When Bloggers Examine the Treasury Market

This is how the blogosphere glosses the news: first the WSJ runs a slightly overheated article about the most recent Treasury auctions, talking in the headline about “Debt Fears” in the market for US government bonds. It’s one of those spectacularly meaningless headlines onto which all manner of rank speculation can be piled: a market went down, so you can talk about “fear” (although for some reason when a market goes up you get many fewer headlines about “greed”), and yes, it’s a debt market, so you can talk about “debt fear”.

This is not helpful, and Paul Krugman pointed out as much with a handy little chart showing that Treasury bonds have been pretty flat since the end of 2007, but for that sudden flight-to-quality at the height of the crisis. Brad DeLong had much the same post, with a bit more added snark.

Steve Waldman then responded to Krugman with a few more charts of his own, this time looking at the yield curve rather than nominal yields. The basis for doing this is simple:

The US government’s cost of long-term borrowing can be decomposed into a short-term rate plus a term premium which investors demand to cover the interest-rate and inflation risks of holding long-term bonds. The short-term rate is substantially a function of monetary policy: the Federal Reserve sets an overnight rate that very short-term Treasury rates must generally follow. Since the Federal Reserve has reduced its policy rate to historic lows, the short-term anchor of Treasury borrowing costs has mechanically fallen. But this drop is a function of monetary policy only. It tells us nothing about the market’s concern or lack thereof with the risks of holding Treasuries.

This is true, but at the same time it feels naive to me: I think it’s fair to say that the historical connection between the Fed funds rate and Treasury bond yields is largely lost when the overnight rate is at or near zero. Indeed, that was one of Alan Greenspan’s big mistakes: He dropped rates so far that he lost control of long-term interest rates. If the Fed funds rate is at 5%, you can turn the steering wheel and see an effect at the long end of the curve. If it’s at 1%, that mechanism becomes much squishier.

James Hamilton adds a useful chart of his own, showing that TIPS have been rising in yield even more dramatically than Treasury bonds. Yields on TIPS are real, not nominal, remember, so that chart alone does a lot to debunk any notion that an increase in bond yields is related to an increase in expected future inflation.

Ryan Avent, attempting to adjudicate, can come up with little more than to conclude that “the data bears watching closely, but that’s about the most one can say for now”. I wouldn’t even go that far. If you stare long enough at bond data, you can conclude just about anything you like — you can even start kidding yourself that you’re looking at the market pricing in default risk on Treasuries. My feeling is, looking at Waldman’s charts, that it might be time for the curve-steepening we’ve seen over the past 18 months to start coming to an end and reverting to the mean. But that said, concluding anything much from looking at lines on charts is always a fool’s game, unless your conclusions are independently derived from much harder empirical analysis.