial inade-quacy, mutual insurers normally are enger to build sufficient net worth so as to discard the assecsnent provision and not include it in newly issued policies.
Doubtless, ZL and NEIL aspire soon to outgrow assessability, as did nany large nutuals whose names today are uidely known.
A Michigan statute, for exanple, allous a mutual fire insurance company to drop the assessnent feature when the company's policyholders surplus reaches at least the $200,000 level. MICH. COM-PILCD LAWG ANN., Vol. 28, Chap. 68, Sec. 500.6840(3), p. 813 (vect's,1967).
- The author recognises that in nuclear property insurance NRC could bring its considerable influence to bear upon any recalcitrant utilities that reneged on assessnents.
He does not feel, however, that URC should be called upon for such a role.
- A case in point at the time of this writing is an injunction issued on Febru-ary 25,1982 in the Leon County Circuit Court in Florida and extended the next day by an appeals court. The court restrained the Florida Fatient's Conpensation Fund fron collecting about 85 percent of the retroactive assessnent recently levied by the Fund. The Fund is a state-chartered insurance mechanisn to provide nedical nalpractice liability insurance excess of ;100,000 per clain and 0500,000 per occurrence to about 120 hospitals and about 7,000 physicians and other professionals in Florida. Unable to pay its clains, the Fund recently levied acainst the nenber hospitals for the 1978-1979 plan year an assessment substantially larger than the preniuns initially charged for the year.
A sleeper provision in the legislation creating the fund exenpts physicians from assessnent in excess of the initial prenluns. The bospitals disputed this provision as inequitable and brought suit to block the assessnent.
" Florida Medical Malpractice Fool 'dants Hospitals to Fund Big Deficit," Business Insurance (March 8,1982), pp.1, h7
Po Fecognition of three of the interpretative precepts used by the courts chould be cufficient to establich the potential for dispute.
1.
One's liability to pay an assecsnent and the concequence of nonpaynent, cubject to statutory constraint, depend in each case upon the terns of the contract.
2 No acceccnent, however, is payable unless it is reasonable and levied in strict accordance with the authority conferred by the insurer's charter, by-laws, and policy.
3 No precunption exicts that an accessment ic valid, and the burden of
+
proof rects upon the incurer.
Illuctrative of the grounds on which nutual incurance company policyholders recict payment of acceconents are those enumerated below. Gone proved to be cuccensful defences; others did not.
1.
An automobile insurance policyholder in Kentucky contested liability for an ansecsnent demanded by a nutual incurance company doniciled in Ucw York for the reason that the notice of contingent liability for acccconent was only on the back of the policy. The Court of Appeals of Kentucky held that such contingent liability was excluded on theegrounds that the notice was not an endorcenent.
2 A policyholder of a nutual insurance conpany attacked an accessment as invalid (but wac not successful in escaping liability for it) on the Ground that the Colorado incurance connicsioner unlaufully issued certificates of authority pernitting the company to do business in Gunnariced fron Willian R. Vance and Bulst M. Anderson, Handbook on the Law of Incurance (3d ed.; St. Paul: West Itblishing Co.,1951), pp. 5, 72,1f99-300, 343-3h7 Puqua Euc Line v. Iink,160 3.t'. 24 646 (1942).
=_
81 i
Colorado in various yearc when its capital was impaired. The Supreme Court of Colorado held that the policyholder was liable.
3 A Pennsylvania policyholder of a Pennsylvania mutual insurance company was not liable to pay an assessment for losses incurred before his I
policy was issued, according to a decision of the Pennsylvania Supreme
^
i Court.
4 A Michigan policyholder of an Illinois mutual insurer was not held lia-ble upon being sued in Plchigan for an assessment where no demand was made for payment of the assessment within one year after termination of the policy. The Supreme Court of Michigan held that the rule of j
comity was not allowed to contravene the rights of the Michigan policy-
]
holder.
5 An insurer (Teorless) buying reinsurance from a mutual insurer that issued assessable policies (Federal Mutual Casualty Company) was deemed by the Supreme Court of Wisconsin not to have become a member-policy-holder of the mutual insurer and not to be liable for an assessment levied on behalf of the insolvent mutual insurer by the State's insur-ance commissioner after the mutual insurer had been placed in liquida-tion.
Possible Confluence of Calls. Much of the case law on insurance policy assess-ments involves disputes in which life insurance mutual benefit associations or county fire insurance mutuals were parties. An argument could be made that the behavior of insureds in these cases is not likely to typify the expected behavior Aronoff v. Fioneer Mutual Compensation Company, 304 P. 2d 1083, (1956).
" Capital City Mutual Fire Insurance Company v. Boggs,172 Pa. 91, 33 A.349 (1895).
- Kechn v. Charles J. Rogers, Inc.,18 N.U. 2d 877 (1945).
- Feerless Insurance Company v. Hanson,135 n.v. 2d 258 (1965).
82 of the nucler. utilities that nicht be called upon to pay UML and/or NEIL assess-ments. Ecing large and sophisticated enterprices, so the argument goes, the nuclear utilities would be likely to accept the assessments gracefully. The argument also stresses that the accessments, if called at all, would be levied by a captive insurer that serves only the nuclear utilities. The implication is that co-controllers of an industry captive would have few reasons to disagree.
The author readily agrees that the members of RIL and the members of UEIL l
are nuch more likely to have a connunity of interent than are the insureds who l
made up the nutual benefit societies or the other mutual insurers cited above.
Yet, the occasion that uould produce a need to call for a retrospective preniun adjustnent in UML or UEIL would be, by its very nature, divisive.
It would cre-ate severe conflicts of interests. One or a feu nuclear utilities would have suffered heavy insured loss and would be looking to the insurer for payment. The other members of the insurer would be under pressure to hold down the magnitude of the prenini. adjustment. This pressure would be increased by the absence of certainty as to how soon, if ever, the appropriate federal or state public rec-ulatory agency would allow a utility subject to an assessment to pass it on to its customers in the forn of a rate charge.
Apprehension about such allowance could cause the utilities subject to assecsnent to be less gracious in their response than might otherwise be the case.
No guarantee exists that any retro-spective preniun adjustnent by NML or UEIL would not be contested.
The staff of the General Accounting Office broached this subject with repre-sentatives of the public service commissions in New York, Florida, Illinois,
'Jicconsin, and California. These representatives speculated that some connissions perhaps would allow a pass-through of some or all of the accessnents. The CAO observed, however, that these representatives "could neither connit their commis-clons to this course of action nor predict the actions of any future state con-nissions. " Creater Connitnent Needed to Solve Continuing Problems at Three Mile Island, Staff of U.S. General Accounting Office (1981), pp. 84-85 ~~
83 The probability of content in enhanced by the poccibility of confluence of acccconents under two or nore progranc. An accident that leads to a large accccc-nent by II::L could alco lead to a large acccconent in ITEIL-I and I!EIL-II. Uorce yet, it could alco lead to an anceccnent for liability incurance purpoces under the ternn of the Irice-Andercon Act.
Zvery nuclear utility licennee could be j
cubject to the acccconent related to the Frice-Andercon Act.
The extent of over-lap of the other accecenents would depend on the extent to uhich the nuclear utilition patronizing I;I L alco patronized liEIL-I and IICIL-II.
In the event of the naxinun accencnentn for any one year in all there progranc, a utility in In:L, I;UIL-I, and !iEIL-II, could be called upon to pay acccccnento equal to four-teen (or cono lcwor e/itiple) tinen its IIML preniun, coven and one-half tinen itc
~
ICIL-I icenlun, five timec itc ITEIL-II preniunn, and 05 nillion per reactor for third-party liability clainn under Irice-Andercon.
Cuch a confluence could produce enouch problenn to tenpt even the noct cooperative insured to ceize upon
=
technicalities to delay paynents, if not to try to wriccle out altocether.
Adnittedly, the probability of thic "worct cace" coning to pacc ac depicted otove nay 1e remote. Lefore lem.es extended over a ceven-year period the utility nicht uithdraw fron I.I anr: I;is!L.
.;ven if the " worst cace" did cone to racc, the accentnente ni;;ht be payable over a period of yearn instead of uithin a few dayc.
3 3
Yet, by their nature, incurern are expected to be able to recpond to "norst cacec "
'The acceccable rolicicc in ZL and I; MIL do cpecify that the accestmente are due A
^
+
U.
.J..ec. C210(b) providen that, upon the occurrence of a nuclear inci-dent that producec liability on the part of a nuclear utility in excecc of the liability incurance applicable, the nuclear utility can he accenced up to ;5 nillion tine: eacn licenced reactor owned.
Various other qualifications and limitations appl:/.
r
.* Attention in called in Cection I:: to the even larger contincent liabilit:,
that could accunulate over a neven year period becauce of the fact that a retro-npective preniur adjuctnent renainc callable for cix yearc beyond the end of the year to @ich the call rould apply.
=
m
Bh and payable uithin tuenty daye after they are called.
In evaluating incurance one ic juntified in contenplating the " worst cases" that could occur.
- 'econnended URC Action.
In this author'c judcnont I3C cannot and should not attempt to alter the accessable features in the existing II:L and :CIL charters, by-laws, and policies.
Zven thou6h tre anneccnent incurance may deserve the appellation of " funny innurance," 13C has little choice but to accept asseccment insurance ac being the equivalent of advance preniun innurance. The process of developing the nuclear property insurance synten is too far along to justify a najor d:nnantling and restructuring.
U.F does Fave a choice, Powever, about use of retrospective prenium adjunt-nents in any neu type of incurance that night be developed. At sono point the nultiple and cumulative assessments arising out of nunerous congruent incurance progrann can overwheln the nuclear utilities, suffering as they are fron so many other denande for cach.
To continue to build possible assessments upon ponsible Counnel for II::L and UEIL nade the follouing interesting observation to the author:
If one accepts the prenise that over the intermediate run and the long run the nuclear utilities must bear their own insured losses, perhaps tbc differ-encen between ?.1:L (or USIL) and ANI-MAZRF are more apparent than real.
"'he " worst cace," an nentioned above, would call for asseccnents in UML and NEIL. Were it to fal] upon ANI-l'AIEl', houever, it doubtless would evoke subetantial rate increaces for utilitico inured in ANI-MAU31.
Under either nysten, ecpecially if insured loucen occurred over a period of several years, a nuclear utility--ascuning it wanted to retain its insurance--would have to pay its share of such losses.
In ANI-MATT the paynents would be in the form of increased prenluns; in NML or NEIL they would be in assecsnents an uell, perhapu, as in increased premiums. The further point uns nade that ANI-MAERI prices doubtless would be designed to recap-ture over a few years fron ANI-l'.A3RF insureds ac a group all large paynents of clains to the innureds suffering lossen.
In 1976 1rofessor Ronald U. F.elcher wrote that the investor-owned utilities, the TVA, and the larger municipal-owned utilities probably could meet their assess-nents e.er the near tern.
He saw a potential problem over the longer run in the availability of cacb.
Fe poJnted out that mere retention of earnings was no guarantee of the availability of cash to pay assescrnnts.
He said that utilities should have ntandby letterc of credit for this purpose and, failing to have them, should segrecate cash.
See his Financial Inplications of Retrospective Frenium Assessmentn, U.S. Nuclear Regulatory Connission, NR-AIC-003 (1976).
Gf much inportance, however, is the fact that Professor Melcher was considering only the ussecsnents that could be callable for nuclear liability insurance under the Price-Anderson Act.
He was not taking into account the additional and possibly larger accessnents callable by NHL and NSIL (for the extra expense and the excess prograns)..
I 85 accennnente could cause the entire superstructure to cag, if not to collapre, at the very tine of nociety's nort critical need for it.
In thic author's view, the annennnent procrans now in une have reached the limits of prudence.
He recon-nendo that NRC not go beyond the progranc in existence in accepting acceccable innurance as caticfying nuclear property incurance requirements, b.
lanto Inherent in Dual Supplicn Anotbur problen With nuclear property incurance arices out of the dualien in the nupply at the prinary level. Nuclear utilition purchace incurance fron either ANI-I'LR1 or 1:ML but not from both. As the result, each of these principal courcen of cupply cervec on1. a portion of the narket. A given nuclear utility han (at current linits) no more than )60 nillion from one cource or the other inntoad of having '900 nillion (or yoncibly a conewhat lower amount) from both courcen in the argrerate.
If, indeed, the nuclear utilitico need and are willinc to purchace nore
'ncurance tran they have, the persistence of this dualion in the market place is i
puzzlinr. The following diccuccion of the phenonenon of dualinn in orcanized under four questionc.
"A ce in point bearn on financing the costs of deconniccioning. A recent ctudy, undertaken at the invitation of and with funding provided by lac, explores the reacibility cf une of another captive insurer to treat the costs of prenature deconnincioning of nuclear reactorn.
See Faul L. Chernick, 'llllian B. Fairley, I:ichael b. F. eyer, and Linda '.. Scharff (staff nenbers of Analycic and Inference, Inc.--3cnearch and Conculting), Decicn, Contn, and Acceptability of an L49ctric Utility Celf-Innurance lool for Accuring the Adecuacy of Funds for Nuclear Icuer ilant Deconnincioninc Uxpennen, U.S. Nuclear hegulatory Conniccion, NUIEG/C3-2J70 (19S1). She authorn of thic report conclude generally that acceccable Jncurance nay 're uneful in financing conta of deconniscioning necessitated by an accident but exprens cono recervationn about whether coctc of deconniccioning not related to accidentn are incurable.
Thic author does not take iccue with thene findingn.
Hir only contention in that at come point one nore acceccable incurance progran ic at leant one too nany.
For thic reason, in hic judcnent, N3': chould not encourace the fornation of the captive, unleuc it would incue advance preniun, nonanceccable incurance.
86 Capacity g NI!L and AIII-IM3F itch M Serve All Nuclear Utilities. Attention is given first to the questian of uhether NHL and ANI-!!AEPP each could serve all nuclear utillt'les in need of nuclear property insurance.
In this author's view no limitation in insurance theory requires a negative answer to the question.
Cone adjustnents night be called for, however, in the operation of the NML, ANI-MAi2P, and ITdIl -II programs.
Insurance theory suggests that accuracy of prediction of losses tends to inprove, other things the same, as the nunber of reasonably honogeneous units exposed to loss and subject to the prediction increases. The point is made several times in this report that the operating base of each principal source of nuclear property $nsurance is too small to render the loss experience credible for rate-naking purposes.
Expansion of the base of ANI-MAERP and ITiL would tend to add stability to the underwriting results of each enterprise because of the larger numbers that could then be used by each in predicting future loss frequency and severity. Uhile the resulting numbers would still be too anall to lead to full actuarial reliability, the resulting increased stability would be a novement in the right direction.
To the extent nuclear utilities patronized to the maximum both AUI-CAdnF and NML, the activities of each incuring enterprise would expand. This expansion could introduce some econonies of scale so as to render each enterprise more effi-cient than it is nou.
The possible reduction in the variability of insurea losses and in the magnitude of operating expenses per unit of premium could be examples of econonies of scale because both could lead to rate reductions.
Were the change under discussion to come about, the resulting combined pri-nary capacity would not likely be as large as the sun of the individual capacities of UML and A!!1-MAEEP under the present dualism. The reason is that some of the
+For exanple, the NEIL-II program would require a higher attachnent point, were
!.11 and ANI-IIAERP each to serve all nuclear utilities, l
87 insurance organizations that cell reinsurance to 134L may also sell reinsurance to AMI-MAEPP or may be direct participants either in ANI or MAE7F. One or more of such organizations nicht not want to have the sun of its IU4L and ANI-I4AS7F connit-nents come to rest on any one nuclear reactor.
Any organization in this situa-tion nicht choose to cut back on its connitnent to :31L, its connitnent to ANI-MASRP, or none conbination of the tuo. No reason is apparent uhy it would want to clininate its connitnents altogether.
At the norse, the loss of capacity should not exceed an anount equal to the volune of reinsurance purchased by NHL.
As observed in Section II, B of this report, IC:L has purchased (or uill purchase) about $125 nillion of reinsurance.
The conbined capacity, therefore, should not be any less than about 0775 nillion (that is, the $50 nillion fron m:L plus the A50 million fron AIII-I'AERP less the possible loss of 2125 million from reinsurers that now deal with both HML and AHI-MAE7F). The erosion nicht not be nearly that auch.
The nochanien by neans of which Alii-I:AII and IU:L could serve the same insureds are not complex. At the least integrated extrene, each entity would une its own policy, set its o n rate, and otheruise operate about as it does nou.
At the nost intecrated extrene, the two entities would insea a connon policy and coordinate their pricinc, underwriting,and clains adjusting.
Deveral alternatives lie between the extrenen. Gne of then is the use of separate but unifornly priced policies.
Uced for "ach 3 Gerve All Uuclear Utilities. To grant for the sake of dis-cussion that I1L and ANI-I'.ATI could serve the sane insureds is to confront another question:
dhould they do so? The ansuer to this question depends perhaps on whose point of vieu is beinc considered.
In this discussion society's point of vie:: is adopted as being overriding of all others.
- The extent to which, if at all, any of these alternatives could raise anti-trust problens is a natter the autho-lea'ies to othern.
P8 tiithout any doubt on the part of the author, society has a powerful interest in overy nuclear utility being able to prevent the harn te society that could be produced through release of radioactive contanination. The option of a nuclear utility to expone cociety to a dancer--however renote--carries with it the duty to protect cociety from that danger, chould the danger naterialize. A nuclear utility, therefore, has among other duties the duty to prevent the accidental release of radioactive contamination fron its prenises. In the event of an acci-dent this prevention vould involve, anong other possible actions, the decontani-nation of nuclear utility property including the safe renoval and disposal of con-taninated debris that could not be decontaninated and/or disposed of on the prenises.
Incurance in generally regarded as the nost reliable nethod to provide the funds to pay for the decontanination and debric renoval required by the occurrence of a nuclear accident. The Three Ille Island accident denonstrated that cleanup fundo can be needed in the approxinate anount of at least about 31 billion.
Ilo evidence ic available to chow that any known naxinun exists with regard to possible need for such funds in the future.
The conclusion, therefore, seens inescapable to thic author that nuclear utilities chould purchase and maintain as much incurance ac nade available by the principal sources to apply to the costs of decontamina-tion and debric renoval as discussed above. Failure on the part of a utility to naximice its ability to protect society in this way is, in this author's view, a groce breach of social responcibility.
I The cane cogency, however, does not apply to the purchase and maintenance by l
nuclear utilities of incurance to repair or replace damaged utility property, whatever the peril /c causing the damage.
Society's principal interest it
'n being l
protected from radioactive contanination, not in a particular nuclear utility being able to rectore itself and its production in the event of accident, whether or not the accident involves radioactive contamination.
ilhether or not a "ith respect to Three Mile Icland, Ginna, or any other nuclear utility plant, society's nain concern is with protection againnt radiation rather than with re-start of the plant.
I
89 particular plant ever resumes operation is not such a compelling social issue as to need treatment by means of incurance regulation. Once no further danger of public contamination existn, the fate of a given damaged plant can be left to the werkings of the market place.
For this. reason, the author feels justified in confining farther attention in this report to insurance that covers decontamination and debris removal This expenses incurred as the result of accidental radioactive contanination.
focus also seems to coincide with NRC's nission, namely, to protect public health and promote safe conditions at nuclear sites.
A potential conflict between the interest of nociety and the interest of owners and long-term creditors of nuclear utilities needs to be anticipated at this point and recolved in favor of society. The insurance under consideration should specify that in case it applies to losses other than decontamination and debris removal expenses it is applicable to cuch other losses only to the extent that it is not needed to pay for decontanination and debris renoval. Giving the decontamination and debrin removal coverage precedence over the other coverages might create a problen for the nuclear utilities.' If co, the utilities might The alteration of the IEIL-Il policy to renovo the precedence given to decontanination coverage, ac nentioned in bection III, C, cuggests that this problen could be serious. Apparently, legal counsel for several of the utilities patronizing UDIL-II advised UEIL officials that the utilities would not be able to The participate in the UEIL-II progran unless the precedence were elininated.
problen was said to arine because of provicions in utility bond indentures that call for the insurance proceeds to be used to repair or replace the damaged Tructees of these indentures property that serves as collateral for the bonds.
might feel conpelled by law or by a provision in the indenture to oppose any pro-Even if vicion in the policy that v:ould given precedence to decontamination.
no precedence in the une of the proceeds were indicated in the utility's insur-ance policy, a truntee might feel obliged to attenpt to preserve the proceeds for une in restoring the prcperty. One option for the trustee in the event of an insured lons might be to try to have the insurance proceeds placed in escrow until needed for rectoration. Thece statenents are not intended to inply that the tructee would be disinterested in decontamination but only that the trustee nicht want the decontanination to be financed from other courecc.
90 be notivated to seek larger limits of coverages so that, after payment of the likely decontanination and debric renoval expences, substantial incurance uould renaln. The remainder could then be used to pay for repair or replacenent of danaged property, thereby saticrying utility creditors and owners. On the natter of prioritien, however, thic author's view in that, clearly, society's interect nuct be cerved firnt.
The problen of catisfying the utility creditors and owners is not the prinary reponsibility of the lac.
If the 100 aere to confine its attention to decontanination and debric renoval incurance, one or nore of the exicting principal cources--or one or nore neu courcec--in due course night offer these coverages in a separate contract.
2hether or not such a developnent occurred night not be a natter of creat nonent for 1r Its intereste would be catisfied so long ac the decontamination and debric renoval incurance uere available ac primary incurance, whether or not in a
+
ceparate policy.
A cace can be nade that nuclear utilitiec in the aggregate have not evinced
)
adequate interect in nuclear property incurance.
Compelling reacons exist to cauce then to be interested in incuring their nuclear properties at leact up to--
)
if not beyond--full replacenent cost neu.
Such incurance limits could 1.elp these utilities protect the invectnent ratings of the bonds and the preferred stock they incue. Thece ratings by Standard and Poor and by roody'r Invector Cervice Inc.
are vital to a utility's ability to raice capital. The cost of the capital for a given utility is extrenely censitive to the invectnent rating its securitien re-ceive from these tuo rating organizationc.
Ibilip Eron, ' ice 1recident of ;itibank,
.A.,
han expressed puzzlenent at the lack of denonctrated interect on the part of utilities toward eliminating s: hat F.r. Fron regards as a ceriouc underincurance problen.
Ee says that invector confidence in the utility inductry--particularly the nuclear cector--in fragile.
Fe indicated that he uac disturbed at the underincurance rink prenlun cone util-j itien are already paying.
Ee urged a tinely solution of General Fublic Utility's financial cricin and a bolstering of property incurance prograns of nuclear utilitiec.
"Looking to the Future fron a Banker'c lercpective," addrecc delivered to Zdicon Zlectric Institute Chief dxecutive Conference, Eancan City (Septenter 11, 19P1).
91 Reasons Why Each Not Already Serving All Nuclear Utilities.
If NML and ANI-MAERP can and should serve all nuclear utilities, this question presents itself:
Uhy are they not already doing so? In defense of both entities the point needs to be registered that each one now, doubtless, is willing to sell coverage to any nuclear utility meeting its underwriting standards, whether or not the utility is also insured in the other entity.
Each policy contains an " apportionment clause" that provides, in effect, that in the event the Lnsured has other appli-cable insurance the insurer will pay no more than its pro rata share of any loss.
This clause prevents an insured from collecting more than the amount of the loss.
Thus, technically, each entity holds itself out as being willing to serve all nuclear utilities.
With no nuclear utility presently being served by both entities, the ques-tion posed above must have dimensions other than the obvious. One explanation of the present state of affairs is that no utility wants more than $450 million of primary insurance and would not buy more even on generous terms. The author, how-ever, cannot help but think that this explan:2on lacks substance. The explanation would be more convincing if a few utilities--or at least one--were now using both The fact that none is doing so suggests that a better explanation lies sources.
elsewhere.
- In fact, ANI-MAERP has publicly offered to do so on several occasions.
- For example, if NU (a nuclear utility) was covered by a $400 million NML policy and a $400 million ANI-MAERP policy, each policy would pay NU one-half of any loss up to $800 million. For simplicity, this example ignores all deductibles.
- To the author's knowledge only one plant site namely, Three Mile Island, has ever been e rved by both.
In 1980 both ANI-MAERP and NML provided full-limit coverage on Unit i, but Unit 1 was not operating. As the author understands, GPU acquired NML insurance on Unit 1 apparently because of some indecision on the part of ANI-MAERP after the accident at Unit 2 to continue full-limit coverage on Unit 1.
(The expectation was that Unit 1, which had been shut down for rou-tine maintenance before the accident at Unit 2, would soon be back in operation.)
The newly acquired NML coverage applied at a time when the ANI-MAERP coverage also applied. Subsequently--and with Unit 1 still out of operation--the NKL coverage was dropped.
92 Another explanation is that the terms of the insurance are not considered by the prospective purchasers as being reasonable, given the fact that substantially tore insurance would be purchased from the two than previously would have been bought from one. This explanation impresses the author as being at the heart of the matter.
It deserves careful elaboration.
A Coinsurance Problem. The foregoing discussion of coinsurance indicated that equity requires that the rate (price per $100 of insurance) go down, other things the same, as the amount of insurance bought approaches the value of the exposure insured. The point was.aade that, even when the coinsurance clause is used in the policy, the rate for insurance based on a 90 percent clause is a bit lower than the rate for insurance based on 80 percent coinsurance and much lower than the rate based, say, on 50 percent coinsurance. The reason, as observed in Sec-tion III, A, is simply that losses tend to be small relative to total value exposed. The probability of a loss of $2X is less than the probability of a loss of $X, whatever the value of X so long as X is positive. For this reason insurers characteristically offer (by means of the coinsurance clause) what amounts to a quantity discount in the unit price of insurance to any insured agreeing to main-tain at least a specified ratio of insurance to value.
This phenomenon suggeststhat, if a nuclear utility purchases $450 million from each of two insurers ($900 million in the aggregate), the rate for each $450 million unit should be lower than the rate that would be proper were the insured to purchase only $450 million in the aggregate.
ANI-MAERP is on record as being willing to reduce its rate in such a circumstance to reflect the larger ratio of insurance to value.
NML is not willing to do so.
With two policies in force in equal amount, each would pay one-half of any loss up to $900 million and less than one-half for any larger loss.
Statement made to author by ANI. In the TMI case as cited above ANI-MAERP set its rate so as to recognize the NML insurance. ANI letter of March 15, 1982 to author.
Statement made to author by NML.
93 An elaboration about how the pricing occurs may clarify the matter. Suppose a nu'elear utility buys insurance on a plant that is worth $1 billion. The maxi-rum primary insurance available from the insurer is only $450 million. An " agreed amount endorsement" is added to the policy. The effect of the endorsement is that
$450 million of insurance will satisfy the coinsurance clause, no matter what the value of the property may prove to be at the time of the loss. The rate the insured pays, however, is determined by the ratio of insurance to value at the time the policy is issued.
In this example the ratio is 45 percent because the value is $1 billion and the insurance is only $450 million. The 45 percent coin-surance rate, as indicated by the table of multiples used by the insurer, is 1.6767 times the 90 percent coinsurance rate.
If the insured were to purchase $450 million of insurance from NML and another
$450 million from ANI-MAERP, a decision would have to be made by each insurer as to what rate to charge. ANI-MAERP has indicated it would be willing to use the 90 percent coinsurance rate or one close to it (according to a statement made by an ANI official to the author). NML would use the 45 percent coinsurance rate, which is about 68 percent higher than the 90 percent coinsurance rate. ANI-MAERP's apparent willingness to discount its rate would give recognition to the fact that insurance was being bought also from NML.
In contrast, NML's apparent unwilling-ness to build a comparable discount into its rate would be to ignore the fact that insurance was being bought also from ANI-MAERP.
If the NML annual premium per reactor for $450 million of insurance at the 45 percent coinsurance percentage rate were $2 million, the amount at issue could be in excess of $800,000 per reactor per year (thatis,$2,000,000less($2,000,000 divided by 1.68) exceeds
$800,000]. The substantial size of the amount at issue perhapa has dissuaded nuclear utilities from using both sources simultaneously under the present system.
- Statement made to author by NML.
94 Speculation about NML's Posture. If this explanation is reliable, the related question still remains as to why NHL has assumed such a posture. The -
author does not profess to know. Two speculations have come to his attention, but neither one impresses him as fully satisfying.
One view is that at least some NML members are fearful that cooperation be-tween NML and ANI-MAERP might lead to antitrust complications. This view appears to the author to be unwarranted for several reasons. First, the change under discussion need not substantially alter the competition that presently exists between NML and ANI-MAERP, except that each would serve the same clients. Most of the advantages alleged for the present competitive system could still apply.
Second, had the insurers been genuinely concerned about antitrust probleme, they likely would have worded their by-laws and policies so as to preclude any possi-bility of cooperative action. Such wording was not included. Third, by virtue of not conducting an insurance business (or any other business) in the United States, NML presumably is immune to antitrust prosecution by United States author-ities.
Fourth, the author sees no basis for apprehension that NML members--as distinct from NML--would become vulnerable to antitrust charges by virtue of mere membership in a mutual insurer that recognized in its pricing the existence of n
insurance as sold by a competitor.
A different explanation of NML's posture--and one more persuasive to the author--is that the utilities represented in NML and NEIL prefer assessment insurance as provided by utility industry captives over advance premium insarance
- ANI-MAERP, which does business in the United States, is on record as being willing to recognize in its own rates the existence of NML insurance. ANI-MAERP counsel sees "no difficulty" to be created by an insured purchasing both ANI-MAERP and NML coverages. Letter of March 15, 1982 to author from ANI.
The author's views are those of a lay person. He purports no expertise in antitrust matters.
- "Or at least enough of them do so to chart the activity of NML and NEIL accordingly.
s 95 c.e provided by conventional insurers. If so, they might prefer to see growth on the part of NML and NEIL rather than on the part of ANI-MAERP. They might view growth of ANI-MAERP as stinting the growth of the NEIL-II program. To this author, such a view, while parochial, could explain the strange state of affairs.
He thinks that society's interest would be served by maximizing the amount of decontamination and debris removal insurance that can be put into place, irre-spective of the principal source. On this basis any strategem that would detract from the amount of such insurance available or that would delay its availability would not be in the public interest.
Recommended NRC Action. Assuming that corrective action by NRC is needed, the question arises as to what, if anything, NRC should do. The author is per-suaded that NRC should bring pressure upon both NML and ANI-MAERP to offer the insurance upon reasonable terms to all nuclear utilities. He is further persuaded that the necessary pressure can be brought to bear indirectly upon NML and ANI-MAERP by placing pressure upon nuclear utilities.
A venerable doctrine in state regulation of insurance is that rates must be adequate, not excessive, and not unfairly discriminatory. The unfair discrimina-tion portion af this threefold criterion has application here. The argument can be advanced that an insurer would be discriminating unfairly if it charged ths same rate to an insured that bought a specified amount of insurance from each of two principal sources as it charged to a comparable insured that bought such amount from only one principal source.
While NRC has substantial direct control over applicants and licensees, it has only indirect control over insurers. Consequently, NRC's ability to correct
- This point is abstruse but can be illustrated as follows:
Suppose NU1 (a nuclear utility) buys $450 million of insurance from NML and another $450 million from ANI-MAERP. Suppose NU2 (another nuclear utility) buys $450 million only from NML. The argument is that NML could be unfairly discriminatory in charging NU1 the same rate as it charges NU2 If another figure (such as $375 million) were substituted for $450 m1211on, the argument could still apply.
96 the situation under discussion is limited in a practical sense to action that it can take with respect to nuclear utilities. With them, however, its options are broad.
The author recommends that NRC change its regulation /s as appropriate to provide in substance the following:
1.
That the issuance and/or continuation of a license to construct and/or operate a nuclear utility reactor is contingent upon the applicant or licensee providing evidence of being an insured in an insurance policy (a) that includes decontamination and debris removal coverage, (b) that stipulates that such coverage takes precedence over all other coverages that may be included in the policy, and (c) that is issued by a prin-cipal source whose rates are not unfairly discriminatory.
2 That failure by a principal source to treat comparable insurance from another principal source as it would treat its own insurance in estab-lishing its own rate is prima facie evidence of unfair discrimination.
3 That primary insurance issued by ANI-MAERP and primary insurance issued by NML are deemed to be comparable for purposes of this recommendation.
h Disinclination of Utilities M Purchase Available Insurance Another problem exists with respect to nuclear property insurance.
Aside from not purt:hasing the primary iffsurance from both primary sources, some utili-ties have refrained in other ways from purchasing all the insurance presently l
available.
Ways in Which the Disinclination Manifests Itself. Aside from the dualism in the primary structure, the disinclination of some utilities to insure to the maximum of available coverage manifests itself in three ways -
1.
As of January 5,1982, six utilities patronizing ANI-MAERP had not pur-chased the maximum amount ($375 million) of insurance that has been 1
l l
t
97 available since about mid-1981.
Five of the six were still at the
$300 million level; the other was close to $375 million. As observed elsewhere, ANI-MAERP raised its limit to $450 million as of January 1, 1982. Whether or not the nuclear utilities will move up to the higher limit with due diligence and dispatch remains to be seen. At the fast-est, the process might take several months.
At less than the fastest it could easily take about a year because some utilities, only recently having renewed their policies, might wait until the next routine renewal to increase the limit. At the slowest, the move up to the higher limits might never be made by some utilities.
2 Similarly, two NML-insured sites as of January 6,1982 were insured for substantially less than the $450 million available from NML since August, 1981.
3 As explained in Section II, C, only thirty-one nuclear utilities had become participants in the NEIL-II program as of early 1982 Not all of the ones that were participating were doing so to the maximum.
NRC's New Regulation. Mindful of the disinclination of some nuclear utili ties to purchase all the nuclear property insurance that is available, NRC in August, 1981 proposed to require such purchase.
The proposed change in the regulation
- Information in letter from ANI to author as cited above.
- Time is required to alter budgets, secure regulatory approval, and comply with other formalities.
- Information in January 6,1982 letter to the author from NML.
- Additionally, not all nuclear utilities participate in NEIL-I and only one participates in APIC.
With attention in this report, towever, now focused on.'y on decontamination and debris removal insurance, these time element coverages are outside this focus.
- Proposed 10 CFR 50 54(v) and 10 CFR 50.55(f) as published in 46 Federal Register No. 159 (August 18,1981), p. 41790. Prior to this regulation, NRC did not require that nuclear property insurance be purchased. As to insurance, the proposal was deemed to be an interim measure, pending a more definitive regulation that would be promulgated later.
98 would have required nuclear utility licensees or construction permit holders within ninety days from the effective date of the regulation to:
have and maintain the maximum available amount of commercial on-site property damage insurance or demonstrate to the sat-isfaction of the commission that it possesses an ec aivalent amount of protection covering such facility.*
Individuals and organizations responding to NRC's invitation for comment on the proposed regulation pointed out the difficulty that nuclear utilities cculd encounter in complying with such a regulation. Questions such as the following were asked by the commentators:
1.
If a nuclear utility could buy more covera 6e in the United States or abroad but only at a demonstrably exorbitant rate, would it neverthe-less be obliged to do so?
2 How would one know whether or not one had the " maximum avai?able amount" of insurance?
3.
Does the regulation imply that each nuclear utility would have to pur-chase insurance from both ANI-MAERP and NML, regardless of cost?
4 If not and if one principal source (ANI-MAERP or NML) offered a few dol-lars more of coverage than the other, would each nuclear utility insured by the other insurer have to switch to avoid violation of the regulation?
5 Would the insurance still be "available" even though the insurer refused to sell it to a given nuclear utility except on terms regarded by that utility as unreasonable?
These comments led the NRC staff to suggest to the Commission several alter-ations in the proposed interim regulation. These suggested alterations were designed to overcome most of the difficulties identified by the commentators referred to above. On March 11, 1982 the Commission approved the regulation in Ibid.
99 a form close to the altered form suggested by the staff.
Upon becoming effec-tive, the regulation will require nuclear utilities to purchase the maximum pri-mary insurance available on reasonable tenns from either ANI-MAERP or NML plus the maximum excess insurance available on reasonable terms from the one principal The regulation urges but does not require that the incurance include source.
coverage for reasonable decontamination and cleanup costs.
The author regards this regulation as a stopgap measure. He offers the fol-lowing wproach as one that would best serve the long-range public interest with respect to nuclear property insurance. This approach seems consistent with--
but goes substantially beyond--the interim regulation.
Suggested Alternative Approach for NRC. Because of the virtual impossibility of articulating an unqualified but meaningful blanket rule, NRC should abandon its effort to do so.
It could, instead, approach the problem by allowing for indi-vidual differences among nuclear utilities and by recognizing that insurers have to retain the option to refuse or restrict coverage where their underwriting standards are not met.
Under this approach NRC could require that each nuclear utility purchase and maintain all the decontamination and debris removal insurance that is available to that utility from each principal source on reasonable terms.
Obviously, NRC would have to specify how differences between the insurer and the utility as to whether the proposed terms (including rates) were reasonable would be resolved in the given case.
In the author's view, NRC has neither the necessary expertise nor authority to resolve such disputes. A more cxpeditious procedure would be for NRC further 10 CFR 50 54(w).
..As explained elsewhere in this report, NBC need not concern itself with whether or not the insurance covers property damage or includes perils other than accidental radioactive contamination. NRC should be vitally concerned that the insurance will pay for decontaninating radioactive property on the site and for removal and aisposal of contaminated debris.
100 to require that, as a prerequisite to having its license issued or continued, the applicantorlicensee(thatis,theutility) agree in case of dispute about reasonableness of the terms of the proposed insurance t.o submit to arbitration the question of whether or not the terms are reasonable. The amount of insurance 1
the utility ultimately would be alle and willing to purchase, once known by NRC, could then be used by NRC, along with all the other information that NRC uses, to determine whether or not to issue (or continue) the license.
NRC could also require that the insurer, as a prerequisite for gaining NRC's acceptance of any insurance it might provide the applicant or the licensee. -agree to arbitrate the reasonableness of the terms. Tne insurer likely would be willing to enter this arbitration for the following reason.
If the terms were deemed reasonable by the arbitrators, the utility would be obliged to buy the insurance under such terms.
If the terms were deemed to be unreasonable, the utility would have no duty to accept them and the insurer would have no obligation to change them. The insurer might decide voluntarily to change the terms. The purpose of the arbitration, however, would be merely to determine whether the terms were reasonable. The purpose would not be to force the insurer to change the terms of the insurance, should arbitration show the terms to be unreasonable.
l l
Interestingly, the ANI-MAERP, NML, and NEIL policies already contain an arbi-tration (or appraisal) provision for use at the option of either party to the contract in case of a dispute about the amount of a claim. The ANI-MAERP appraisal clause, taken with minor modifications from the standard fire policy, reads as l
follows:
In case the Insured and Insurer shall fail to agree as to the actual cash value or the amount of the loss, then, on the written demand of either, each shall select a competent and disinterested appraiser and' notify the other of the appraiser selected within twenty (20) days of such demand. The appraisers shall first select a competent and disinterested umpires and failing for fifteen (15)
I days to agree upon such an umpire, then, on request of the Insured or the Insurer, such umpire shall be selected by a judge of a court l
t l
1
101 of record in the state in which the property covered is located.
The appraisers shall then appraise the loss, stating separately actual cash value and loss to each items and, failing to agree, shall submit their differences, only, to the umpire. An award in writing, so itemized, of any two when filed with the Insurers shall detennine the amount of actual cash value and loss. Each appraiser shall be paid by the party selecting him and the expenses of the appraisal and umpire shall be paid by the parties equally.
Although this procedure has to do only with disputes about the amount of an insured loss, it could easily be modified to fit a dispute about whether terms of a proposed insurance transaction were or were not reasonable.
Ona useful circumstance is that many persons in the insurance community are familiar with this procedure. The clause appears in numerous policies and the process has besn invoked many times. Use of this process would obviate the necessity for NRC to take upon itself any adjudicatory role in disputes about the reasonable-ness of the terms of proposed nuclear property insurance transactions. Yet, use of the process would allow NRC to perform its duty to maximize the volume of de-contamination and debris removal insurance that can be issued on reasonable terms.
D.
Lack of Statistical Stability in Annual Operating Data Another problem embodied in nuclear property insurance is the lack of sta-bility in the annual operating data of the insurers. The annual underwriting gain or loss figures are particularly lacking in actuarial reliability. What may appear to be a handsome cumulative underwriting gain may prove later to have been only a mirage. The number of reactor units exposed to loss is too small to allow the law of large numbers to stabilize the predictions. A large loss, such as at Three Mile Island, may consume all the cumulative underwriting as well as invest-ment gain and even may require the insurers to invade capital in order to pay
- Of interest is the point that the aim of the arbitration would be merely to establish whether or not all the terms collectively were reasonable.
If they were, the nuclear utility would be obliged to accept them.
If they were not, it would not be so obliged. The arbitrarion would not involve any restatement of terms held to be unreasonable.
102 the resulting claims. Worse yet, no guarantee exists that Three Mile Island is the largest nuclear property insurance loss that can be suffered.
Artificiality of Annual Accounting. The artificiality of annual underwric-ing gain or loss figures is troublesome for all principal sources of nuclear property insurance.
It complicates ratemaking and management of cash flow. The problem is more severe, however, for ANI-MAERP than it is for NML or NEIL. The reason is twofold. First, NML and NEIL, relying as they do on using assessments if and as the need arises, do not depend as heavily as do the insurers in ANI-MAERP on advance premiums to pay for insured losses. Consequently, year-to-year volatility does not pose as large a problem in establishing advance premiums for NML and NEIL as it does for ANI-MAERP.
The second reason that year-to-year -
dility is more troublesome for ANI-MAERP than for NML or NEIL bears on income taxation. Being domiciled in Bermuda.
NML and NEIL are subject neither to the United States federal income taxes nor to any substantial income tax in Bermuda. An underwriting gain in a given year for NML or NEIL does not create for that year any federal income tax liability. By contrast, most of the insurers ani reinsurers participating in ANI-MAERP are sub-ject to United States federal income taxes. Annual underwriting gain or loss as reported by ANI or MAERP influences the tay ble income of these companies. The l
l annual underwriting gains consistently rept rted by ANI-MAERP from 1957 through the middle 1970's led to increases in the taxable incomes of the participating companies.
Assuming that each company each year was in a position of underwriting profit, l
a part of the underwriting gain derived each year from ANI-MAERP had to be paid in income taxes. As matters turned out, the figures reported during these years
- The reason is that advance premiups do not constitute as large a portion of
(
total income for NML and NEIL as for ANI-MAERP. By the same logic, volatility i
creates a potentially horrendous problem for NML and NEIL--but not for ANI-MAERP with re6ard to assessments. To date NML has not suffered any serious loss: neither has the NEIL-II program. A sizeable loss has occurred in NEIL-I.
l
103 as annual underwriting gains were not real income after all because of the large less at Three Mile Island. Admittedly, the participating companies were allowed to offset their shares of the ANI-MAERP underwriting losses in recent years against what otherwise would have been their taxable income. The offset, however, was not necessarily complete. Moreover, the companies lost the income that could have been earned on the money paid as taxes on " income" that turned out not really to be income. As a result, the insurcrs could not accumulate as much money to pay claims as they could have done in the absence of the requirement that they report each year's underwriting figure and pay a tax on any annual gain.
In this author's judgment, some sort of leveling arrangment for federal in-come tax purposes is needed in ANI-MAERP to recognize the lack of stability in the annual underwriting gain or loss data. The participating insurers and re-insurers need a mechanism by means of which to average the annual underwriting results over a decade or longer period. A ten-year or longer moving average annual underwriting gain or loss figure might be much more meaningful than any-one of the annual figures used to compute the moving average.
Industry Credit Rating Plan. The author thinks that an " Industry Credit Rating Plan," similar to but perhaps simpler than the one in use for the nuclear liability insurance program, would be in the public interest. Under the liability insurance plan ANI and MAELU are pemitted special tax treatment by the Internal Revenue Service under a private letter ruling.
An amount equal to about 70 percent
- Underwriting gain or loss is the difference between earned premiums and the sum of incurred losses, loss adjustment expenses, and underwriting expenses. The incurred losses figure for a given year includes estimates of amounts that will become payable because of insured contingencies that occurred during that year.
For purposes of arriving at taxable income, these estimates have to be confined to contingencies that have already occurred. Except as to life insurance and certain long-term disability insurance contracts, insurers normally are not allowed to include in the incurred losses for any year estimates of the cost of contingencies that may occur in subsequent years. Given this constraint, nuclear property insurers cannot for tax purposes charge against current income any allowance for insured contingencies that have not yet occurred.
In arriving at their taxaole income from year to year the insurers are not allowed to anticipate the occurrence in a future year of an accident such as the one that occurred at Three Mile Island.
104 of each year's premiums is placed in a separate account from which incurred losses and loss adjustment expenses arededucted. Any amount left over a the end of the year is considered as " unearned premium." As such, it is not income to the participating insurers but is a current liability of ANI or MAELU and is subject to being refunded ultimately to the insured utility if not needed for losses or loss adjustment expenses.
Any unearned premium that remains in the account for ten years, not having been needed for losses or loss adjustment expenses during any year in the ten-year period ending with the current year, is refunded to the insureds.
If ANI-MAELU should have bad loss experience over a given ten-year period, little or no unearned premiums would remain at the end of that period to be refunded.
With good loss experience, however, the refunds could be large, as, indeed, they have been.
Under the nuclear liability insurance program up to 1980, roughly 55 per-n cent of the premiums paid to ANI-MAELU through 1971 have been refunded.
i The plan described above has both advantages and disadvantages to the parti-cipating insurers.
It is advantageous to them in that it obviates their being required to pay federal income taxes on " income" that may prove to have been illusory.
It is disadvantageous in providing that any " unearned premium" placed in the special account and not used over a ten-year period must be returned to the insureds.
Recommended NRC Action. To this author's way of thinking, the ANI-MAELU plan l
l could serve a useful purpose by being extended to nuclear property insurance.
l Alternatively, a simpler system might be equally useful.
If the objective is merely to reduce the year-to-year fluctuations in insurer taxable income, the i
+ In fact, if the loss ratio for the period should exceed 70 percent, the special account would not be adequate to absorb the incurred losses and loss adjustment expenses.
n American Nuclear Insurers Reports, as cited earlier, p. 4
(
.--,.,.J-
,,_m.
_m.
7..
.-- - - - - - - _ - + -
105 dsvice to achieve the objective could be less elaborate than the ANI-MAELU plan.
For purposes of determining federal income tax each year for each insurer in ANI-MAERP some type of average annual undorwriting gain or loss could be substituted for the actual gain or loss the insurer derived from ANI-MAFRP for that year.
For example, the figure to be substituted could be the adjusted average of the annual underwriting gain or loss figures for the last twenty years.
The average would need to be adjusted to account for growth or decline in premium volume over the twenty-year period.
This approach would not require the establishment of any extra unearned pre-mium reserve.
It would not impose any restrictions on use by ANI-MAERP of pre-mium income.
It would not involve the insureds in any direct manner. Over the long run it would not necessarily change the totality of taxes payable. For any twenty-year period, however, the total taxable
.come could be larger or smaller than under the present system, depending on the size and timing of the gains and losses.
Its sole purpose would be to reduce for tax purposes the variability of annual underwriting gain or loss more so than could be accomplished by the oper-ating loss ca:Ty-backs and carry-forwards available to the insurers under the present system of income taxation.
The author recommends, therefore, that NRC endorse the need to define for federal income tax purposes the annual underwriting gain or loss of insurers and reinsurers participating in ANI-MAERP in such a way as to reduce the year-to-year volatility of such gain or loss. The author is aware that distinct limitations exist on the ability of NRC to influence federal income taxation in the United States. Because this report is addressed to the NRC, however, this recommendation
- Use of a twenty-year average would be consistent with the Nuclear Property Insurance Industry Experience Guide, as described elsewhere in this report.
- One way to make this adjustment would be to compute a weighted average of underwriting gain or loss with the weights in proportion to annual premium volumes.
106 is also addressed to the NRC.
If the recommendation has merit, nothing prevents its also being acted upon by others, including the companies that would be affected by it.
E Prohibitions on Purchase of Mutual Insurance A peculiar provision exists in the constitution of the state of Texas. This provision would create a problem in implementing the recommendation that each nuclear utility buy all the insurance reasonably available from each principal source.
The Problem. An excerpt from a letter to NRC from an attorney for a Texas municipality spells out the problem:
Section 52 of Article III of the Texas Constitution denies a Texas city the power "to lend its credit or to grant public money or thing of value in aid of, or to, any individual, association or j
corporation whatsoever, or to become a stockholder in such corpora-tion, association, or compan;'." The Texas Supreme Court has con-strued this provision as prohibiting a Texas city from purchasing insurance which provides for assessment of policyholders or from purchasing insurance from mutual insurance associations where the policyholder becomes the equivalent of a stockholder in the company. City of Tyler v. Texas Employers Insurance Association, 288 S.W. 409 (Tex. Comm'n App. 1926, judgmt. adopted); reh'g denied 294 S.W. 195 (Tex. Comm'n App. 1927).
Lewis v.
Independent School District of Austin, 139 Tex. 83, 161 S.W. 2d 530 (1942).*
The author of the letter proceeded to argue that Texas cities should not be l
required by NRC to purchase insurance coverage that they are prohibited ')y Texas i
law from purchasing.
He argued further that, with the cities not being allowed legally to purchase insurance from mutual insurers, the cities should not be required to provide the " equivalent amount of protection" referred to in the proposed URC regulation.
i l
+
Letter of October 14, 1981 to NRC from an assistant city attorney for i
Austin, Texas. The author of the letter was commenting on NRC's proposed regu-lation 10 CFR 50. Similar letters were addressed to the NRC on behalf of the City Public Service Board of San Antonio and Houston Lighting and Power Company.
The author of the first letter cited in this footnote observed that the Supreme Courts of Louisiana and Idaho had reached decisions on the matter similar to those of the Texas Supreme Court.
- - - - - - - - - - - - - + - - - - - - - - - --
107 The prohibition cited in the excerpt from the letter is troublesome on three counts. First, it seems to preclude municipally owned nuclear utilities from buying insurance from NML. Second, it also seems to preclude them from buying either NEIL-I or NEIL-II insurance. Third, it even raises questions about their buying insurance with which MAERP is identified. With insurance to cover nuclear utility decontamination and debris removal being in short supply, this prohibition is serious. The public interest weighs arainst Texas municipal nuclear utilities being excluded from a major portion of the market.
Recommended NRC Action.
The long-range solution to this problem seems to be the promulgation of an NRC regulation that would have the effect of preempting any state or local law that would be in conflict with the regulation. Whether or not NRC has the authority to issue a regulation preempting state and local law is a question beyond the expertise of this author.
He recommends that NRC seek appropriate legal counsel about the issuance of a regulation to legitimize the patronage by nunicipal and other publicly owned nuclear utilities of MAERP, NML, NEIL, and other mutual-oriented suppliers of nuclear property insurance.
Pending the effective date of such regulation /s, other expedients need to be considered. One such expedient could be the use of non-mutual insurers to front for the mutual insurers in any case where a prohibition of the sort under dis-cussion here is encountered.
NML and NEIL could organize and use stock company subsidiaries or could eqpage other stock company insurers to issue coverage to municipally owned nuclear utilities and possibly other utilities in Texas.
- 10 CFR 50 54(w), as cited earlier, seems to dodge this issue by merely requir-ing that a utility subject to such a prohibition need buy only the insurance reason-ably available to that utility. One can argue that if the rtility is prohibited from buying the insurance, such insurance is not available to that utility.
The author understands that the fronting technique is already being used to avoid probleus in the purchase of ANI-MAERP insurance by Texas municipalities that have ownership interests in nucleer reactors.
108 The author does not view the fronting device as the perfect solution to the problem.
Its use smacks of subterfuge. One should always be suspicious about the use of a substitute to alter the form but not the substance of a transaction, even when the substitute is needed to satisfy a bothersome legal technicality.
Here as in other parts of this report, however, the public good is deemed to be served best through maximum availability of decontamination and debris removal insurance. Such a rationale convinces the author that fronting in this situation is consistent with the public interest.
F.
Avoidance of Accidents Doubtless, the most serious threat to the continued availability of nuclear property insurance is the threat of fbture accidents.
If the frequency and/or severity of future accidents at nuclear plants is widely perceived as substantial and growing, no amount of activity by the NRC would necessarily succeed in pre-serving the demand for and supply of nuclear property insurance.
Fragility of the Insurance Mechanism.
Insurance theory makes clear the fact that insurance is a fragile mechanism. The range of insurability is narrow.
The mechanism can function only when the probability of loss during any given year is I
low. Otherwise, the premiums necessary to pay the expected losses are larger than the insuring public is willing to pay.
With the demand for the insurance eroded, insurers tend to reduce supply because of the absence of expectation that premiums would be adequate to pay the losses and expenses.
The exact identification of the upper limit of insurability as measured by the probability of a total loss during a given year is open to dispute. The higher the catastrophic potential, the lower the upper limit of insurability as measured
- Adverse selection begins to manifest itself with the insureds who are less likely to produce claims dropping out first. Their departure from the pool of l
insurers simply worsens an already bad situation, l
0 109 by probability. The author sunaises that the upper limit of insurability of nuclear property insurance is very low--perhaps as low as 0.2 percent.
If this surmise is indicative of reality, the voluntary supply of nuclear property insur-ance in the private sector is vulnerable to even small increases in loss frequency or severity.
NRC's Critical Function. The author suggests that the most effective con-tribution NRC can make to the quantity and reliability of nuclear property insur-ance is to promote safety in the operation of nuclear reactors. NRC's proper function with respect to nuclear property insurance, therefore, is consistent with one of NRC's principal roles generally, namely, protection of the public against radioactive contamination.
VIII.
SUMMARY
OF RECOMMENDATIONS For clarity and emphasis the author's recommendations to NRC are restated in sammary form as follows:
A.
That NRC not go beyond the types of insurance already in existence in accepting assessable insurance as satisfying any nuclear property insur-ance requirements it may impose.
B.
That NRC change its regulation /s as appropriate to provide in substance the following:
1.
Thattheissuanceand/orcontinuationofalicensetoconstructand/
or operate a nuclear utility reactor is contingent upon the applicant or licensee providing evidence of being an insured in an insurance policy (a) that includes decontamination and debris removal cover-age, (b) that stipulates that such coverage takes precedence over all other coverages in the policy, and (c) that is issued by a principal source whose rates are not unfairly discriminatory.
l 110 j
2 That failure by a principal sourt:e to treat comparable insurance l
from another principal source as it would treat its own insurance in establishing its own rate is prima facie evidence of unfair discrimination.
3 That primary insurance issued by ANI-MAERP and primary insurance issued by NML are deemed to be comparable for purposes of this recommendation.
C.
That NRC in due course promulgate a permanent regulation to require:
1.
That each applicant or licensee as a prerequisite to hsving its license issued or continued purchase and maintain all the decon-tamination and debris removal insurance that is available to that applicant or licensee from each principal source on reasonable terms.
2 That each applicant or licensee as a prerequisite to having its license issued or continued agree in case of dispute between it and the insurer about reasonableness of the terms of the proposed insur-ance to submit to arbitration the question of whether or not the terms are reasonable.
3 That 3ach insurer in each principal source as a prerequisite for gaining NRC's acceptance of any insurance it might provide the applicant or licensee agree in case of dispute between it and the applicant or licensee about reasonableness of the termr. af the pro-posed insurance to submit to arbitration the question of whether or not the terms are reasonable (but without any obligation to alter anyterm/sdeemedunreasonable).
D.
Th-t NRC endorse the need to define for federal income tax purposes the annual underwriting gain or loss of insurers and reinsurers participating
111 in ANI-MAERP in such a way as to reduce the year-to-year volatility of such gain or loss.
E.
That NRC seek appropriate legal counsel about the issuance of a regula-tion to legitimize, where not. prohibited by state or local law, the patronage by municipal and other publicly owned nuclear utilities of MAERP, NML, NEIL, and other mutual-oriented suppliers of nuclear property insurance, and that, pending the effective date of such regula-tion, NRC accept as in the public interest the practice of a stock incurer fronting for a mutual insurer in complying witt. the prohibitions.
F.
That NRC contribute to the development of nuclear property insurance by continuing to protect the public against radioactive centamination.
l
I 112 M
==
B IB LIOGRAlmY
_- m C he rn ic k, I aul L., W illiam B. Fairley, Michael B. Meyer, and Linda fl. Scharff.
- esign, ':osts, and Acceptability of an Electric Utility Self-Insurance Fool for Assuming Adequacy of Wnds for Nuclear l ower Ilant Decommission Expense.
U.".
I Nuclear "erulatory :ommission NUFFT,/N2370 (1981).
Available for purchase E
from the National Technical Information Service, Springfield, VA, 22161.
M Lleaning Up the Hemains of Nuclear Facilities--A Multi-billion Dollar Problem.
Deport to the Consrress by the Comptroller General of the United States, EMD 9 46, 1977. Available for purchase from National Technical Information Service.
[
- nri ng f i e ld, VA, 22161.
Cleanup of Three Mile Island Unit No. 2 Task Force on Nuclear Institutional Issues
_==
to Mdison :lec t ric Insti tute Board of Direc tors, 1981. Availability unknown by autho,.
Ienc ri pt ion of Mxcess Iroperty Insurance Program.
Nuclear Electric Insurance I1mited, 19'i.
Availabi11ty unknown by author.
- rea t e r
- ommitment
- .eeded to.lolve :ontinuing I roblems at Three Mile Is land.
Staff of
lone ra l Accounting O f f ic e, 19P1, Chapter 6.
Available for purchase frem the National ec hn ic a l In fo ma t ion Service, Springfield, VA, 22161.
m uurley, 9 1.
" Di sc unsion of an Actuarial Note on Experience Pating Nuclear Irop-erty Innurant e' J'roceedines of the <:asualty Actuarial Society, Vol. 60 (May,
10"O, pp. thE-111.
Available from the :asualty Actuarial Society, 200 Fast C nd
'treet, New York, NY, 10014 Innc, 'ohn D.
"' oft pots in Insurance cheory" in John D.
Long, Ed. Is sues in
- ngurance.
'nd ed.,
Ma l v e"n, 1 A:
American Institute for Property and Liability j
'nde rwr i t e rs, :nc 1991, Vol.
pp. 3M -432 Available for Purc hase from
=
Ameri an :nstitute, I rovidence and Surartown Roads, Malvern, IA, 193 '.
Mc'!ure, J.
"An Actuarial Note on Experience Hating Nuclear Iroperty Insurance."
l'roceedines of the <:asualty Actuarial,loriety, Vol. 69 (November, 1972), pn. 160-l
Available from :asualty Ac t uarial :;oc iety, 200 Fast 42nd ?treet, New York, M
NY, 1C 'lu.
"Ni< lear :n:urance
'ac t e and 2icures' American Nuclear :nsurers Reports, No. 1, evises
,v arch, lou Available upon request from Americar. Nuclear Insurers, "uildine 1,
h e Mxc"ance, Fa m ! nc t o n,
'T, Me l..
u6 ele"e',
sonal i W.
financial :mplications of Het"ospective I remium Assessments on v
< l ec + r i,
't il! + 1e!
Nuclear "erulatory lomnission, NU - A IG-00 3 ( 19761 M
Available cor m rehase crom the National Tec hnical Information
'e rvic e, ' p r in r'-
Cie'1, VA,
' i + 1.
M Nuclear : owe *, ' a f e t ', 1 : ns uran c e--
- s sue s of the 19"C's--The I ns u ranc e Indus t ry's Viewpoin+.
',11ance of American :nsurers, Ame ric an Insu ranc e Association, Y
Na+ tonal AsSoetTtton of Independent :nsurers, Mutual Atomic Energy Reinsurance M
1001, Anertean Nuclea: Insurers, no date.
Available upon request from American d
Nuclear nsure- ;, Fullding 3, The {xc hance, Farmington,
'T, 06032 w
C
.A
113 Summary of Operatioris. Nuclear Mutual Limited,1981. Available upon request from Nuclear Mutual Limited, P.O. Box 2025. Hamilton 5, Bermuda.
=
O.S. NUCLE AR REGUL ATORY COMMISSION BIBLIOGRAPHIC DATA SHEET NUR EG-0891
- 2. (Leave blank) 4 TITLE AND SUBTtTLE (Add Volume No. of wormonete) nuclear Property Insurance: Status and Outlook 3 RECIPIENT S Access oN NO.
5.
ATE REPORT COMPLETED EuNoHtsi M ON TH l YEAR Dr. John D. Long, Consultant April 1982 DATE REPOHT f 5SUEO 9 PE HF ORMING ORGANIZATION N AME AND M AILING ADDRESS (tactude lip Codel "ONT" l "^a Of fice of State Programs May 1982 U.S. Nuclear Regulatory Commission 6
'L'*' **a*>
Washington, D.C.
20555 8 (Leave blanki
- 12. SPONSORING ORG ANIZ ATioN N AME AND M AILING ADDRESS (include 2,0 Codel PRoJE CT/ TASK! WORK UNIT No.
Of fice of State Programs U.S. Nuclear Regulatory Commission
" "'N" d. " '
m Washington, D.C.
20555 M
PE RIOD COV E RE D (inclusere datesj
- 3. T YPE OF RE POR T Technical Information To present M
15 SUPPLEME N TARY NOTE S 16 ARST R ACT 000 **wds or len/ The report addresses the problem of the unavailablity of adequate levels of preperty insurance for commercial power reactors to pay for decontamination and cleanup costs aris'.ng from accidents. The report is designed to answer six questions, as follows:
=
l.
What has been the development of each principal source of nuclear property insurance used as of early 1932 by nucl~ utilities in the United States?
2.
What are some of the distinguishing features of nuclear property insurance as offered by the pr ncipal sources ?
i 3.
How much nuclear property insurance war offered by each of these sources as of January 1, 19C ?
Assuming that present plans came to fruition, how much nuclear property insurance is y
4 likely to be of fered by each of these sources as of January 1,1983?
5.
What, if any, principal sources of nuclear property insurance are likely to emerge in the private sector by January 1,19R3?
f). What problems serious enough to warrant action of the NRC exist with respect to nuclear
=
property insurance and what action should NRC take in response to each problem?
17a DE SC RIP T O H S 17 Q y WORDS AND DOCUME N T AN ALYSIS Property Insurance M
Decontamination Nuclear Industry Financing 17h IDE N 1lF it RS OPE N E N DE D 1E RVS
" s' c ' * " c ' ^ ss " ""
- 2
' " ^ " ' '
iR Av An ARiun ST ATm NT unel m ifind 22 es u Unlimited 20 su n yss in s ov, unclassifled S
sac roRu 33s a m M
-B I
UNITED STATES
, ou,,, ct,ss,,,
i NUCLEAR REGULATORY COMMISSION acstact a rtis raeo WASHINGTON. D.C 20%5
.$s'.*c' c l
- t avet to 11[_, _
06 F O At BUSINE SS Pt P.AL T Y f 0A FRN A f f USE $3X) 2 5 5 5 0 7 9,3 7 I AfuNhC 1 ANISID CF TEDC PGLicy S l>;agyacc"h"Mc4ncNs ocr e, hASHIAGICN DC 23555