a decommissioning trust suffers the same vulnerabilities. That suggests a need for termination insurance.With the growing prospect of some form of electric utility deregulation, and the possible end of rate regulation (em with no guaranteed source of income for utilities that hold licenses to operate nuclear reactors (em it's high time to reexamine the funding process for nuclear decommissioning and consider a policy alternative to ensure that funds will be available when needed to retire the nation's nuclear plants.
Investor-owned utilities with nuclear power plants are accumulating billions of dollars through the ratemaking process to fund Nuclear Decommissioning Trusts (NDTs). The U.S. Nuclear Regulatory Commission (NRC) mandates NDTs as an alternative to the two other decommissioning assurance options (em i.e., use of prepayment or surety bond.
When a nuclear plant is permanently shut down, several alternatives are available to handle decontamination, dismantling, and site restoration. But no matter what alternative is chosen, the responsibility to provide funds for future decommissioning currently falls to the nuclear plant owner (em a liability accepted as a condition of holding an operating license.
However, because the time horizon for actually settling this obligation can span many years, several public policy concerns remain. A 1993 article about the economics of decommissioning, illustrated the application of Coase's Theorem on externalities to the thorny question of how to decide, from a public policy standpoint, whether to have prompt decommissioning (DECON) or delayed decommissioning with interim safe storage (SAFSTOR).1 Among other important considerations, the article revealed the impact of the assignment of property rights and the choice of the discount rate (used to calculate the present value of the liability).
Now, a new generation of concerns that affect NDTs has emerged:
s State initiatives on electric restructuring, in California and elsewhere
s The February 7, 1996, Exposure Draft issued by the Financial Accounting Standards Board (FASB) on a proposed accounting standard for the closure and removal of long-lived assets, including nuclear plants2
s The bankruptcy of Orange County, CA (and the disclosures of massive losses by other organizations) (em due in part to so-called "derivatives" transactions
s The Notice of Proposed Rulemaking issued by the NRC in April, 1996, to examine financial assurance requirements for nuclear decommissioning.3
Any of the first three developments might warrant a look at NDTs. The NRC action forces the issue. The current climate may even suggest a more fundamental rethinking than the NRC contemplates.
Contemporary arrangements for trust-investment management seek a target rate of return. If the portfolio is risky, it can easily have a future value far greater than the plant owner's initial investment (em or far less. If far greater, then a small initial investment will suffice to pay off the future decommissioning cost. But, if the value turns out to be far less at that distant time, the plant owner may no longer be around to make up the shortfall.
The efforts now underway for electric industry restructuring and deregulation call into question implicit assumptions that underpin the establishment of the present NDT structures. The possibility of a shortfall in NDT funds is, thus, both highlighted and exacerbated.
This climate suggests the idea of termination insurance to cover any shortfall in NDTs. Termination insurance would pay a contracted amount to a nuclear plant owner/ operator (or other beneficiary) in the event of a premature shutdown, given stipulated terms and conditions of operation and shutdown. [Some limited coverage is already available from property damage insurance to apply to any NDT shortfall in the case of shutdown due to an accident.] At the least, any reexamination of trust management should deal with the notion that the plant owner in effect holds a put option to defray its decommissioning liability, and the fact that the cost of such insurance can be estimated using option-pricing theory.
The Present Approach and its Problems
The present approach calls for a dollar estimate of future nuclear decommissioning costs (the "certification amount"), in accordance with NRC regulations.4 This dollar amount forms the basis for periodic accruals under ratemaking to fund the NDT, in the form of an external sinking fund. Revenues collected to fund the trust are currently deductible for income tax purposes, provided that the funds go into a so-called Qualified Trust set up in accordance with section 468A of the Internal Revenue Code.
This procedure invites an analogy to the operation and actuarial basis of a defined-benefit pension plan. Since the expected payout dates in most cases extend many years in the future, a gap usually exists between the amount of actual assets on hand, and the ultimate liability amount. The current market value of the assets plus the expected income and expected future accruals equals the present value of the liability, discounted at an appropriate rate.
In practice, the approach to nuclear decommissioning tends to be bifurcated: An engineering effort focuses on estimating and minimizing the liability in cost terms. Separately, a financial activity related to the trust fund(s) focuses on maximizing the investment returns to the NDT. Now that IRS regulations no longer restrict qualified NDTs to "Black Lung" trust investments,5 state public utility commissions (PUCs) have typically endorsed investing a high percentage of the funds in common stocks, especially for NDTs contemplating funds
accumulation and investment over many years.
The presumption is that stocks represent prudent long-term holdings because they offer a higher expected rate of return than bonds, and the probability of realizing the expected rate of return increases with the length of the holding period. Yet, because returns on stocks generally prove more volatile than returns on high-grade fixed-income securities, NDT managers presumably would shift funds judiciously from stocks to fixed-income assets as decommissioning becomes imminent, so that the money needed to pay for decommissioning can actually be obtained by liquidating noncash assets in the NDT at the time required.
This bifurcated approach presents problems. First, it is unnecessarily risky, because no advantage is taken of the possibility to better match the portfolio asset structure to the characteristics of the decommissioning liability. Furthermore, although approved NDT contributions are tax deductible, the trust income, unlike pension trust income, is taxable to the (utility) owner-operator (unless invested in tax-exempt securities like municipal bonds). The NDT financial position in the long run is thus highly dependent on, among other things, the tax laws, the asset allocation over time, and the market-timing skills of the investment managers.
Second, the potential arises for a shortfall of assets (a gap) should the plant shut down prematurely without termination gap insurance. It does not yet appear that utilities formally integrate their nuclear plant insurance approaches (handled by the "risk management" department) with their plans for decommissioning.
Third, a plant owner may wish to "abandon" the plant for business and/or tax purposes. This option becomes conceivable now with electric industry restructuring and the revaluation of generating assets, that is now under way.
Estimates of the expected cost to decommission using the DECON prompt dismantling alternative currently run to about $300 million in 1996 dollars. Putting the plant in SAFSTOR would entail, alternatively, ongoing surveillance, monitoring, and maintenance expenditures equivalent (in present value) to at least as large a fund.6 If only a fraction of the required sum has been accumulated in the NDT at the time of shutdown, the actions that could be taken by the parties will depend on what terms and conditions are then operative under regulatory or contractual arrangements in force to minimize the adverse consequences to public health and safety.
Changing the Approach
and its Consequences
Modern financial theory, pension fund management, and insurance/actuarial practice suggest that contingent claims analysis and an Asset-Liability (A-L) management approach could lead to superior results from a public policy standpoint. Such an approach entails A-L matching while maintaining an optimal portfolio. What is termed "the gap" (the asset shortfall) is the mathematical analog to the
concept of "negative surplus" in the model of a defined-benefit pension plan. If the modeling of the "liabilities" (stream of costs over time) is integrated with that of the "assets" (NDT), the gap can be better monitored and controlled. Techniques for option pricing and portfolio optimization should allow one to design an NDT portfolio that minimizes or eliminates shortfall risk.
The riskier the assets, or the longer the time until decommissioning, the larger the potential swing in the future value of the portfolio. It is this latter result that proves crucial in the context of the NDT. To the extent that funds are required at a date certain, the holding of stocks and bonds rather than a dedicated and immunized portfolio (including, e.g., AAA-rated, zero-coupon, inflation-indexed bonds) could lead to a shortfall on the date(s) certain for disbursement from the NDT.
By analogy to recent work by Zvi Bodie7 and earlier work by Jack Treynor8 that focused on corporate defined-benefit pension funds, one can show that the assurance provided by the nuclear plant owner to cover decommissioning costs involves a put option.9 The analogy to pension funds also calls attention to the concept of termination insurance, as provided by the Pension Benefit Guaranty Corporation (PBGC) under the ERISA legislation.10
Put options fall under the rubric of "derivatives" (em financial instruments that can be used in hedging or risk-minimizing transactions. However, as the recent, well-publicized experience of Orange County (as well as several large corporations) shows, derivatives can be structured and used in transactions that embody speculative bets on specific financial outcomes (specifically, betting on changes in interest rates) and a high degree of reliance on settlement (i.e., no default) by the counterparty.
The put option in the NDT case captures the potential for the plant owner to effectively settle the decommissioning obligation if the value of investments in the NDT at the time of decommissioning turns out to be less than the expenditures required at that time, and if the plant owner has no recourse to sufficient additional funds either through emergency ratemaking or proceeds from gap insurance. The value of a put option can never be negative, nor less than the difference, in this case, between the cost of decommissioning (the "exercise price") and the value of the NDT portfolio (and any other assets available) for settlement of the decommissioning obligation. Of course, even if there were no assets available, the value of the put here can never exceed the cost of decommissioning.
The cost of insuring against a "gap" due to (a) a portfolio
shortfall from risky investment strategy, or (b) a premature shutdown, can be formulated in terms of the value of a put option, using an option-pricing model. Figures 1 and 2, respectively, illustrate the nuclear plant owner's synthetic balance sheets, valuing the put option as an asset. When the NDT is fully funded on an actuarial basis (Figure 1), the NDT holds a put. The plant owner provides a guarantee valued as the liability G, and might own some fraction, Æ (Æ The value of the put option increases, the greater the weighting of stocks in the NDT portfolio and the amount of time available to exercise the option (i.e., time remaining before decommissioning). This fact could affect a plant owner's business strategy: The owner might wish to choose 60-year SAFSTOR rather than prompt DECON; or (as suggested above) under new competitive conditions, abandon the operation. If sufficient funding is not available when the nuclear plant is shut down, and there is no further recourse against the plant owner(s), consumers or taxpayers must make up the shortfall or endure a long period of unexpected environmental anxiety.
FASB's Proposed Statement of Financial Accounting Standards, Accounting for Certain Liabilities Related to Closure or Removal of Long-Lived Assets, is likely to add impetus for a change in the approach to NDT management. The Exposure Draft proposes recognizing and measuring a liability like that for decommissioning, as the obligation is incurred.11 The Exposure Draft proposes extensive, footnoted disclosures to provide information about the nature and status of any trust funds, insurance, or guarantees to cover the liability.12
The feasibility of termination gap insurance should be studied as a means of extinguishing this liability in the case of a premature shutdown. Such an event represents to some extent a controllable risk, and to some extent a poolable risk. The NRC looked into some form of termination insurance back in the 1980s, and reportedly found the insurance industry unenthusiastic about underwriting the risk because of a moral hazard pitfall.
Further effort is required to estimate the insurance parameters. The magnitude of any given loss event should be relatively small compared to the underwriting and loss absorption capacity of the property/liability insurance industry. Various institutional arrangements like the insurance pools that have been developed for nuclear power will likely provide instructive precedents. The NRC's April 8 Federal Register notice seems to indicate that it will encourage putting some new assurance mechanism in place. t
Donald Korn, a chartered financial analyst, heads a management consulting firm, DHK Associates, in Los Altos, CA. The author gratefully acknowledges the helpful comments of Zvi Bodie (Boston University), Geoffry Rothwell (Stanford University), and Jack L. Treynor (president of Tregnor Capital Management, Inc.). Any errors of ommission or commission remain the responsibility of the author.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.