The Quants, a tale of non-fiction horror

[nancylebov]

The Quants by Scott Patterson, is an account of the people who did quantitative analysis of the stock market, and made a major contribution to the financial crisis.

The short version is that some very smart people didn't have quite enough sense and/or enough virtue.

I used to have a theory that working excessively long hours (90 to 100 hour weeks) at the top of the financial system was a major contributor to crisis, because if you're running that low on sleep, it's hard to think about whether what you're doing makes sense. Besides, I thought it was an amusing irony that if the analysts had been protected by a union or regulations, things would have gone better.

I abandon that theory-- those guys were so convinced they knew what they were doing and so focused on money that they wouldn't have done any better if they'd been getting eight hours of sleep every night. One or two of them might have been hit by inspiration, but it wouldn't have been enough to do any good.

A lot of this starts with Edward Thorp (Beat the Dealer, Beat the Market). He invented card counting (finding a tiny flaw in the rules of blackjack which used to make it possible for someone with a good memory to come out ahead).

The thing about Thorp is that he knew he was operating in a real world-- he knew things could go wrong and he used an equation to be sure of how much to bet safely. (Sorry I don't have time to look up the equation, I'm typing in haste-- the book has to go back to the library and I should be doing ICON prep.) He kept using the equation and thinking about risks when he applied computer analysis to the market.

There were others (notably Taleb, author The Black Swan and Soros) who thought about what they were doing.

However, there were a bunch of guys who thought you could understand the market by just studying the market. They forgot there was a real world which was bigger than their minds. In particular, a common strategy was to find something they thought would go up, and bet big on it. Then they'd figure out something they thought would go down if that went up, and they'd short sell the second thing. Short selling has unlimited possibilities for loss.

This kind of thing isn't nonsense-- some of those funds were bringing in returns of 40% or nearly for multiple years.

I learned about short squeezes from the book. For some reason, my mind doesn't quite wrap itself around short selling. Fortunately, it's possible to treat short selling as a black box. You make money if the price goes down. You lose money if the price goes up.

A short squeeze is what happens when a lot of people sell something short, and then see the price going up. So they all want to buy it now because they want to not lose more money as it goes higher. The price goes up very fast.[1]

Anyway, it was all chugging along nicely and then things happened and both the bets were wrong. This is a way to lose a lot of money very fast.

None of these guys had experience in the market, and I suspect this contributed to their not knowing that things occasionally get very weird.

bradhicks has written about the process of someone finding a new good investment, and then other people trying to piggyback on the insight without understanding it, and buying things which look kind of the good investment, but which are really less and less like it. And, of course, there are plenty of people who'll sell to the unthoughtful buyers.

If this book is correct, there was a lot more wishful thinking than conscious fraud.

I have a notion that part of what went wrong was a shift in the felt definition of "risk". Instead of seeing risk as a region which included some unknown unknowns, quants came to see risk as a perfectly well-defined chance to make money.

I'm planning to read Michael Lewis' The Big Short-- it's an account of three people who figured out by different routes that the financial system was going sour, resisted the social pressure (one of them is autistic) and made a lot of money thereby. All of them placed their bets and then tried to warn people. No one listened.

[1] I'm adding a concept to my mental toolbox. Knowing about a thing isn't enough-- it's worth asking what happens if you have a whole lot of that thing.

Addendum: In the very popular Rich Dad, Poor Dad (published in 2000), the underlying premise was that doing something useful for your living was for suckers. Nothing wrong with doing something useful as a hobby, but you should be making your living in some way that you shuffle the risk off on to someone else. There were government guarantees which made it possible to do this in real estate.

This looked sociopathic to me at the at the time, and no one else seemed to see it that way.

Addendum 2: I just found something I wanted to include Paul Wilmott could see that the quants didn't understand enough, and they were going to blow up the financial system. In 2003, he started a training program so that there would be people who understood what the equations meant, but it was too late.

Comments

[xiphias.livejournal.com]

One of my friends worked as a quant. She believes that what she was doing was ethical and reasonable: she was finding patterns, predicting them, and then kicking her analysis upstairs to other people. She was careful to explain what the assumptions she was working under were, and in what situations the algorithms would fail.

The problem was that the people making the trades didn't understand the algorithms. They'd plug numbers in one end and do whatever the other end said, without actually understanding if the algorithm was relevant in that case -- without actually understanding that an algorithm might NOT be relevant in every case.

[andrewducker]

Yup, wishful thinking is pretty much the stuff that bubbles are made of. When it was DotCom time, people were massively investing in anything with "Net" or ".Com" in the name, no matter whether it actually had anything to do with the internet, or was at all profitable (or might ever be). The same thing happened this time round.

And because the government (and upper managers/directors) tend to assume that the people doing the investing _must_ know what they're doing, this goes unchecked. There was a risk person at a large UK bank who was sacked for saying "This is crazy! It's all going to come crashing down!" about 18 months before it all fell apart. I could tell it was all going to crash. Friends at other financial instituations knew the same thing. But high level people seemed to exist in their own world.

[whswhs.livejournal.com]

I recommend Harold Vogel's book on bubbles and crashes, available from Cambridge University Press (http://www.cambridge.org/us/catalogue/catalogue.asp?isbn=9780521199674). I copy edited it a year ago and learned a lot (and Vogel and I had a very good professional relationship). Vogel starts out by looking at the role of credit expansion and inflation as creating a milieu in which bubbles can start. But what he says is that you get an actual bubble when people see prices rising extremely fast, and become panicked that they will lost out if they don't jump onto the rising bubble . . . and you get a period when economic rationality breaks down.

As I understand it, there are investment advisors who know better than to get involved with bubbles. But typically they get called into the main office, and told that their clients are complaining because they're earning 10% when everyone else is earning 40%, and they'd better change their standards if they want to keep their jobs. Classic perverse incentives.

[nancylebov.livejournal.com]

I have no idea how you'd get to a bubble-resistant society from here.

[rimrunner.livejournal.com]

I feel unusually fortunate in my choice of advisers, then, since both of them have said that trying to go beyond 10% is very risky (for reasons outlined in this post and the comments).

Or maybe it's me, because if I was the type of person to insist on 40% they would probably try to find me something that offered that rate of return. I'm not. I got lucky once (worked for Amazon.com very early in the company's existence) but luck was all it was. I know enough to know how much I don't know about how investing and finance work.

Benjamin Graham describes a tech bubble in The Intelligent Investor. Except for the companies and the raw amount of money involved, it looks a lot like the dot-com bubble.

[redbird]

I made money as a side effect of a short squeeze, a few years ago. (I posted about it at the time, under the heading "Step 3, Profit."(

Someone tried shorting large amounts of Travelzoo stock, without registering that most of the stock belonged to one person, and a lot of the rest was in tiny amounts that had been given away when the company started (the typical stockholder had three shares). That meant the underlying assumption of short sales, namely that enough of the stock to cover the short will always be available, failed. And suddenly the more-or-less worthless stock I had gotten free and mostly ignored was worth significant money.

The company founder and majority stockholder did very well out of the whole thing, I believe. If there's a message here, beyond "selling short is risky," it's that if you're going to try it anyway, your research should include finding out who currently holds the stock.

[womzilla.livejournal.com]

This points out what I was going to say to Nancy: a true "short squeeze" generally comes when it's discovered that there are more people shorting a stock than there are shares readily available to cover the shorts. As in the Travelzoo case, everyone short has to buy shares *somewhere*, and the price can skyrocket.

There was a massive short squeeze in Volkswagon last year that IIRC wiped out a major European hedge fund. They simply hadn't done enough research on how many shares were floating around. (I think that some large percentage of the shares that they assumed were available were actually being held by a single investor using several brokers to obscure his interest; the investor was performing a sub rosa buyout.)

[bradhicks.livejournal.com]

I forget the origin of the rule, but I've run into very few exceptions to it: At every poker table, there is one person there that was invited to be The Sucker. Nobody there likes him, so they won't feel bad when they clip him, and they are confident that he won't be able to tell when they're cheating him. So when you sit down at any poker table, look around and see if you can spot The Sucker. If you can't? It's you.

Ever since deregulation and the defunding of the SEC, Wall Street has gone back to being what the Chicago Board of Trade has always been: a rigged game. And unless you're some CEO's friend or some top trader's friend, and I mean real friend, someone who's known them personally since before they got that job? If you sit down at their poker table, you're The Sucker.

Maybe if you put your money in an insured deposit account you'll lose money relative to inflation. In fact, you probably will. But if you sit down at a rigged poker table with your money, you won't just lose money relative to inflation; whenever they're ready to clean you out, you will lose it all.

Don't be The Sucker.

[nancylebov.livejournal.com]

Unfortunately, it's even tougher than that. The level of self-delusion at the top was so high that the CEO might be giving their best friend the best advice the CEO can imagine- and still be utterly wrong.

Also, though this is a much lower probability problem, I think Madoff was ripping off people who were close to him.