FHA reform: if not now, when?

The Center for American Progress (CAP) released a short report on the financial status of the FHA earlier this week. Its title, “Too Early to Sound the FHA Alarm,” reflects its criticism of my recent call for an immediate recapitalization of FHA’s main insurance fund based on analysis which showed the fund to be insolvent because of substantial underestimation of risk and future losses by the FHA and its outside actuary (my report is available here.)

The CAP report does not delve into the details of my estimates of the underestimation of future losses on FHA mortgage guarantees. Rather, it makes the interesting argument that FHA should not be reined in or recapitalized because it is needed as part of countercyclical macroeconomic and housing policy. I have real sympathy with this view, and noted in an editorial published last month by The American that we needed a multi-year plan to reduce FHA’s scale to a more manageable level because an immediate, sharp reduction would be too risky given how fragile the housing market remains.

However, that does not mean FHA should not be recapitalized now. It should, and a key reason why is that we need to start recognizing the true costs of this program. Basic economics tells us that we want the program’s size to be such that the marginal benefit from the last mortgage guaranteed just equals the expected costs from that guarantee. If we keep pretending that FHA only provides benefits and no real costs, then it will forever be too large and too risky, and will become increasingly so over time. This is already happening, as its total outstanding insurance guarantees on single family mortgages now amount to just below 7 percent of annual gross domestic product.

As my November editorial in The American noted, FHA is risky not just because of its size but because its business model shares key risk traits with the now-infamous CDO-squared (collateralized debt obligations-squared, or CDO2) securities that were among the first to fail in the financial crisis. For those not familiar with those securities, a CDO2 essentially was backed by a pool of high loan-to-value mortgages with little or no home equity cushioning them against losses in the event house prices declined. Because all the loans shared the same weakness (little or no equity cushion), they were all susceptible to any economy-wide shock that lowered house prices or raised unemployment. When house prices fell by a modest amount at the beginning of the crisis, all the loans were underwater, making them more likely to default. That is precisely what happened, as we know.

FHA obviously is not issuing CDO2 securities, but it shares the same financial weaknesses. It guarantees pools of highly leveraged assets, most of which have little or no equity cushion, which leaves them more likely to fail en masse if house prices fall further. My report shows that over half the existing pool already is underwater, so the risk of future default already is very high. What this means is that using FHA in the way encouraged by CAP is very risky countercyclical economic policy that could end up imposing very high costs on the nation. This is an example of why economists and policymakers should abide by a version of Hippocratic Oath in medicine: first, do no harm. Alternatively, we need to develop much less risky countercyclical policy interventions.

We need to recapitalize FHA now and put it on a sound financial footing now precisely because the potential harm from its risky business model is so great. This does not mean shutting them down or shrinking them substantially right now because of on-going weakness in the housing market. However, it is imperative that this be done on an orderly basis over the next few years. For those who believe otherwise and that we should just keep increasing the risk level, I have a simple question: if not now, when?

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>