Sunday, April 25, 2010

Biking, Banking, and Bumblebees

Hank Paulson was an Eagle Scout and he rides a bike with his wife. He can't be all bad.

So I learned in Andrew Ross Sorkin's epic about the recent financial crisis, Too Big to Fail. Sorkin's is a detailed work based on hundreds of interviews.

But it blows.

Don't read it before Michael Lewis's The Big Short.

The Big Short
is just as good as Lewis's first book, Liar's Poker, which is saying a lot. I read Liar's Poker decades ago and was gripped by its ambiguity: Wall Street as morally repulsive but at the same time, attractive as a pure sport. This latest book gives the reader that same mixed feeling as it follows the tiny band of investors who made bazillions while Wall Street collapsed. They're making money because they were right, and they didn't buy the hype around them. But it's kind of like collecting on life insurance. You feel kind of creepy about it.

It's immensely satisfying when the light bulb comes on and you say to yourself, "Oh, THAT makes sense!" I call it the explanogasm. That happened a lot while reading Lewis's book. You'll not only learn, in plain English, what CDO's and mezzanine tranches are, but why they were created, why they are so complicated, and why they have names like "mezzanine".

You should also read When Genius Failed, which is about Long Term Capital Management, a hedge fund filled with Nobel Prize winners that went belly up in 1998. I just finished the book, but it was written in 2000. It not only plausibly explains how LTCM melted down, but actually predicts, with uncanny accuracy, the more recent crisis. If you want to know how we got where we are, you have to read it.

While you're at it, one more book on the financial apocalypse that is still better than Sorkin's: The Quants: How a New Breed of Math Whizzes Conquered Wall Street and Nearly Destroyed It. Quants shares many of the faults of Too Big to Fail, (more on this later) but is somehow less pretentious about it all. Read that first too.

Why am I slamming this Too Big to Fail book, which the Financial Times calls "an extraordinary achievement that will be hard to surpass as the definitive account"? Because it's annoying.

More precisely... you know how when Hollywood follows a band of youthful adventurers (see... Lord of the Rings, Harry Potter, etc.) or a team of soldiers or whatever, during the battle scene, anonymous unattractive people are slain by the thousands with nary a sidelong camera glimpse (or, if it's a Mel Gibson film, just enough to capture some creatively gorey, porny shots of parts of bodies being plied or cleaved from the other parts of those bodies) and the mood is all cheery and "yeah, yeah, we're kicking ass!" but THEN one of the actual characters you're following gets wounded or killed, and the orchestra swells and the sounds of battle fade and everything goes into slow motion as the friends, the central characters you're supposed to like, all get closeups of their misty eyes and furrowed brows, and their attractively dirt/ash/blood smeared faces, and there's a lot of consternation as suddenly they realized they're not invincible, and then their collective moral rage gets vented on some more anonymous, dehumanized "enemy" in novel acts of stylized violence, as they slice and bludgeon their way over to the member of the party in danger, and the rest of the battle just seems to give them the time and space to work through their emotional issues, because, you know, one doesn't interrupt the slowmo, with the swelling orchestra and all?

You know what I'm talking about? It's a kind of narrative focus on the heroes that distorts and distracts from what is actually happening.

Ok, so in this book, the heroes are the assholes who created the problems and the government officials who gifted those assholes staggering amounts of bonus money from the tax receipts of our grandchildren.

Yes it's a detailed book. We get ten pages of hand wringing, mourning, weeping, and gnashing of teeth, over the firing of a guy at Lehman who sank the firm with his greedy, pigheaded decisions. Oh, we're told he wasn't actually fired, but kept on as a consultant so he could keep his benefits and retirement. This is a guy who flew into work on a helicopter from Greenwich every day and frequently sent his wife from NYC to LA on shopping trips on his private jet. Even after Lehman, he'll lead a more privileged life than all but a few hundred people in the world. But when he's forced out of the firm, oh my, all of the cinematic tricks are in play. We're supposed to feel his pain. Well, I don't.

I can't generate sympathy for top bankers when teachers are being fired across the country, when working class people are going homeless as a result of their greed.

Why did the crisis happen?

You need to read the books for a fuller picture, but this is the most basic factor: the incentives are screwed up, at every level.

And we're not doing a thing to fix it.

Therefore, there *will* be more crises.

That's it.

There's also this quanty aspect, which is fascinating to me. This is my own explanation, similar to what I've seen elsewhere, but not exactly the same.

The problem is based on a strain of mathematical logic starting with the liar paradox "THIS SENTENCE IS FALSE" and running up through Godel's incompleteness theorems.

Quants make mathematical models of the markets based on data from past markets. But then they act on the basis of those models and thereby *change* the market.

That change is not data going into the construction of the model. It's totally new stuff, new phenomena. And that's not a simple problem with a simple remedy. If you were to predict the change a model would create and put that change into the model as data, you'd have a different model, because, to stay ahead of the market, you have to do the opposite of what the sucker does. When the model says people normally ZIG, you ZAG. (Well, strictly, you ZIG ahead of them, which is still not ZIGging when they did.) Taking your output as input would therefore dictate a change. And if you put THAT change into the model, you'd have a different model still.

There's no way of getting ahead of it. In principle.

The problem is at the center of the "diagonalization" method of proof. It's how you show that the cardinality of the real numbers is greater than the cardinality of the naturals. It's why there's a "halting problem" in computer science.

Let's see. Can I say this more simply?

If you're an engineer, you might build for fifty year flood, but ignore the hundred year flood. First you find out what those flood levels are, and then you build accordingly. Your analysis does not change the patterns of rainfall.

But in finance, if you determine what the hundred year flood is, and then act as if the hundred year flood almost certainly will not happen, and you leverage up, you'll actually make it happen. Right away.

As soon as a strategy or position is assigned a precise value of risk by a model, and then you act on that model, and you're sizeable, or influential and you give other people the same idea, you change the risk.

This is the difference between limning nature with mathematical formulae and predicting the markets on the basis of the past with analogous formulae. Models in physics don't feed back into the system in this totally perverse way.

There's no getting around this problem, except to make very very small trades and to be very very secretive about it. Anybody want to be an inaugural investor in my new secret micro hedge fund? :D

Ok, one more thing. Bumblebees.

Lang Lang plays flight of the bumblebee on the iPad.

I went for a long ride near Oxford today and everything around here is blooming. Plants in hedgerows having anonymous sex with other plants, on the opposite sides of the road, courtesy of pollinators. Some of those pollinators are bumblebees. I hit three of bumblebees while riding maybe 26 mph with the help of a strong tailwind. One of the bees bounced off my front teeth. (Note to self: don't be a mouthbreather.) The other two successively lodged in my bushy right eyebrow, where one of them stung me and briefly paralyzed my eyelid.

Friggin' bees!

But that was just to get another "B" in the title of this post.

The main thing is that you should read Michael Lewis's latest book. Especially if you have anything to do with lawmaking.

I'm jetlagged and bee stung. I need some sleep.


Calvini said...

Welcome to England, now hold still while we give a beesting to teach you how to have a stiff upper lip. Or eyebrow. Whatever, dear boy.

Soon as I graduate, I'm schooling myself on financial shitstorms. Sounds like awesome stuff.

Chuck Wagon said...

The Big Short is fantastic. I can't speak to the Sorkin book but I give a huge endorsement to The Big Short. It's available as an audiobook, perhaps from your local library system as a loaner download?