It’s generally thought that this year’s recession and fall in share prices came as a bolt from the blue – the result of a genuinely unforeseeable pandemic. But this isn’t entirely true. Last year, one indicator was warning of trouble to come – and it is still sending a warning sign.
I’m speaking of the yield curve. Last year, in both the US and UK, 10-year government bond yields fell below three-month money rates. In both countries, this had in the past been a harbinger of recession. In the UK, inversions of the yield curve in 1973, 1979, 1988 and 2006 all led to recessions a few months later. Indeed, as former MPC member Charles Goodhart and colleagues pointed out last year, the curve’s predictive power dates back to the 19th century. For this reason, Arturo Estrella – the former economist at the New York Fed who was one of the first to document the link between inverted yield curves and recessions – said last August that the chance of recession was “pretty high”. And the Federal Reserve cut interest rates last year in part because the yield curve was signalling downside risks to economic activity.
And because share prices fall in recessions, the yield curve has also predicted bear markets. For example, since 1985 there have been 174 months that ended with 10-year yields being below three-month money rates. On 67 of these occasions, the All-Share index fell in the following three years. But after the 214 months when 10-year yields were above three-month money, the index fell only 19 times. A fall in share prices over a three-year period is therefore more than four times as likely after the yield curve has been inverted than after it has been upward sloping.
There’s a good reason for this. To see it, consider why anyone would want to own a bond when cash pays a higher return. It’s because they expect short-term interest rates to fall, and so want to lock in a longer-term rate. And why would they expect rates to fall? Because they expect the economy to hit hard times. In this way, the yield curve embodies the wisdom of crowds – it forecasts recessions far better than economists do.
Both history and theory, therefore, told us last year that the yield curve was a warning sign.
But, but, but. There’s an obvious problem here. While the gilt market might be a good economic forecaster, it seems absurd to give it credit for predicting a pandemic and its inevitable consequence, recession. Surely, then, in this case the yield curve just got lucky.
Or did it? As the philosopher Michael Polanyi pointed out, we all have tacit knowledge – inarticulable hunches and gut feels. An inverted yield curve might be an aggregation of these – of senses of foreboding that something might go wrong. Even if people had no idea of where trouble would come from, they might still have had a feeling of our vulnerability. If so, the yield curve was telling us something useful.
Which is a big problem, because the message it is sending now is a worrying one. Ten-year gilt yields, at under 0.2 per cent, are below three-month money rates. Which tells us there’s a heightened risk of falling share prices and economic weakness. The gilt market seems to have no confidence in the possibility of a V-shaped recovery.
You might think this inversion is artificial, the result of the Bank of England buying gilts. I’m not so sure. Recent auctions of gilts have been heavily oversubscribed, with bids of more than twice the amount of gilts being sold. There’s strong demand for gilts from the private sector, as well as the Bank.
Instead, there might be two other reasons to doubt the message of the yield curve.
One is that yields might be depressed not so much by our economic prospects but also by fears that other investors will turn gloomy – investors might be scared that others will be scared. In other words, yields might be low as the result of an information cascade, of the blind leading the blind. This means that one of the key conditions for the wisdom of crowds to operate – that people’s beliefs be uncorrelated – does not hold.
Another is simply that we are in uncharted waters. Our situation is without precedent. This means we can’t be at all sure which past rules still work and which don’t.
On the other hand, though, last year’s experience – not to mention that of 2006-07 as well – warns us not to dismiss the message of yield curves.
For me, all this warns us of the need for caution. Personally, I had quite a high cash weighting before the crisis, and I will keep it.
