Putting a Price on Risk

I spent most of the day in Richmond yesterday attending the first Summit on Virginia's Energy Future. I'll write more about the main topic of that session next week, but a statistic from one of the panelists stuck in my mind for the entire drive home. In describing the risks that utilities take on when investing in new power plants, the President and Chief Nuclear Officer of Dominion Virginia Power, David Heacock, explained that over the sixty year life of such a facility, the cumulative difference between their high and low long-term natural gas price forecasts amounted to $7 billion, equivalent to the entire up-front cost of a nuclear power plant. He also suggested that the value of the difference between their high and low forecasts for the price likely to be imposed on CO2 emissions was in the same ballpark. Despite the recent financial crisis and accompanying loss of confidence in sophisticated risk-monetizing mechanisms that failed so spectacularly to account for low-probability events, some businesses have no choice but to assess risk in terms of its dollar impact. And as government fills in for a number of hopefully-temporary gaps in various markets, it must also grapple with risk in this way.

The President's proposal to quadruple the total loan guarantees available for new nuclear power plants has raised some concerns about the cost of backing loans to an industry that has suffered spectacular defaults in the past. Doubtless many of my readers are too young to remember the WPPSS (or "Whoops") default in the early 1980s. The amount in question, $2.25 billion, would seem more like a rounding error in today's inflated terms, but that was a lot of money at the time, and it caused quite a stir. I don't believe the Whoops precedent is relevant to today's emerging nuclear renaissance, other than as a reminder--as if we needed one after the last couple of years--that the risk of default is never zero, and a loan guarantee always costs something.

I also find it interesting that worries about the cost of such guarantees have come up mainly in the context of nuclear power, while loan guarantees, loans and outright grants to a variety of "green" projects and firms have attracted little comment along these lines. A case in point is the widely-celebrated $529 million federal loan--that's loan, not loan guarantee--to Fisker Automotive. As with today's nukes vs. Whoops, there may be no direct analogies to the DeLorean experience or various other sorry episodes in the history of the car business, other than to remind us that the risk of default on that loan is also not zero.

As another speaker at yesterday's session pointed out, we are in an extraordinary time, in the aftermath of a financial crisis and with credit for many firms still frozen. At such times, the government may have to step into roles that are otherwise better left to the private sector, such as financing auto start-ups and backstopping loans to power plants. When that happens, our proper attitude towards the risk that we are taking on collectively is neither to sweep it under the carpet, as has largely been done with various green loans and loan guarantees, nor to assume it approaches 100%, as some seem to be doing in the case of nuclear power. The magnitude of these risks can be quantified and weighed against the cost of doing nothing. It can also potentially be reduced through judicious diversification--recognizing that the government itself controls some of the key risks of default through another of its powers, to regulate. With great power comes great responsibility, and those wielding it today should consider that carefully, as they would be called to account later should some recipient of one of these loans or loan guarantees ever default. Now, that's a 100% certainty.