The King of Quants: A Conversation with Emanuel Derman, Part 3

In this, the third and final part of our conversation with financial engineer and author of My Life as a Quant, Emanuel Derman, TradingMarkets CEO and co-founder Larry Connors discuss the practical side of financial engineering, how to get started in the field, what opportunities there are for mathematically inclined analysts and traders and whether or not financial institutions were greedy or just got it plain wrong when it came to the subprime mortgage debacle.

For Part 2, click here.

For Part 1, click here.

Larry Connors: I'll give you a hypothetical: I've got a son or daughter that's 18 years old and decides he or she decides to become a financial engineer. Give me the path for them for the next ten years.

Emanuel Derman: Get a bachelor's degree in solid subjects like math, statistics, applied math, - physics, yes or no. I'm a physicist but I don't think that's critical. But a lot of rigorous things.

Financial modeling isn't that rigorous and though many people may not agree with me but I think it's a mistake to start it too early. That is, unless you want to become a trader when you're 18 and know that for sure.

But if you want to be a financial engineer I think you first need a really firm basis in skills that are not ephemeral, you know? You're always going to need to know real analysis or computer programming or statistics.

I'm not saying don't do any financial engineering, but I'm a little bit wary of jumping straight into the field.

Connors: Yes.

Derman: And then, in an ideal world, learn some finance and then work for two years in the financial business. That will help you decide whether you like the research side or the trading and sales side of the business.

Then, if you want to be a financial engineer, come back and get a master's degree or a PhD in that field. And if you don't, go back and work on Wall Street or get an MBA or do something completely different.

But there is more than one way to skin a cat. You can go straight to a master's in financial engineering. Most of our students in the program actually come directly from undergraduate school, get a masters and then go out into the world and get jobs.

Connors: What's usually their first job? Do you see a common theme as a first job when they come out?

Derman: Well, I'll tell you. One common theme is that 90% of them are foreign-born, foreign-educated.

Connors: Interesting.

Derman: Americans - I am too. But Native Americans don't want to go that way, though more of them do now that quantitative people can take positions and actually trade and do statistical arbitrage. It's better than ten years ago. Because firms see that there's money to be made by using quantitative techniques.

Most of our students are foreign-born and most of them get jobs afterwards - exactly what depends on their skills.

Some as traders. Some of them as quantitative people on a trading desk helping the traders and building models. Some go into risk management. Many move into hedge funds, increasingly. I think the big difference in the last seven or eight years is that most people used to go the sell side. And now many go to the buy side, to asset management, and to hedge funds. That's where the jobs are.

Connors: The last question that I have is this. Let's look ahead over the next two-three-four-five years. We're building models today, hypothetically.

Where are the biggest edges? I remember having a conversation with a neighbor of mine - this was back on the West Coast - who was one of the larger, better commodity traders in the world back in the '80s.

When I asked him how he did it, he basically said he was in markets that nobody else was in and was doing things that, really, nobody else was looking at.

Where would you send someone today to find edges?

Derman: Well, just before I answer that, I sort of think that's true of life in general, especially if you're a little bit lazy. Whether you're in physics or you're in finance, the way to have the biggest bang for your buck is to go into some area - whether you're in trading or in financial engineering - where nobody's done virtually anything.

I mean, BDT was a little bit like that. You know, there are many more sophisticated approaches now, but you get a lot of impact if you grasp the essence of the problem even if your mathematics isn't that sophisticated.

So it's always good to be first through the door, you know? To be the guy that does the first order thing and let other people add the corrections is a nice piece of luck.

Connors: Would you say that possibly the international markets may have these opportunities? Where would you look?

Derman: People were doing emerging markets a few years ago. Japan in the early '90s was a lucrative place to trade volatility because people didn't think in volatility terms.

Ten years ago the people who bought options in Europe during the Internet boom didn't really care about volatility. They cared about getting a cheap way to leverage a bet on pharmaceutical or Internet stocks.

On the other hand, the investment banks or the trading desks that catered to them didn't care about stock up or stock down, they cared about volatility. And I'm making this a little bit more organized and simplistic than it sounds, but desks could charge higher volatility to people who didn't care about volatility but cared about stock direction.

If you can find places where people have different aims and different metrics then you can make money between the cracks - It's sort of like lotteries in the sense that people will pay a small amount of money for less-than-fair odds because they don't really care about the odds.

You can get more of that in places where there are new products, and particularly high margin products. Desks make money by custom tailoring - either by custom tailoring for people who want something that they can't acquire off-the-shelf.

Connors: So it sounds these types of edges, these opportunities, will exist over the next "x" amount of years until those markets become mature.

Derman: I think so, although it's probably harder now because everybody's plunging in and markets go electronic very quickly, where as before, they weren't -

Connors: Yes.

Derman: And so I think those opportunities get harder.

I guess there two ways to make money: one is sort of an asset management approach, where you take positions. The other way is the service side, where you're a wholesaler and you're just trying to make the spread, you know, or you make fancy stuff for people who can't get what they want in a world of simple listed securities.

As I was saying about bond options earlier. You know, you could buy bond options 20 years ago on the Chicago Board of Options Exchange, but they were generic. You couldn't buy one written on the particular bond that you owned. Am I making sense?

Connors: Yes, yes.

Derman: In other words, if you were Fidelity, you could sell a call on some bond future, you know? And you could use that in a complicated way to hedge or speculate. But that wasn't the call you wanted—what you really wanted was, say, a call on the bond you owned, the 6% of 2010.

Connors: Yes.

Derman: To please you, a structuring or trading desk could make a custom call for you (Fidelity, say), and then the trading desk would hedge it with bond futures which took some skill and some models.

You see a similar phenomenon in the last five years with variance swaps, which I used to work on, or volatility swaps, where buy-side firms want to take a speculative bet on volatility -

Connors: Yes.

Derman: Let's say on S&P 500 volatility - there are complicated ways to do it by buying whole portfolios of options but it takes a lot of technology - and what a lot of trading desks have offered instead is to say to the buy-side, "I'll simply sell you the volatility bet, and I'll charge you a spread; and I'll take the trouble to do the technological finance engineering part to give you what you want."

Connors: So why do we see - if they were just playing the house, why do we see this type of meltdown that we're seeing? Why did so many of them get into trouble? Is it just - is it improper models? Is it greed?

Derman: I think the models for the subprimes were not that sophisticated. The real problem was: what do you assume is gonna happen to housing values. And what's gonna happen to default rates. And I think if you get those things right, I think what comes after that is not so sophisticated.

It's the inputs that are bad there. People assumed the good times would last forever.

And they hold on too much to their own. I was asking somebody last night - we had a guy talk at a seminar that I organized - a banker from a major Wall Street firm. Somebody asked him the same question, and he said that the CDO issuers had a pipeline of mortgages coming in and had a bunch on their books when the trouble started. And when things staredt to go bad they couldn't get rid of it fast enough.

Connors: Yes. Yes.

Derman: But I think it's about believing your own story to some extent. I think it's a cross between believing your own story, believing the future's going to be better than you think. And a little bit of greed.

Maybe not even greed. Just riding a good thing for too long, the classic mistake.