America's bond market is upside down. Is the economy about to capsize as well?
The long and the short of it

SHORTLY before America's last recession, which began in March 2001, something odd happened to interest rates. Short-term rates rose above long. The same thing happened before the recessions of 1990, 1981, 1980, 1973, 1969 and 1960. A dark omen, then, but why worry about it now? In recent months, yields on short-term securities have crept up on those offered by longer-dated instruments. In the last week of December, it was (slightly) cheaper for the American government to borrow for ten years than for two.

This is unusual. The government borrows by selling a variety of IOUs, which promise to give the buyer his money back sooner (three-month bills, for example) or later (eg, ten-year Treasuries). Normally, the longer the maturity, the higher the yield a security must offer: the “yield curve” slopes upwards. Markets take this to be the natural state of affairs (though just why it should be so has taxed some of the best economists).

When things are upended, the yield curve is said to be “inverted”, a condition now exciting much chatter among analysts. Despite all this talk, the yield curve is not yet inverted across its full length. The yield on two-year Treasuries may have risen above that on ten-year bonds, but the rate on three-month bills still falls short by about 0.4 percentage points. The spread between ten-year and three-month securities has been this narrow twice before (in 1998 and 1995) without a recession ensuing. Nonetheless, the ironing-out of the yield curve is not normally welcome news. According to a statistical model estimated by Arturo Estrella*, an economist at the Federal Reserve Bank of New York, a spread of 0.4 points, averaged over a month, has historically signalled an 18% chance of recession within a year.

What gives the yield curve its predictive power? Long-term rates represent, in part, the market's expectations for future short-term rates. To see why, consider an investor who wants to lend for ten years. He could sink his money into a ten-year bond for the duration of its life. Alternatively, he could buy a five-year bond today, rolling his money over into another in five years' time. Suppose the five-year rate is 5% now, but the investor expects it to rise to 10% in five years' time. In that case, ten-year bonds must offer a yield of about 7.5% today to attract his money. On the other hand, if the investor expects five-year rates to fall to just 3% in five years' time, he will accept a ten-year yield of only about 4% today. In this case, long rates will fall below short now, in anticipation of even lower short rates later.

An inverted yield curve, then, suggests that short-term rates are higher today than they will be in the future. But why should this necessarily spell recession? Normally, it is because the Federal Reserve is in the midst of a campaign against inflation. To win this battle, short-term rates are sometimes raised high enough to induce a recession, which squeezes inflation out of the system. In due course, lower inflation will pave the way for lower short-term rates. But before this happens, long-term bond yields fall in anticipation of the future victory. In this case, an inverted yield curve is just a measure of the Fed's power.

Alternatively, inversions may be a measure of the Fed's ignorance. The bond market may know something the central bankers don't. Long-term rates may be subdued, because the market anticipates a recession that will eventually force the Fed to loosen monetary policy. But short-term rates remain high, because the Fed has yet to act on what the bond market foresees.

A portent and a puzzle

So does the flatness in today's yield curve mean that the Fed is trying to engineer a recession? Hardly. The Fed's rate-setting committee no longer describes its monetary stance as “accommodative”, but neither is it trying to cage a runaway economy. At 4.25%, its key rate is still much lower than the rate of growth in nominal GDP (more than 7%, annualised, in the third quarter), which serves as one crude measure of policy's tightness. According to the minutes of its December meeting, released this week, some members of the committee reckon that the federal funds rate is probably within a neutral range, one that should allow the economy to grow at close to its full potential.

In contrast to previous inversions, the yield curve is flat not because short rates are unusually high, but because long rates are unusually low. Yields on ten-year Treasuries have hovered around 4-4.5%, even as the Fed has hoisted short-term rates 13 times. Alan Greenspan, the Fed's chairman, himself does not fully understand why this is so—no doubt it has much to do with foreign purchases of long-dated American securities by oil producers and Asian central banks. Nonetheless, on this reading, the bond market offers a puzzling “conundrum”, as Mr Greenspan has put it, not a worrying omen. Optimists find comforting parallels in the events of 1966. In the last few months of that year, the interest rate on three-month bills edged above that on ten-year bonds, but no recession followed—the only time a fully inverted yield curve has cried wolf. Then, as now, long-term rates were unusually low, averaging under 5%.

The pessimists, however, look back five years, not 40. In the second half of 2000 the yield curve inverted, and then, as now, the vast majority of commentators dismissed it, arguing that the old portent had nothing to say about the new economy. Three months into 2001, the economy slipped into recession.

Although monetary policy may not be that tight as yet, the economy's strength may rely, more than most realise, on interest rates remaining low. To a disturbing degree, America's economy is still debt-led. Can this borrowing continue to drive growth now that interest rates are no longer “accommodative” and house prices are starting to cool? The answer is not easy to find in the bond market. But it will decide whether America's economy is as flat in the year ahead as the yield curve is today.

* “The Yield Curve as a Leading Indicator: Frequently Asked Questions”. October 2005. Available at  www.newyorkfed.org/research/capital_markets/ycfaq.pdf