Returns to the carry trade

Romesh Vaitilingam interviews Martin Eichenbaum for Vox

February 2009

Transcription of an VoxEU audio interview [http://www.voxeu.org/index.php?q=node/30]

Romesh: Welcome to Vox Talks, a series of audio interviews with leading Vaitilingam: economists from around the world. My name is Romesh Vaitilingam and today's interview is with Martin Eichenbaum, Professor of Economics at Northwestern University. Martin and I met at the American Economic Association's annual meetings in San Francisco in January 2009, where we spoke about his research on the carry trade.

Martin: Well, the carry trade is actually a particular trade that financial Eichenbaum: folks make all the time, and that is you borrow at low interest rate currency, you lend into high interest rate currency and then you take the proceeds and come back into your initial currency. That's going to be a great strategy as long as the currencies you are getting don't depreciate in value.

Well it turns out that you might think that the returns or the payoffs to this particular speculation strategy are just risky, and economists for a long time have said, well sure, you can make returns on this strategy, but it is full of risk. Well it turns out that that is just plain wrong in this context. The payoffs for the carry trade are uncorrelated with virtually any measure of risk that you care to point out.

To make this very dramatic, suppose you are a money manager, you have stock market returns. And they say, well, give me some money and I will invest it in the carry trade. Are you increasing risk? No, you are actually decreasing risk and you are increasing the rate of return by combining those two assets together.

Now, am I starting to run a mutual fund, why am I even talking about this? Well, it turns out that in traditional macro models, we have nothing about these financial flows. So if you go to the IMF model, basically their model assumes that most of the financial flows between countries reflect goods and services. In reality, the vast majority of financial flows have to do with the speculative strategies and not trading goods and services.

Moreover, the standard models used by the Board of Governors, the IMF, at their core is a relationship called uncovered interest rate parity. Without being fancy, that says that the carry trade strategy can't be profitable and it is. So we are working with models for policy that are wildly inconsistent with the profitability of this widely practiced financial strategy.

Romesh: OK, you are looking for an explanation for something that people say shouldn't really be there?

Martin: I want to even make it really tougher now. So you might say, well suppose with some probability, a low probability, say you are earning pesos, that's the high interest rate currency for the sake of argument, and they depreciate while you are holding them. So it is true, you made this little tiny interest rate advantage by lending into pesos, but you now get killed because the peso depreciated. The phrase that some people use is "you are picking up pennies in front of trucks." So we devised a strategy which rule that out.

We just basically said, well what if you just buy a put or a call on pesos depending on whether you are lending or borrowing pesos. So what that does is, basically it protects you from a big appreciation or depreciation, whatever it is you are afraid of.

Your initial reaction would be: that can't make any sense, the premiums on those puts and calls would have to wipe out the profit. Guess what? When you actually look at the data we got from the Chicago Mercantile Exchange, it doesn't. In fact, they are so cheap those puts and calls, that it basically doesn't affect anything.

So now, I have made this thing really crazy. So you are waiting for a truck but you just bought insurance against the truck, so even if it comes, you are fine.

Romesh: So now you got to get to the root of this problem?

Martin: No, I am not going to have a great answer; what we do say is, there are trucks, but they don't come in the form of very bad returns. The question is, when you lose a buck, how much does it hurt you? And we have lots of fancy names for that, the marginal utility of consumption, all sorts of ways of evaluating that. What we devised is a nonparametric way, that is to say we don't take an explicit stand on the functional form of what this object is. But what the truck is, it turns out when you lose in the carry trade, we are saying it is very very painful, it is particularly painful.

So think about just what is going on now, the price of risk has gone crazy. Anybody who lost a buck last month in the carry trade, it really hurts. So it is not that they lost an extraordinary amount, some of them did, but not on average. It really hurt them badly.

So that leads you into questions of how do people evaluate risk under various scenarios. And maybe that leads you away from rational expectations where they just understand the nature of the world all the time.

And it could be that what these peso states are or these truck states, if you like, are situations where you really don't know what is going on anymore and you are incredibly averse to risk or you are incredibly averse to losing money in any trade. Working that out takes you literally to the advances, which we are not doing, but people like Lars Hansen and Tom Sargent are, of modeling financial traders in a world where they don't really know the right model.

No one has quite made the connection yet between this carry trade which remains an enormous puzzle and the kind of research that people like Hansen and Sargent are doing, but that's clearly the link. And if we believe that, what does that say about rational expectations in our day-to-day business cycle models?

So the stakes there, what starts off as a very narrowly investment financing project, very quickly become issues that cut to the core of how we model macro-economic behavior.

Romesh: How does it feel at a time like this of extreme exchange rate volatility? It seems to be that exchange rate volatility happens in short periods like it has recently with the sterling going down massively against the dollar, the euro being up?

Martin: Absolutely. Volatility is enormous. Well, so one question is what happens to the carry trade strategies that we are proposing over the last couple of months? That's one way to think about it. Did you get taken to the cleaners? Well, it turns out if you did very simple carry trade strategies, you probably did lose a lot, but if you did our hedge carry trade strategy, you lost a lot less than you did in any other financial market. So it is a difference between losing 20% in stocks and 3% when this truck came with this strategy; with 3% you don't have a party, but that is very different.

Romesh: So currencies trading would have been the thing to be in?

Martin: Well, let me very careful answering.

Romesh: Sure.

Martin: If you pursue the strategy that we are discussing, which is a variant of the simplest kind, see what a lot of guys do is, they specialize in dollar/kiwi or dollar/pound. The first thing you have to remember is you should never do that. So if you have a dollar that you want to invest in the carry trade, break it up into 10 little packets, to take an example, and do carry trade, US dollar versus nine other currencies or ten other currencies. That has enormously positive effects of diversification. So the folks who you report reading the newspapers who lost just tons: a) we are not doing that and b) they weren't buying options, these puts and calls to protect themselves on the downside.

Romesh: Yeah.

Martin: So you know, some people really took -- if you were doing yen versus certain other currencies, you got killed in August and September, but these more subtle strategies, you are doing pretty well relatively speaking.

Romesh: Great. Martin Eichenbaum, thank you very much.