Peter Orszag's short letter to Congress on the cost of the proposed help for Fannie and Freddie is well worth reading in full. It is a bit misleading to say, as most reports did, that the Congressional Budget Office believes the bail-out "could cost $25 billion"--a less scary figure than many on Capitol Hill had apparently been expecting. This is a probability-weighted average of two more likely, but very different, scenarios. One (rated a better-than-50 percent bet) is that the Treasury will not need to hand over anything at all. If all goes well, merely announcing that it is ready to do whatever it takes will be enough to reassure the markets. The other possibility is that the markets are not reassured, in which case the necessary bail-out would likely be much more than $25 billion, and possibly more than $100 billion.
CBO's estimate recognizes that there is a significant chance--probably better than 50 percent--that the proposed new authority for the Secretary would not be used before it expired at the end of December 2009. If the proposal is enacted, private markets might be sufficiently reassured to provide the GSEs with adequate capital to continue operations without any infusion of funds from the Treasury; during that time, it is possible that expectations about the duration and depth of the downturn in the housing market may brighten. Under that scenario, the temporary authority would not be used and thus would involve no budgetary cost. In CBO's view, however, that scenario is far from the only possible result. Indeed, many analysts and traders believe that there is a significant likelihood that conditions in the housing and financial markets could deteriorate more than already reflected on the GSEs' balance sheets, and such continuing problems would increase the probability that this new authority would have to be used. CBO's cost estimate therefore accounts for both the possibility that federal funds would not have to be expended under the new authority and the possibility that the government would have to use that authority to provide assistance to the GSEs... CBO's estimate of $25 billion in costs over the 2009-2010 period reflects a probability-weighted average of how large those injections might need to be, including zero as a potential outcome.
And the letter's last paragraph makes a point worth pondering:
A strong argument can be made that if the Treasury used the proposed authority, the GSEs' operations should be incorporated directly into the federal budget. That is, the proposal, especially to the extent it would result in any government acquisition of an equity stake in the GSEs, raises a significant budgetary question. Currently, data on the GSEs are reported along with federal budget information each year, but the activity of those entities is not encompassed within the budget. That treatment could change if the federal government's financial stake or control changes in a significant way.The