Arnold Kling  

Reflections on Freddie Mac

Symposium on Bank Size... If You (Insurers) Pay, They (D...

This morning, I was interviewed for background about Freddie Mac. Two interesting questions that were asked:

1. Given that I now believe that Freddie and Fannie should be phased out, how did I feel about working for Freddie?

At the time (the late 1980's and early 1990's), those of us who were economists generally thought that policy makers were largely misguided in their attempt to socially engineer the mortgage market. However, taking the policy goals as given, we felt that at Freddie Mac we were operating very conscientiously from a public trust standpoint. In particular, we believed that we were preventing a recurrence of what had just happened to the S&L industry.

--We marked our assets and liabilities to market, so that we would not be disguising our failures.

--We presented in our annual reports an estimate of the effect of a significant increase or decrease in interest rates on our overall market value. This overall "NPV curve" was flat, which showed that we were carefully hedged with respect to interest rates.

--We used a severe stress test ("the Depression scenario") to set capital requirements with respect to credit risk.

2. Given that Freddie Mac used a severe Depression scenario in this way, why did it not have the capital to survive the drop in home prices that actually took place in recent years?

I don't know the answer. (The crisis hit a decade after I had left the firm.) But I have two suspicions. First, I have read that some of what Freddie and Fannie called capital was really soft. The worst example was counting tax-loss carry-forwards as capital. Second, I suspect that Freddie did not use its internal risk models to evaluate securities with agency ratings. So if somebody put together a security backed by shaky loans ("Alt-A" loans, as they were called), and the security was rated AAA, the security was assumed to survive the stress test, even though Freddie's own models would have predicted that the security would blow up under a stress test scenario. In a sense, Freddie deliberately chose to be fooled by the rating agencies, rather than apply its own know-how in evaluating the risk of the securities that it bought. Again, I do not know for sure that this is how things went down, but it is my suspicion.

Overall, if you go back to the early 1990's, we were at a stage of technocratic optimism. The old regulatory mechanism had failed (as shown by the S&L debacle), we had learned its lessons, and we were going to do things right. Now, I would say that no matter how optimistic you are when you come up with a new regulatory mechanism, the tendency is for regulatory mechanisms to degrade over time. No matter how impregnable the regulations seem when you first put them in place, eventually a government guarantee leads to disaster. For me, that is the main lesson of the Freddie Mac story.

Comments and Sharing

COMMENTS (1 to date)
Will Ambrosini writes:

Its interesting that this regulatory degradation happened in an environment where you claim everyone was acting "conscientiously from a public trust standpoint". Its usually assumed that the regulated find "loopholes" or engage in other nefarious acts because they explicitly want to work around the regulation. In other words, regulatory degradation happens even when everyone has good intentions.

Comments for this entry have been closed
Return to top