This could matter a lot. "The slope of the yield curve is a reliable predictor of future real economic activity," say economists at the New York Fed. On the last three occasions when the curve became inverted, with 10-year yields falling below two-year ones, recessions followed.
There's a simple reason for this. To see it, just ask why anyone would want to buy (say) a 10-year bond if a two-year one yields more. The answer is that they would do so if they expected short-term yields to fall, so that when the two-year bond matures they will face low returns on their reinvestment. But, of course, short-term yields would fall if the economy is weak.
It's not surprise, therefore, that the yield curve should predict economic activity. What is a surprise, though, is that for years this has not been sufficiently well known. In 2008, economists at the San Francisco Fed pointed out that the yield curve did a better of forecasting the economy than economists. The fact that the recession of that year (which followed an inverted yield curve) took many economists by surprise only strengthened their point.
If you think this just shows the folly of economic forecasting, think again. Equity investors haven't heeded the message of the yield curve, either. The last two yield curve inversions, in 2000 and 2007, led to slumps in share prices. If investors had heeded the message of the yield curve, they would have saved themselves trillions of dollars.
You might think, therefore, that the recent flattening of the yield curve is worrying.
It is a concern, but not yet a pressing one. Much of the predictive power of the curve comes when it is inverted - because recessions and bear markets then follow. When the curve is upward sloping, as it still is, movements in it are less worrying.
Some simple statistics tell us this. Since 1987 the correlation between the gap between 10- and two-year yields and subsequent two-yearly changes in industrial production has been 0.3. But if we look only at periods when 10-year yields have been above two-year ones, this correlation has been 0.16. And while the correlation between the yield gap and subsequent two-yearly returns on equities has been 0.16 for the whole post-1987 sample, it has been minus 0.05 at times when the yield curve has sloped upwards.
The recent flattening of the curve is therefore not telling us much. But this could change. If long-dated yields stay around their current levels as shorter-dated ones rise next year then the curve could become inverted. That would be the time to expect recession and sell equities.
Such an inversion might seem a distant possibility. I mention it to point out that many investors are worrying about the wrong thing. They're worrying that there's a 'bond bubble', which could burst when QE ends. From the point of view of equity investors, however, higher bond yields are to be welcomed because they would be a sign of a reduced chance of recession and bear market. In this sense, what should worry equity investors is not a sell-off of bonds but the lack of one.
