Some investors are worried about the possibility that the US yield curve might invert – that longer-dated yields could fall below shorter-dated ones. They are right to worry.
My chart shows why. It shows one measure of the yield curve (the gap between 10-year Treasury yields and the Fed funds rate) and annual US industrial production growth. This looks messy. But look at what happens when 10-year yields are below the Fed funds rate. We've seen this four times in the last 30 years. On three of those occasions, a recession followed.
Yes, the time lag is variable. And it’s not the case that more inverted yield curves lead to worse recessions; the curve was less inverted in 2007 than it was in 1989 or 2000 but the following recession was far nastier.
Nevertheless, there is predictability here. Let’s consider only end-month data from January 1987 to April 2015. This gives us 340 data points. Of these, we’ve had 290 upward-sloping curves on my definition and 50 inverted ones. After the 290 upward-sloping occasions industrial production grew by an average of 2.4 per cent per year in the following three years. After the 50 inversions, however, it fell by an average of 1 per cent per year in the following three years.
The yield curve, then, predicts recessions. Given that economists have been terrible at forecasting these – as the IMF’s Prakash Loungani has pointed out for years – this means the curve sends a useful message.
There’s a simple reason for this. When long-dated yields are below shorter-dated ones, it’s because investors expect short-term interest rates to fall: for example, in early 2007 they held 10-year bonds on yields of below 5 per cent rather than cash on yields of more than 5 per cent because they thought cash returns would fall. But why might they expect short-term rates to fall? It’s because they fear a recession. In this sense, the yield curve captures the dispersed wisdom of crowds.
Luckily, there’s little chance of the yield curve inverting very soon, in the sense of 10-year yields falling below the Fed funds rate. But it could do so next year. Futures markets are pricing in a fed funds rate next December of 2.7 per cent. It would take only a moderate rally in 10-year bonds to then invert the curve.
Even the most parochial of equity investors should worry about this simply because if the yield curve predicts a recession it also predicts falling share prices.
In those 50 months when the US yield curve was inverted, the All-Share index fell by an average of 9.1 per cent in the following three years. It posted some sort of fall on 36 of those occasions – 72 per cent of the time. By contrast, in the 290 months when the curve was upward sloping the market rose by an average of 22.8 per cent and fell over the three years in only 15.5 per cent of cases. This means a fall in the index is 4.6 times more likely in the three years following an inverted yield curve than it is in the three following an upward-sloping curve.
As you’d expect given that yield curves predict recessions, it is cyclical stocks that are especially sensitive to inversions of the yield curve. To take a few examples, the FTSE construction sector rises by an average of 25.7 per cent in three years following an upward-sloping yield curve but falls by an average of 11.5 per cent in the three following an inverted curve. Transport stocks rise by an average of 25.7 per cent after upward slopes and fall by an average of 19.1 per cent after inverted ones. General retailers rise by 17.7 per cent on average after upward-sloping curves but fall by 4 per cent on average after inverted ones. And banks rise by an average of 30.3 per cent after upward slopes but fall by an average of 9.2 per cent after inversions.
Herein, though, lies both a comfort and a concern for investors.
The comfort is that the Federal Reserve should now be wise to the fact that yield curves predict recessions. This should stop it raising the Fed funds rate if doing so would cause an inversion of the curve.
I don’t say this as an idle hope. Back in 2007 John Williams co-wrote a paper in which he showed that “forecasters have generally placed too little weight on yield curve information when projecting declines in the aggregate economy”. He wrote that when the yield curve was inverted, and the fact that the economy slumped a few months later proved him right. He now sits on the FOMC and so helps set interest rates.
The concern, however, is that equity investors should know this as well. In the past, they largely ignored the yield curve, which meant that share prices stayed too high when the curve was inverted and only fell as the reality of recession hit them. Now, however, investors are unlikely to repeat this error. Instead, any hint that the yield curve might invert should cause them to sell shares generally and especially cyclical ones long before any recession. This means that we might well not be able to wait until the yield curve actually inverts before we rotate out of cyclicals. Which means that such stocks are riskier now, and stock-picking is harder.
