Monday, September 22, 2008

Success Through Failure


'No more mistakes, and you're through', was one of the big corporate talks given by John Cleese. His funny, but serious point, is that if you don't make mistakes, you also won't make any great innovations. The book Success Through Failure, by Henry Petroski, argues that mistakes are essential: "We learn wisdom from failure much more than success. We often discover what will do by finding out what will not do; and probably he who never made a mistake never made a discovery". He notes that large bridge failures tend to happen at intervals of approximately thirty years: 1847, 1879, 1907, 1940, 1970. He suggests this is because as time goes on, people forget about failure, and push extensions too far. When was the last residential housing crisis? Before any current politicians, businessmen were around, so it was easy to dismiss those highlighting risks.

The mortgage mistake seems to have been wanting an ownership society that was not consistent with paying its bills; some people saw this coming, and we would have been wise to heed them. Freddie Mac’s former chief risk officer David A. Andrukonis, recalled telling the Bill Syron, who was rewarded for his Boston Fed pieces on mortgage lending discrimination by being made CEO of Freddie Mac, in mid-2004 that the company was buying bad loans that “would likely pose an enormous financial and reputational risk to the company and the country.” Syron disagreed, and either fired him, or he left for other reasons. Andrukonis was right, Syron was wrong. Syron pocketed $38MM over his tenure. I doubt Andrukonis got anything close to that before leaving. Stan Liebowitz had written about the declining underwriting standards. He too was ignored.

I think mistakes as learning tools are highly overrated, because they often have little generality. There are a million ways to do something wrong, and one way to do it right, so going one by one through the wrong ways is a very inefficient way to learn. Theory beats trial and error. Nevertheless, at the margins, we will make mistakes, and can't get too down about them. The key is, the policy on homeownership was debatable before, now it's not. As a society, we can move on to the next mistake. Trial and error is not the best way to learn as a general strategy, but it is part of our overall strategy, that we use towards those risks we take where, from a societal standpoint, the risks were not obvious.

Those who knew this was going to happen should not be mad, but rather, have a kind of wistful regret, comforted by the effect of the correct call on their personal portfolio. If they did not have a sufficient amount of capital to take such bets, then they also did not lose money. So those who are full of self-righteous fury on the risk takers, are hardly credible. There are those like David Andrukonis, whose career suffered, and Bill Syron, who made millions perpetuating this mistake, are rare, and probably both going to be involved in a lot of litigation. Hopefully, Syron will be made to pay like the Enron executives did, but he has a lot more friends than they did.

So now, as we try to fix the problem, it is best to mind two independent issues. One is preventing a banking crisis caused by a run, because the nature of banking is that if all the lenders try to withdraw their money, the banks are all insolvent because their assets are illiquid. This is especially so for banks that have lots of mortgages whose prices are currently unknown,and so assumed to be of very low value. A sequence of bank failures does no one any good. Secondly, we need to adjust underwriting standards back to the bad old days, where people needed to verify income, have a proven credit record, and a 20% down payment. Unfortunately, the bill commented upon today by legislators included saving bad credit card debt, forgiving the current mortgage owners on their payments, and limits on executive pay. These all may have some merit, but they a distraction, and generally difficult implement well (remember, the Clinton surtax on million dollar salaries gave rise to the stock option boom we saw in the 1990's).

5 comments:

Anonymous said...

your take on all this is wrong because other countries who don't have a F&F equivalent experienced bigger bubbles, but whatever. how do you adjust for confirmation bias? in general.

Eric Falkenstein said...

F&F?

Anonymous said...

fnm/fre

Eric Falkenstein said...

I haven't seen data on housing prices, or mortgage defaults in other countries. Do you have any?

I think Fannie and Freddie are basically causes and symptoms. Their mortgage 'innovations' could not have occurred without the zeitgeist that such changes were not just innocuous, but just and good. Academics and politicians saw these important changes as good things. Investor at the bottom were cowed by the academics, who 'proved' that the old underwriting standards were arbitrary and outdated, and they had no hard data to prove the new mortgage rules were too lax.

As per how can I know I'm not shoehorning information into my interpretation. I was led to this interpretation by what I see to be the facts, and researching the data and various others' interpretations. I may be wrong, but if I were to fill my head with an infinite regress of doubts about my biases and assumptions, I would just say 'I don't know anything', which some might say, 'great!', but that is quite nihilistic epistemologically, and not very helpful. No one wants to read someone, or to manage someone, or to invest with someone, who merely says and acts as if he knows nothing. To say 'everyone knows nothing' is the kind of thing that appeals to the inner Holden Caulfield in us, but it's just as naive as thinking experts know everything.

Anonymous said...

foreclosure rates variances are not really comparable b/c in other countries they come after you if you don't pay. you can't just send the keys. but pretty much everywhere they doubled or tripled from a year ago. however when i look at price increases and price/income ratios i expect a tsunami. australia for example is smth like 8 to 1.
i was asking about the confirmation bias not necessarily in this case, but if you knew of good (quantitative) methods to guard against it. so i guess is just gather as much facts and use common sense.