October 7, 2007
Brooklyn, New York
n+1: All right, let’s get to it. Is America now a Third World country?
HFM: No, we’re a First World country with a weak currency. From time to time, the dollar’s very weak; from time to time, it’s very strong; and unfortunately what tends to happen is people tend to extrapolate. But in reality, over the very very long term, currency processes tend to be fairly stable and mean-reverting. So the dollar’s very weak today, but that’s no reason to believe the dollar’s going to be weak forever or that, because it’s weak today, it’s going to get dramatically weaker tomorrow.
n+1: But you, in your work, are not dealing with the long term . . .
HFM: No, we’re dealing with the short term. But I’ll tell you, in our work we don’t trade the G-7 crosses because we don’t feel we have an edge on that. Dollar-sterling, dollar-euro, or dollar-yen: it’s amazing how many brilliant investors have gotten egg on their face trying to trade the G-7 crosses. I can think of so many examples—where people make these really strong calls, that seem very sensible, and then get killed. A very good example is Julian Robertson in the late ’90s being short the yen against the dollar. Japan had just gone through this horrible deflation, the economy was in the shitter, the banking system was rotten. All these things should lead a currency to trade weaker, and he got very very long the dollar, short the yen, and a lot of people did alongside him, and basically there was a two- or three-week period in ’98 when they had the financial crisis and the yen actually strengthened 10 or 15 percent. I can’t remember the exact numbers, but all these guys just got carried out.
n+1: “Carried out,” is that a term of art?
HFM: Carried out . . . like basically they’re carried out on a board, they’re dead. Another example of that, a personal example: Generally every year, at the beginning of the year, banks that we deal with will have events, dinners or lunches, where they gather some of their big clients and discuss themes for the coming year. They encourage everybody to, you know, go around the table: “What’s your best trade idea for the coming year?” And at the beginning of 2005 I was at a dinner with some fairly prominent macro investors, and it was almost like a bidding war. The first guy would say, “I think the best trade is short dollar, long euro, it’s going up to $1.45.” At the time, I forget, maybe it was $1.30. And the next guy would go, “No, no, you’re so naïve. $1.45? It’s going to $1.60!” And it was a competition for who could be more bearish on the dollar and win the prize and be the least naïve person at the table. “It’s going to $1.65 and probably higher! Maybe $1.75!” At the eighteen-month horizon.
Now considering that everyone at the table being super-bearish on the dollar probably meant that they were already short the dollar and long the euro, I went back and looked at my portfolio and said: “Any position I have that’s euro-bullish and dollar-bearish, I’m going to reverse it, because if everybody already said ‘I hate the dollar,’ they’ve already positioned for it, who’s left?” Who’s left to actually make this move happen? And who’s on the other side of that trade? On the other side of the trade is the official sector that has all sorts of other incentives, nonfinancial incentives, political incentives. They want to keep their currency weak to promote growth or exports or jobs; or they have pegs, peg-regimes that they need to defend, and they don’t really care about maximizing profit on their reserves, they’re not a bank trying to maximize profits, they have broad policy objectives—and infinite firepower.
n+1: So you did well.
HFM: Well, we didn’t lose. I mean, I don’t bet on this process, but sometimes there are positions you have that have a certain derivative exposure to the dollar-euro, and we tried to be careful not to take too much of that. Because we thought that this consensus, this super-strong consensus that the dollar’s got to go weaker, actually represented a risk that the dollar would go in the other direction.
n+1: How do you know all this stuff?
HFM: All what stuff?
n+1: All the stuff you know? Did you go to—
HFM: I didn’t go to business school. I did not major in economics. I learned the old-fashioned way by apprenticing to a very talented investor. I didn’t even know what a hedge fund was when I started. I’m sure today I would never get hired.
n+1: Really?
HFM: Yeah, it would be impossible because I had no background, or I had a very exiguous background in finance. The guy who hired me always talked about hiring good intellectual athletes, people who were mentally agile in an all-around way, and that the specifics of finance you could learn, which I think is true. But at the time, I mean, no hedge fund was really flooded with applicants, and that allowed him to let his mind range a little bit and consider different kinds of candidates. Today we have a recruiting group, and what do they do?—they throw resumes at you, and it’s, like, one business school guy, one finance major after another, kids who from the time they were 12 years old were watching Jim Cramer and dreaming of working in a hedge fund. And I think in reality that, if anything, they’re less likely to make good investors than people with more interesting backgrounds.
n+1: Why?
HFM: Because in the end the way you make a ton of money is calling paradigm shifts, and people who are real finance types, maybe they can work really well within the paradigm of a particular market or a particular set of rules—and you can make money doing that—but the people who make huge money, the George Soroses and Julian Robertsons, they’re the people who can step back and see when the paradigm is going to shift, and I think that comes from having a broader experience, a little bit of a different approach to how you think about things.
n+1: What’s a paradigm shift in finance?
HFM: Well, a paradigm shift in finance is maybe what we’ve gone through in the subprime market and the spillover that’s had in other markets, where there were really basic assumptions that, you know, were wrong.
The thing is that nobody has enough brainpower to question every assumption, to think about every single facet of an investment. There are certain things you need to take for granted. And people took for granted the idea that, “OK, something that Moody’s rates triple-A must be money-good, so I’m going to worry about the other things I’m investing in, but when it comes time to say, ‘Where am I going to put my cash?,’ I’ll just leave it in triple-A commercial paper.” It could be the case that, yeah, the power’s going to fail in my office, and maybe the water supply is going to fail, and I should plan for that, but you only have so much brainpower, so you think about what you think are the relevant factors, the factors that are likely to change. But often some of your assumptions are wrong.
n+1: So the Moody’s ratings were like the water running . . .
HFM: Exactly. Triple-A is triple-A. But there were people who made a ton of money in the subprime crisis because they looked at the collateral that underlay a lot of these CDOs [collateralized debt obligations] and commercial paper programs that were highly rated and they said, “Wait a second. Underlying this are loans that have been made to people who really shouldn’t own houses—they’re not financially prepared to own houses. The underwriting standards are materially worse than they’ve been in previous years; the amount of construction in particular markets is just totally out of proportion with the sort of household formation that’s going on; the rating agencies are kind of asleep at the switch, they’re not changing their assumptions and therefore, OK, notwithstanding something may be rated triple-A, I can come up with a realistic scenario where those securities are impaired.” And pricing on triple-A CDO paper was very, very rich. Spreads were very, very tight, and these guys said, “You know what? These assumptions that triple-A is money-good, or the assumptions that underlie Moody’s ratings . . .”
n+1: “Money-good”?
HFM: In other words, if you buy a bond, you’re going to get back your principal. It’s money-good. You’re going to get a hundred cents on the dollar back.
But in reality this was wrong, and people were able to short triple-A securities very cheaply. They weren’t paying a lot to be short and they made huge money on triple-A securities and triple-A CDO paper that now trades at fifty cents on the dollar. I mean that’s like the water’s not running today, right? The sun didn’t rise. But if you were trained in finance, you probably are more likely to take for granted that, “The rating agencies have a very sound process, credit analysis, the same process that I’ve been trained in, all the assumptions that I use are kind of the same as the assumptions they use.” In the same fashion, if you assess the attractiveness of a trade based on historical data from a time when people weren’t really actively doing that trade, and then suddenly everybody’s doing that trade, the behavior of the trade will be different. And if you’re trained the same way as everybody else, in general you’re all going to behave the same. And when everyone behaves the same, that makes trades a lot riskier: everybody’s buying at the same time, you get bubbles, everybody’s selling at the same time, you get crashes.
A good example of that is . . . I don’t know if you’ve heard about the problems that cropped up over the summer in a type of business called statistical arbitrage? Stat arb?
n+1: . . .
HFM: Quantitative trading?
n+1: . . .
HFM: Goldman Sachs had a fund that lost 30 percent, and Highbridge had a fund that lost a lot of money. Stat arb is, basically, computerized trading of a huge universe of stocks based on a set of models. And those models can be technical models like momentum or mean reversion, or fundamental models like just, “Buy stocks that have high cash-flow yields and sell stocks that have low cash-flow yields.” That’s a gross simplification, but the core of it is that there are predictable relationships between stock price history and future performance, or between fundamental variables of a company and stock price performance, and these are broadly reliable. No given stock is going to perform in line with the models. But if you’re trading a universe of five thousand stocks, in general you’ll have enough of an edge that you’ll make money.
n+1: And so the computers themselves are making these trades?
HFM: You build the models and the computer does the trading. You actually do all the analysis. But it’s too many stocks for a human brain to handle, so guys with a lot of physics and hardcore statistics backgrounds come up with ideas about models that might lead to excess return, and then they test them and all these models get incorporated into a bigger system that trades stocks in an automated way.
n+1: So the computers are running the . . .
HFM: Yeah, the computer is sending out the orders and doing the trading.
n+1: It’s just a couple steps from that to the computers enslaving—
HFM: Yes, but I for one welcome our computer trading masters.
People call it “black box trading,” because sometimes you don’t even know why the black box is doing what it’s doing, because the whole idea is that if you could, you should be doing it yourself. But it’s something that’s done on such a big scale that a human brain can’t do it in real time. The problem is that the DNA of a lot of these models are very, very similar—it’s like an ecosystem with no biodiversity, because most of the people who do stat arb can trace their lineage, their intellectual lineage, back to four or five guys who really started the whole discipline in the ’70s and ’80s.
And what happened is, in August a few of these funds that have big black box trading books suffered losses in other businesses and decided to reduce risk, so they dialed down their black boxes. The black box system started unwinding its positions, and every black box is so similar that everybody was long the same stocks and short the same stocks. When one fund starts selling off its longs and buying back its shorts, that causes losses for the next black box, and the people who run that black box say, “Oh gosh! I’m losing a lot more money than I thought I could. My risk model is no longer relevant; let me turn down my black box.” And basically you had an avalanche where everybody’s black box was being shut off, causing incredibly bizarre behavior in the market.
n+1: By the black boxes?
HFM: Well, in the part of the profit-and-loss that they were generating to the point where, to give you an example from our black box system, because we have one . . .
n+1: A big black box?
HFM: Actually I think it’s gray, and it’s not in our main office, it’s off site. And we made sure it has no arms or legs or anything it could use to enslave us. But we had a loss over the course of like three days that was a ten-sigma event, meaning it should never happen based on the statistical models that underlie it. Why? Because the model doesn’t assume that everybody else is trading the same model as you are. So that’s sort of a meta-model factor. The model doesn’t know that there are other black boxes out there.
n+1: What’s a “ten-sigma event”?
HFM: Meaning that it’s ten standard deviations from the mean . . . meaning it’s basically impossible, you know? But returns are very fat-tailed, so the joke that we always say is, “Oh my God, today I had a loss that’s a six-sigma event! I mean, that’s the first time that’s happened in three months!” It’s like a once-in-ten-thousand-years event, and I haven’t had one in the last three months.
n+1: It happens all the time?
HFM: Those kind of things do happen, yeah. And usually it happens because there’s a flaw in the model and in the assumptions that people made. Black box trading is small relative to all trading, but there was a ton of it going on and it was all very similar.
n+1: So why did all of the hedge funds have this subprime mortgage paper?
HFM: Well, some hedge funds did and some didn’t. Some hedge funds made a lot of money being short it. Some hedge funds lost money being long it. Where the losses are concentrated, though, are not so much in the hedge fund world. The losses are concentrated at banks—a lot of European banks, Asian banks. Even the Chinese central bank has exposure.
People talked about this being a hedge fund problem, but it wasn’t really a hedge fund problem. There were some hedge funds that were taking pure subprime exposure, but most hedge funds were basically in the CDO business, so they would buy all sorts of mortgage pools. They buy mortgages, and then they package them and tranche the pools of mortgages into various tranches from senior to equity. So basically you have a number of tranches of paper that are backed by the mortgage pools and there’s a cash-flow waterfall—the cash comes in from those mortgages, and a certain tranche has the first priority. Then you have descending order of priority, and the hedge fund would usually keep the last piece, which is known as the equity, or the residual, as opposed to the stuff that was triple-A—that’s the most senior paper.
So if you had a pool of half a billion dollars of mortgages, maybe there would be 300 million dollars of triple-A paper you would sell to fund that, and then there would be smaller tranches of more junior paper. And the buyers of that paper, particularly the very senior paper, the triple-A paper, were not mortgage experts—they say, “It’s triple-A? I’ll buy it.” These are money market funds; the spread that they’re getting paid is very small, so they can’t hire a lot of analysts to go and dig in the mortgage pools and really understand them, they just rely on the rating agencies, and that’s their downfall.
It’s kind of an interesting interaction, in the sense that a lot of this mortgage project was almost created by the bid for the CDO paper rather than the reverse. I mean, the traditional way to think about financing is, “OK, I find an investment opportunity that on its face I think is a good opportunity. I want to deploy capital on that opportunity. Now I go look for funding. I think that making mortgage loans is a good investment, so I will make mortgage loans. Then I will seek to fund that activity, perhaps by issuing CDO paper, issuing the triple-A, double-A, A, and down the chain.” But what happened is, you had the creation of so many vehicles designed to buy that paper that the dynamic flipped around. It was almost as if the demand for that paper created the mortgages.
n+1: Created the loans?
HFM: Called forth the loans, because it became a really profitable business. You saw where you could issue these liabilities. Say I could issue these liabilities at a weighted average cost of LIBOR [London Interbank Offered Rate] plus one-fifty, and I know all I have to do is just push that money out the door, LIBOR plus three hundred, and I’ll make a huge amount of money from that origination activity or just on the equity piece that I keep, which is highly, highly leveraged. The person who really knows the mortgages is not the person who is really taking most of the risk. The people who were taking most of the risk were the undifferentiated mass of buyers out there.
n+1: Right, and when you say the person who knows the mortgage, meaning the person who knows that the person they find on the street . . .
HFM: May not be a good credit, right? In financial markets, bad things tend to happen when you divorce the people who take the risk from the people who understand the risk. In the subprime market, that distance just increased and increased and increased. It started out that you had mortgage companies that would keep some of the stuff on their own books. Subprime lenders, it wasn’t a big business—it was specialty lenders, and they made risky loans, and they would keep a lot of it on their books.
But then these guys were like, “Well, you know, there are hedge fund buyers for pools that we put together,” and then the hedge fund buyers say, “You know what? We need to fund, we need to leverage this, so how can we leverage this? Oh, I have an idea, let’s create a CDO and issue paper against it to fund ourselves,” and then you get buyers of that paper. The buyers of that paper, they’re more ratings-sensitive than fundamentals-sensitive, so they’re quite divorced from the details. Then vehicles were set up that had a mandate to kind of robotically buy that paper and fund themselves through issuing paper in the market.
n+1: Black boxes?
HFM: No, not the black boxes. But there wasn’t a lot of human judgment going on. In reality those guys were so far from the true collateral that underlay the paper—they have no idea. It’s like they’re buying collateralized paper of a conduit, the conduit’s buying triple-A paper of a CDO, the CDO is set up by a hedge fund that’s bought mortgage pools from a mortgage originator, and the mortgage originator is the one who realizes that they lent half a million dollars on a house in Stockton, California to . . . someone who makes fifty thousand a year. That’s where the specific knowledge about the risk resides, but the ultimate risk-taker is very very far away from that.
So what happened is this machine—the big machine that wanted to gobble up rated paper—needed to be fed. There were people who could make a lot of money feeding the machine, and they were like, “We need to keep originating mortgages, and feeding them to the machine.” Someone is hungry for paper, paper will be created. And that’s almost never a good thing, when lending decisions are being driven by the fact that many, many steps down the chain there’s just someone who wants to buy paper.
n+1: Why was all the press about the mortgage crisis about the hedge funds?
HFM: People like talking about hedge funds. They like to blame us for everything. And there were hedge funds that lost a lot of money.
We’ve had our share of lumps from the black boxes and subprime at our fund, but we’re still standing.
n+1: You lost on the subprime?
HFM: We did. We were involved in creating CDOs.
n+1: You were?
HFM: Yeah, yeah. Not me, personally. But we have people who did it. They would buy mortgage pools, package them into CDOs, and sell off senior liabilities. We kept the equity pieces ourselves, and, you know, those equity pieces are worth—they’re worth pretty much zero. But the amount of money we lost was kind of small relative to the amount that was lost on the whole. Because—OK, we had the equity on a CDO with half a billion dollars in mortgage collateral, and we issued paper for, you know, 450 million dollars, and kept 50 million dollars of the most junior piece for ourselves. OK, so we lost 50 million dollars. If that mortgage pool is now only worth 300 million dollars, that’s 200 million dollars of losses, 150 million dollars in losses are borne by the people who bought the CDO paper.
n+1: From you?
HFM: From—yeah, from the CDO we set up.
n+1: Are they mad at you?
HFM: Well, our CDO paper performed better than average. So I think they’re happy we did a better job than our competitors—but they’re not happy they lost money.
n+1: Is the person who ran that—is he going to get fired?
HFM: He was already fired.
n+1: Really? He’s gone?
HFM: He’s gone. He was a guy who knew the market really, really well, who was a real expert in the nuts and bolts of mortgage lending, and really knew the collateral—but he was a true believer. And there were other people at the firm, say, at the middle of last year, who were not mortgage experts, who were saying, “I see the run-up in housing prices in some of these geographies, and I just don’t really get it. I go down to Florida and see the forest of cranes, and I wonder, who’s going to be in all those apartments? And I hear about all sorts of friends who are getting loans to buy apartments or houses speculatively and are lying about the fact that it’s not a primary residence, and I see these commercials on TV about low-doc, no-doc mortgages—and there is no way, there is no way that this is not going to end badly. And I see that these mortgages are being created by this massive demand for CDO paper, by this robotic bid, and this is the perfect example of a bubble—and we should be short, we should be short subprime paper.”
n+1: This is what guys do? They travel around Florida, they watch TV?
HFM: Just in your normal life. I mean, I trade a different market, I don’t trade subprime, but I travel for other reasons, and some of my partners do the same thing. And a number of us thought, “This is just crazy. We should be short. This is a bubble waiting to be popped.” But the person who was the expert, who traded subprime paper and issued the CDOs, was a true believer in the paradigm: “In 2003, people said that the credit quality of the average subprime mortgage was deteriorating, and now look, those mortgages have performed fine. The subprime market works.”
And, hey, he was the expert—you defer to the expert.
n+1: He didn’t listen.
HFM: But he’s the expert, right? It’s a tough thing. If you have somebody who’s really trained in the mortgage business, he’s been in the mortgage business for fifteen years, in equilibrium he’ll do a great job. He’ll be able to pick, of the mortgage pools out there, which is the good one, which is the bad one. He did a very good job of that, because the ones that he picked were better than the market. But the guy who’s just buried in the forest . . . he’s not going to see the big picture, he’s not going to catch the paradigm shift.
n+1: And now he’s gone.
HFM: And now he—will have plenty of time to think about the big picture.
n+1: [laughs evil laugh].
HFM: [also laughs evil laugh].
