Credit Risk and the GSEs

Eakinomics: Credit Risk and the GSEs The core business of Fannie Mae and Freddie Mac – the housing government-sponsored enterprises (GSEs) – is purchasing mortgages from banks or other originators, bundling them into mortgage-backed securities (MBS), and providing a guarantee to MBS investors that they will get their money even if the underlying borrower fails to make payments on their mortgage. The quality of the origination process – who qualifies, the amount of a down-payment, and so forth – determines how much credit risk there is in the system. The only question is who bears that risk. Clearly the point of the guarantee is to make sure the answer is not investors. But that means the GSEs are holding all the credit risk. As we saw in the financial crisis, the GSEs did not have the financial wherewithal to bear that risk and it was assumed by the taxpayers. A key mission for policymakers is to avoid repeating this episode. One strategy to avoid concentrating too much credit risk in the GSEs is the use of Credit Risk Transfers (CRTs). AAF’s Thomas Wade has a nice primer on CRTs, but the basic idea is to shift that risk to a willing private sector investor (for a price). Specifically, CRTs are issued by the GSEs as unsecured debt obligations (no collateral) which means that the buyer might not be fully repaid. To offset this risk, unsecured debt obligations typically carry higher interest rates. The GSEs pay interest on these notes and repay principal based on the performance of an underlying pool of loans. If this loan pool incurs losses, the value of the notes is written down and the GSE is no longer obligated to pay that portion of the principal to its investors. In effect the burden of losses is shifted to the private sector. A second strategy to reduce taxpayer risk is to have the GSEs hold more capital. The Federal Housing Finance Agency has proposed a rule dictating the capital requirements of GSEs. Taken at face value it seems like a step in the right direction: hold more capital, calculate capital based on riskiness of assets, and add special buffers so that capital is available for specific contingencies. Unfortunately, as proposed the capital rule removes the incentive for CRTs. (See Wade’s exposition for a complete discussion.) How does this happen? An alternative to the risk-based capital computation is instead a simple leverage ratio requirement: the GSEs must hold capital equal to 4 percent of total assets. This alternative capital calculation appears likely to be the binding constraint on holding capital. Notice that it is computed without regard to the riskiness of assets, so there will be no benefit to moving some of the risk via CRTs – and removing a risk-based constraint on the GSEs could encourage them to take on more risky business practices. In sum, there has been important progress in address the exposure of taxpayers to risk from the GSEs. But there could be even more progress if the capital rule is modified to retain the incentives for CRTs.