Thursday, April 2, 2009

Stiglitz is an Idiot

Ok, I read on Tyler Cowen's blog that using negative words like "idiot" and "moron" increases the number of commenters, and presumably readers. Clearly my favorite flâneur, Nassim Taleb, has climbed the best-seller ranks with this rhetorical style, so, I'll give it a try. I think Stiglitz is intelligent and educated. But he is also incredibly biased in a totally predictable, simple-minded way. This is not a partisan critique, just look at his activity: he gives speeches about once every two days. Guys like that don't have time to look at the detailed data, or derive a fresh view, but rather fit it to their internal templates. They are using their position to 'make a difference', which sounds nice, but it just means you are an amoral, confabulating advocate. Think lawyer.

In that way, he is not thinking when he writes, merely advocating the best case for his big picture idea that markets should be sublimated to government control whenever there is not perfect information (no uncertainty or asymmetric information, ie always). He edited volumes 1-3 of the collected works of Paul Samuelson, with a heavy emphasis on market failures. He loves pointing out scenarios where a market equilibrium is not a Pareto Optimum or vice versa (Pareto Optimum being the idea you can't make anyone better off without making someone worse off, a rather modest definition of optimality). Government failure is not interesting to those obsessed with market failure, a common blind spot in the Marxist tradition (very little in Das Capital outlines how communism actually works, it's mainly a criticism of free markets).

Anyway, in today's NYT he rails against Geithner's plan for several dubious reasons, but fundamentally because it is a subsidy to banks and investors. It isn't clear if his distinction between banks and investors is the difference between management and equity, or equity owners and bond owners. Whatever, his Big Idea is that there is always imperfect information, therefore markets are not optimal, therefore have the politburo control it. Here's the Nobel Prize winning argumentation in block quotes:

Let’s take a moment to remember what caused this mess in the first place. Banks got themselves, and our economy, into trouble by overleveraging — that is, using relatively little capital of their own, they borrowed heavily to buy extremely risky real estate assets.

Where's the data? Bank leverage, as measured by common equity to total assets, did not materially change over the past 15 years for banks. Did you mean "Investment Banks"? That did change, but only for a handful (ie, Citi and Morgan Stanley). It's not true merely because you say it every other day when lecturing as Expert on Everything.
In the process, they used overly complex instruments like collateralized debt obligations.

Derivatives are the tail of the dog. If mortgage prices did not collapse, CDOs would not have collapsed. CDO's facilitated risk transfer, if anything, it prevented a worse crisis by moving the effect of the housing price collapse out of banks and into Hedge Funds and Government (ie Fannie and Freddie). This diversification is generally considered a good thing, but here it is insinuated that it was the fuel for the risk taking. Sorry, the impetus was at the other end, with $5 Trillion in Community Reinvestment Act pledges for investing in 'traditionally underserved communities'--ie, groups previously not lent to because they were bad credit risks. Mortgage innovations like low, and even no, money down, implies the greatest leverage was to the individuals, not the banks.
The prospect of high compensation gave managers incentives to be shortsighted and undertake excessive risk, rather than lend money prudently.
Anytime a bubble collapses we can say this, but this criticism needs more flesh on it. It seems to imply that any payment scheme that is a function of revenue generated, that does not have an infinite clawback provision, encourages excessive risk taking. That's a nice bargaining position for a socialist like Stiglitz. That may seem like a partisan shot, but it's true: he's always for giving the state more control into economic activity, to the point that it always has veto power and generally sets the agenda and mission statement, compensation, etc. His ends dictate, with perfect predictability, his interpretation of events and his prescription for rectifying the situation.
Banks made all these mistakes without anyone knowing, partly because so much of what they were doing was “off balance sheet” financing.

Partly to be sure. But for commercial banks this was pretty small. The biggest problem was on balance sheet. Look at the data. Look at a regional bank like KeyCorp. Their stock has fallen 75%, and they lost $1B last year. Most of this was an increase in the provision for loan losses on their books. They had $1.25B in credit default swaps at the end of 2008, and noted (page 60 of 10-K) "These swaps did not have a significant effect on Key’s operating results for 2008." Banks, as opposed to Investment Banks, are not primarily hurting because of derivatives. Their exposure would have only been worse without them.

The problems of AIG and Lehman were of taking too much risk in CDOs and credit default swaps, and to this day I have not seen specifics of their balance sheet. That is, until I see what the reference assets for the swaps, I can't assess their actual solvency, or potential for insolvency (and thus a logical run by short-term debt holders). But those problems are not the problems of our banks, 7000 little companies that try to intermediate between savers and investors. Most Commercial Banks are not Investment Banks, their balance sheets are not dominated by marketable securities.

Stiglitz notes the problem of offering banks basically 6:1 leverage, and a 50% match-funding, by looking at the following scenario. Assume there is an asset with a price of $150, that will pay off $200 or $0 next period. A $12 investment by private companies, a $12 match funding by the government, and a $126 senior debt by government. The return on the 'total value' of this transaction is either +33% or -100% [$200 or $0]. Well, with 6:1 leverage and match funding, that means a 208% return in the up state, or a -100% return in the down state. The expected return: 54% {(208-100)/2}. What a deal! Meanwhile, government gets an expected -40% return on 7 times the capital.

A key here seems to be what the prices and returns really are. If you suppose such a bad transaction, clearly, it is a greater subsidy to the private sector. But given the government is trying to fix things without nationalization, presumably a subsidy is a given. The issue is, are these deals this much of a give away? I doubt it. When was the last time a pool of mortgages was worth zero. Geithner's plan refers to pools of mortgages, not mortgage tranches. A tranche of a pool of mortgages can easily go to zero, especially the mezzanine tranches, but the pool itself? If this seems possible to you, you are, well, an idiot.

It seems like you are giving money away, to fat-cats. The problem is, these are mortgages, so they are not 'idiosyncratic risk assets'. You can not buy 100 of them, with a 54% return, and lock in that return. Instead, investing through time, and I doubt any investors are willing to put up large amounts (billions) in investments with a 50% chance of a negative 100% return. Further, the auction method of selling assets implies that even if this were a case of idiosyncratic risk (say, adverse selection made it 50% of these were worth $200, 50% $0), then somehow investors would be colluding to keep the return at these insanely high (54%) levels. Ah, the deals Captains of Industry make in the back rooms. The nice thing about greed is that it implies competition, and competition abhors easy profits the way nature abhors a vacuum.

Stiglitz notes how efficient government is in nationalizing banks. Who's to say how much value was saved, or destroyed, via government's take over of IndyMac (sold at a loss of $9B last January)? But this will only be worse if they take on lots of banks because then their portfolios will be politicized to the n-th degree, as then broader considerations will be in play. Consider that once they took over IndyMac they halted foreclosure on mortgages, leaving zombie properties where 'owners' had no equity in the properties, and bank's could not resell them. That merely added to their losses, and didn't help the little guy, it just delayed the inevitable resale of that property to someone who was a viable homeowner. Do that with 100 IndyMacs and you have really depressed the market, and a larger impact on poor communities that do not need more unowned properties. When the effective owner has no money in the house, it is not maintained, managed, or re-allocated. It just serves as a place for punks to cause trouble. Indeed, the 'owner' of the house in foreclosure limbo has zero incentive to do anything with the property, just wait and pray for hyperinflation (a potential plank II of the Treasury plan).

After the Great Depression, John Kenneth Galbraith wrote 'The Great Crash', and it was a best seller. The thesis was that greed and short-sighted risk taking, especially by corporate executives, caused the Stock market crash, which then caused the Depression. This was the standard perception for decades. Today, I doubt more than 10% of economists think that greed or short-sightedness was a singular, or useful explanation for the 1929 crash and subsequent collapse, or that the Great Depression was caused by the stock market crash.

No comments:

Post a Comment