One of those seemingly-plausible analyses that could just as easily be made in exactly the opposite direction

Fivethirtyeight commenter TGGP links to a news article about zillionaire financier Peter Theil, who “predicted which firms would be bailed out based on whether they leaned Republican or Democratic.” In the words of reporter Peter Robinson, Theil “possesses a preternatural ability to spot patterns that others miss.”

I’ll repeat a bunch of Theil’s reasoning, because on one level it’s interesting while on another level I find it hard to take completely seriously as it stands..

From Robinson’s article:

More than a year ago, Peter explained, he recognized that the leading investment banks were in much deeper trouble than their share prices reflected. . . . Peter mulled the problem. Then it came to him. “I realized that, the more Republican the institution,” he explained, “the sooner it would go down.” Peter did some research, investigating, for example, the campaign contributions that leading figures at each investment bank had made. Then he drew up a simple list, ranking the institutions from most to least Republican: Bear Stearns, Lehman Brothers, Merrill Lynch, Morgan Stanley and Goldman Sachs. (Yes, Republican Hank Paulson is an alumnus of Goldman Sachs. But Peter’s research showed that the Democrats at Goldman heavily outnumbered the Republicans.)

Peter drew up this list, as I’ve said, more than a year ago. What happened next? Bear Stearns collapsed. Lehman Brothers declared the biggest bankruptcy in history while Merrill Lynch disposed of itself in a fire sale to Bank of America (nyse: BAC – news – people ). Then Morgan Stanley and Goldman Sachs escaped to safety.

Events unfolded, in other words, just about as Peter had predicted.safety.

Why? Why did the most Republican banks suffer mortal wounds while the least Republican survived? “I have three explanations,” Peter told me. “I’m not sure which one is right, but I am sure at least one of them is.” [emphasis added]

1. A hostile federal bureaucracy. Despite a Republican in the White House, the bureaucrats who staffed the regulatory agencies, the Treasury and the Fed remain unfriendly to the GOP. Consciously or not, they proved unforgiving toward Bear, Lehman and Merrill but sympathetic toward Morgan and Goldman.

2. An inability to accept reality. At the most Republican institutions, the principals believed in free markets–only too devoutly. Even Milton Friedman would never have argued that markets work perfectly all the time, only that they work a lot better than government intervention. But at Bear, Lehman and Merrill, folks became convinced that the markets possessed almost magical properties. When trouble started, they literally couldn’t believe it.

3. Uncoolness. All the investment banks recruited at the same elite universities, and political correctness at such schools is profound. (If you want proof, just look at the last election cycle. The faculty at Harvard contributed to Democrats over Republicans by a ratio of 93 to seven.) At Yale and Berkeley and Wharton–at all the elite schools–the Democratic Party is cool; the Republican Party, decidedly not. The least Republican investment banks were therefore able to snap up the best talent, leaving the most Republican firms to pick through the leftovers. “In big financial institutions,” Peter said, “it could be that a Republican profile now correlates with technocratic incompetence.”

A hostile federal bureaucracy, an inability to accept reality, or acute uncoolness. Long before this past Tuesday, any of these three explanations would suggest, the Republican Party had already been placed on the defensive.

OK, now here’s why I think this reasoning is, even if correct, a bit suspect. All three of these arguments could so easily be made in the opposite direction.

1. A hostile Federal bureaucracy. In 2008, the Treasury had been in Republican hands for nearly eight years. And, even in Democratic administration, the Treasury is traditionally the home of bankers, not bomb-throwers.

2. An inability to accept reality. Republicans are more likely to understand the judgments of markets and realize that when the economy is going in the tank, that the government has no magic want to fix things.

3. Uncoolness. The default ideological position at Harvard, Yale, Wharton, etc., is to be a liberal Democrat. The Republicans among them are more likely to be able to think on their own and handle themselves well in a crisis.

I’m not saying that the above points are correct either–after all, it was Bear Stearns and Lehman that collapsed first. But, thinking prospectively, I don’t see Theil’s arguments as convincing, and I can well understand why, as the reporter notes, “he never even considered trading on his analysis.” He’s the expert here and I’m not–but in this case perhaps his expertise has led him to a bit of overconfidence in his analyses.

6 thoughts on “One of those seemingly-plausible analyses that could just as easily be made in exactly the opposite direction

  1. Why not evaluate this claim on statistical grounds, in the style of the Lady Tasting Tea?

    There is a 1:120 chance of getting the list precisely ordered. He does well by that test. If we were to relax the burden and say he had to pick the 3 losing banks out of the total 5, he has a 1:10 chance of picking the set. Still not a dead easy task, but far less difficult.

  2. Theil's analysis is hooey.

    One, we know a good deal about how the Bear-Lehman denouement unfolded, and it had absolutely nothing to do with the perceived politics of the firms involved.

    Two, the perceived differences of these firms in ideology is de minimis. I work in the industry, and the furthest left you would ever peg any of them is Goldman's Eisenhower Republicanism.

    Three, not only can you read all three explanations at 180 degrees to those presented, but none survives an encounter with the facts. In aggregate, career bureaucrats are not partisans, those whom imperfect markets made rich do not trust markets blindly, and business schools do not turn out partisan Democrats.

    Full disclosure: I work in the industry for a large, well-known firm whose founder is quite Republican. We did not need a bailout.

  3. Was the list actually only 5 long? How many possible but worthless patterns can one find in such a list? This is more like rearranging letters to see how many words you can make and then picking the ones you like to say, "See, this pattern of words is hidden inside this list."

  4. Personally, I didn't take Thiel's analysis seriously. As he said, he didn't put his money then where his mouth is now. I think he just came up with a provocative post-hoc theory.

  5. For what it's worth, his rankings would also be about right for prestige/reputation, which suggests to me that #3 (Thiel's version) might be the closest to reality. But I'm not sure at all that he has the right mechanism/causality.

  6. Mark Fredrickson:

    "Why not evaluate this claim on statistical grounds, in the style of the Lady Tasting Tea?

    How about: http://en.wikipedia.org/wiki/Confounder

    "There is a 1:120 chance of getting the list precisely ordered."

    That's an interesting statement – can you justify it? Hint – an appeal to the binomial distribution is not woth anything here.

Comments are closed.