The Destination Currency in the Global Carry Trade: USD?
If you are an investor of the Ben Graham school, you’ve lived your life looking for “value” investments with a “margin of safety.” Periodically, if you are a pure value investor, then you go through long periods of pulling your hair out when momentum rules the day, even if you believe – as GMO’s Ben Inker eloquently stated in last month’s letter – that in the long run, no factor is as important to investment returns as valuation.
This is one of those times. Stocks have been egregiously overvalued (using the Shiller CAPE, or Tobin’s Q, or any of a dozen other traditional value metrics) for a very long time now. Ten-year Treasuries are at 1.80% in an environment where median inflation is at 2.5% and rising, and where the Fed’s target for inflation is above the long-term nominal yield. TIPS yields are significantly better, but 10-year real yields at 0.23% won’t make you rich. Commodities are very cheap, but that’s just a bubble in the other direction. The bottom line is that the last few years have not been a great time to be purely a value investor. The value investor laments “why?”, and tries to incorporate some momentum metrics into his or her approach, to at least avoid the value traps.
Well, here is one reason why: the US is the destination currency in the global carry trade.
A “carry trade” is one in which regular returns can be earned simply on the difference in yields between different instruments. If I can borrow at LIBOR flat and lend at LIBOR+2%, I am in a carry trade. Carry trades that are riskless and result from one’s market position (e.g., if I am a bank and I can borrow from 5-year CD customers at 0.5% and invest in 5-year Treasuries at 1.35%) are usually more like accrual trades, and are not what we are talking about here. We are talking about positions that imply some risk, even if it is believed to be small. For example, because we are pretty sure that the Fed will not tighten aggressively any time soon, we could simply buy 2-year Treasuries at 0.88% and borrow the money in overnight repo markets at 0.40% and earn 48bps per year for two years. This will work unless overnight interest rates rise appreciably above 88bps.
We all know that carry trades can be terribly dangerous. Carry trades are implicit short-option bets where you make a little money a lot of the time, and then get run over with some (unknown) frequency and lose a lot of money occasionally. But they are seductive bets since we all like to think we will see the train coming and leap free just in time. There’s a reason these bets exist – someone wants the other side, after all.
Carry trades in currency-land are some of the most common and most curious of all. If I borrow money for three years in Japan and lend it in Brazil, then I expect to make a huge interest spread. Of course, though, this is entirely reflected in the 3-year forward rate between yen and real, which is set precisely in this way (covered-interest arbitrage, it is called). So, to make money on the Yen/Real carry bet, you need to carry the trade and reverse the exchange rate bet at the end. If the Real has appreciated, or has been stable, or has declined only a little, then you “won” the carry trade. But all you really did was bet against the forward exchange rate. Still, lots and lots of investors make precisely this sort of bet: borrowing money is low-interest rate currencies, investing in high-interest-rate currencies, and betting that the latter currency will at least not decline very much.
How does this get back to the value question?
Over the last several years, the US interest rate advantage relative to Europe and Japan has grown. This should mean that the dollar is expected to weaken going forward, so that someone who borrows in Euro to invest in the US ought to expect to lose on the future exchange rate when they cash out their dollars. And indeed, as the interest rate advantage has widened so has the steepness of the forward points curve that expresses this relationship. But, because investors like to go to higher-yielding currencies, the dollar in fact has strengthened.
This flow is a lot like what happens to people on a ship that has foundered on rocks. Someone lowers a lifeboat, which looks like a great deal. So people begin to pour into the lifeboat, and they keep doing so until it ceases, suddenly, to be a good deal. Then all of those people start to wish they had stayed on the ship and waited for help.
In any event, back to value: the chart below (source: Bloomberg) shows the difference between the 10-year US$ Libor swap rate minus the 10-year Euribor swap rate, in white and plotted in percentage terms on the right-hand scale. The yellow line is the S&P 500, and is plotted on the left-hand scale. Notice anything interesting?
The next chart shows a longer time scale. You can see that this is not a phenomenon unique to the last few years.
Yes, the correlation isn’t perfect but to me, it’s striking. And we can probably do better. After all, the chart above is just showing the level of equity prices, not whether they are overvalued or undervalued, and my thesis is that the fact that the US is the high-yielding currency in the carry trade causes the angst for value investors. We can show this by looking at the interest rate spread as above, but this time against a measure of valuation. I’ve chosen, for simplicity, the Shiller Cyclically-Adjusted P/E (CAPE) (Source: http://www.econ.yale.edu/~shiller/data.htm)
Now, I should take pains to point out that I have not proven any causality here. It may turn out, in fact, that the causality runs the other way: overheated markets lead to tight US monetary policy that causes the interest rate spread to widen. I am skeptical of that, because I can’t recall many episodes in the last couple of decades where frothy markets led to tight monetary policy, but the point is that this chart is only suggestive of a relationship, not indicative of it. Still, it is highly suggestive!
The implication, if there is a causal relationship here, is interesting. It suggests that we need not fear these levels of valuation, as long as interest rates continue to suggest that the US is a good place to keep your money (that is, as long as you aren’t afraid of the dollar weakening). That, in turn, suggests that we ought to keep an eye on rates of change: if the ECB tightens more, or eases less, than is priced into European markets (which seems unlikely), or the Fed tightens less, or eases more, than is priced into US markets (which seems more likely, but not super likely since not much is presently priced in), or the dollar trend changes clearly. When one of those things happens, it will be a sign that not only are the future returns to equities looking unrewarding, but the more immediate returns as well.