A Lot Can Happen In 10 Years

A little debate about interest rates took place this week. Here's Krugman in the NYTimes:

Well, let’s compare with Germany, also perceived as almost completely safe — but paying much lower interest. Why?

The crucial point here is that German bonds are denominated in euros, while U.S. bonds are denominated in dollars. And what that means in turn is that higher U.S. rates don’t reflect fear of default; they reflect the expectation that the dollar will fall against the euro over the decade ahead.

But why should we expect a falling dollar vis-a-vis the euro? One big reason is that European inflation is very low and falling, while the U.S. seems to be holding near (although below) its 2 percent target. And other things equal, higher inflation should translate into a falling currency, just to keep competitiveness unchanged. If you look at the expected inflation implied by yields on inflation-protected bonds relative to ordinary bonds, they seem to imply roughly 1.8 percent inflation in the US over the next decade versus half that in the euro area, which means that the inflation differential explains about 60 percent of the interest rate differential.

Beyond that, there is good reason to expect the dollar to fall in real terms over the medium term. Why? The relative strength of the US economy has led to a perception that the Fed will raise rates much sooner than the ECB, which makes dollar assets attractive — and as Rudi Dornbusch explained long ago, what that does is cause your currency to rise until people expect it to fall in the future. The dollar is strong right now because the U.S. economy is doing better than the euro area, and this very strength means that investors expect the dollar to fall in the future.

Matthew Klein at FT Alphaville offers a rebuttal in The dubious relationship between yields and exchange rates, concluding:

Just as the difference between German and American sovereign borrowing costs tell us nothing about default risk, neither do they tell us anything meaningful about future exchange rates, much less market expectations of future exchange rates.

Unfortunately for Krugman's argument, market participants don't buy 10-year bonds expecting to hold them for 10-years. Rather, as 2008 showed, maturity mismatch means banks and investors borrow short and lend long. Their long-term bond holdings are rentals that may be sold at any time, including under extreme duress. Alternatively, sovereign bond yields might be telling us something about default risk in the corporate bond market. OR the slide in rates may be an anticipation of ECB QE, an expectation that German bond yields are turning Japanese.

Finally, even if Krugman was correct about the reason for the difference in interest rates, it still tells us nothing other than relative exchange rates. By the end of 10-years, an investor who buys a 10-year German bond yield will come out ahead of a 10-year U.S. Treasury holder. If the U.S. dollar is starting a 5 to 6 year bull market and investors in the 10-year government bond market are taking into account long-term currency trends along with long-term exchange rates, they are forecasting a period of extreme volatility.