Thursday, October 25, 2007

Nuclear Energy and Loan Guarantees, Part II

Yesterday, I forwarded a note from Richard Myers, NEI's Vice President of Policy Development, concerning nuclear energy and loan guarantees, the issue that Bonnie Raitt, Jackson Browne and Graham Nash came to Washington to agitate about earlier this week.

But while emotional rhetoric is one thing, the facts are another. Here's Part II:

The Loan Guarantees Authorized by the Energy Policy Act Are Not a Subsidy

For the past few weeks, we’ve been regaled with horror stories about the loan guarantees for new nuclear power plants authorized in the 2005 Energy Policy Act. The story line from the anti-nuclear groups goes like this: “huge bailout” and “a blank check for the nuclear power industry” and “massive subsidies for the nuclear power industry.”

Here are the facts.

First, the Energy Policy Act authorizes loan guarantees for a portfolio of 10 clean energy technologies. New nuclear power plants are just one of the 10. Here’s the list, verbatim from the statute:

1. Renewable energy systems

2. Advanced fossil energy technology (including coal gasification)

3. Hydrogen fuel cell technology for residential, industrial, or transportation applications

4. Advanced nuclear energy facilities

5. Carbon capture and sequestration practices and technologies, including agricultural and forestry practices that store and sequester carbon

6. Efficient electrical generation, transmission, and distribution technologies

7. Efficient end-use energy technologies

8. Production facilities for fuel efficient vehicles, including hybrid and advanced diesel vehicles

9. Pollution control equipment

10. Refineries, meaning facilities at which crude oil is refined into gasoline

Second, a subsidy is when the federal government makes a payment to a private party. The energy loan guarantee program works the other way around. The private parties make payments to the federal government in order to receive the loan guarantees. That’s not a subsidy.

Loan guarantee programs deserve further explanation. Until 1990, the federal budget accounted for federal loan programs and federal loan guarantee programs differently. Loan programs – like student loans – appeared in the budget (in budget-speak, they were “scored”) at full face value. Loan guarantee programs were not scored at all.

That didn’t make a lot of sense. Most federal loans are repaid, so why appropriate and score the full amount of the loan? And a small percentage of loan guarantees went into default, so shouldn’t they be scored somehow in the federal budget?

Congress fixed this problem in the 1990 Federal Credit Reform Act (PDF), which created a standardized way of accounting for loan and loan guarantee programs in the federal budget. Federal agencies that provide loans and loan guarantees are required to calculate a “cost,” following standardized protocols. In simple terms, that “cost” is the expected payments by the federal government less expected revenues received by the federal government.

In most loan guarantee programs, this cost appears in the federal budget as an appropriated amount. The energy loan guarantee program created by the 2005 energy legislation took a different and innovative approach. It stipulates that the Department of Energy cannot issue a loan guarantee until and unless the company receiving the loan guarantees has paid the cost of the guarantee (and, by the way, all administrative fees and costs incurred by the agency in administering the program).

So could someone please tell me where the subsidy is – for advanced nuclear plants or any of the other nine technologies eligible?

The procedures for calculating the “cost” of loan guarantees are established and overseen by the Office of Management and Budget, a relatively obscure Executive Branch agency with enormous power. OMB has a model that all federal agencies must use when calculating the cost of a loan guarantee. It includes a number of assumptions. One of the most important is the probability of default: What is the likelihood that a company receives a loan guarantee, raises debt financing in the capital markets, then defaults on the loan, forcing the government to repay the banks?

We can’t speak for the other nine technologies eligible for loan guarantees, but in the case of new nuclear plants the probability of default is pretty close to zero. Why? Because the companies building these new nuclear power plants will have one billion dollars or more of their own equity (actually, their shareholders’ money) invested in the project, side-by-side with the guaranteed debt. In the event of default, the company loses that investment: The government will seize it to help repay the loan. There’s not an electric power company in the United States that can sustain a billion-dollar loss on a single project. That’s why these projects are so well-planned. Why all necessary safety and regulatory approvals are obtained before construction begins. Why due diligence is so disciplined and exhaustive. Why successful completion and operation is (forgive the word) guaranteed.

This is one of those rare cases when the public interest and the private sector’s interest are perfectly aligned, when both parties have a single common interest in success.

Look for more a little bit later.

6 comments:

Anonymous said...

"in the case of new nuclear plants the probability of default is pretty close to zero."

What about WPPSS?

Sovietologist said...

The critical word here is "new." The companies building these plants don't want to end up losing billions like WPPSS did- so they've studied the lessons of the 1970s and 1980s carefully. WPPSS is a good example of why they'll go to extreme lengths to assure their new plants ultimately go onstream- and that's powerful insurance against defaults.

Richard Myers said...

Good question, but WPPSS wouldn't happen in today's world. We have overhauled the licensing process; ensured that all licensing, design and safety reviews are completed before construction begins; improved construction management and construction techniques (based largely on experience building plants in the Far East), and taken many other steps to ensure that all risks are identified and hedged to avoid debacles like WPPSS. WPPSS was also in part a product of a group of small municipal utilities trying to build multiple nuclear power plants at the same: They got way out in front of their headlights and well outside their core competence. Past experience is a useful source of lessons learned and the industry has, in fact, looked back systematically and incorporated those lessons into planning for the next plants. Past experience is not, however, a useful or relevant predictor of what will happen in the future.

KenG said...

104 operating nuclear units. One bond default. The full benefit of knowing exactly what caused that default. Sounds pretty close to zero to me.

Matthew66 said...

Recent experience at the Tennessee Valley Authority's Brown's Ferry plant indicates that the nuclear power industry is more than capable of delivering a project on time and on budget. The key is in developing a detailed plan, a realistic budget and time frame, and hiring a skilled project manager. I have no doubt that all players have learnt lessons from projects that went off track.

One of the features built into the new Energy Policy Act is protection for the first few plants from delays caused by the untried regulatory system. Unfortunately here in the USA we don't make the loser of a law suit pick up the tab for the winners legal fees, like they do in the UK Australia and other places. If we did there would be a strong disincentive to use law suits as a delaying tactic.

Anonymous said...

Credit Derivatives 101:

The fair value of a credit default swap (loan guarantee) is:

CDS = PD*(1-RF)*Notional

where PD is the probability of default, RF is the recovery fraction and the Notional is the loan amount. PD is usually only a few percent (as already noted here), and RF is about 50 to 75%, in most instances. I think nuclear power plants will valuable assets for decades to come, making it likely that recovery fractions will be fairly high. A CDS is a little more complicated than what I shown here; I have conservatively ignored the discounting.

Anyway, in customary fashion the antinukes are confounding the value of the loan guarantee, which is about 1% the value of loan, with the value of the loan. This exaggerates by a factor of 100. If you have ever paid mortgage insurance, you know the difference between PMI and the actual value of the loan.

The antinukes also fail to mention that the federal government will be paid back many times over, in the form of taxes, if the loan guarantees work, which seems very likely. Also, the construction of a nuclear plant averts countless dollars of damage from the burning of fossil fuels. It's a great trade for the fed.