Join our community of smart investors

On positive feedback

Share price falls sometimes feed on themselves, but only in the short term.
May 30, 2019

To what extent are stock markets automatically stabilising? This question gains force from the fact that the US yield curve has inverted again, with 10-year Treasury yields falling below three-month money rates.

This raises the danger of a positive feedback loop. The inverted yield curve might cause investors to fear recession, which could cause them to try to shift out of equities and into bonds with the result that the yield curve inverts even more which, could in turn, intensify fears of recession and so trigger even more selling of equities. We saw a glimpse of just this process last week.

This, however, is by no means the only way in which share price falls can feed upon themselves and generate further falls.

Another mechanism occurs through risk-parity trades. These are based on the idea, pointed out by the Massachusetts Institute of Technology’s Andrew Lo and by Tyler Muir at UCLA and Alan Moreira at the University of Rochester, that there is no systematic correlation between market volatility and subsequent returns. This means you can earn above-average risk-adjusted returns by selling equities when volatility is high. Because share price falls are often accompanied by higher volatility, such trades can cause investors to sell after prices have fallen, driving prices down even more.

Yet another cause of positive feedback loops are information cascades. People don’t act only upon their own information. They also take their cues from what others seem to believe: you’d have to be amazingly arrogant to think you know it all and others know nothing. Other investors’ pessimism can therefore cause us to sell. Christopher Carroll at Johns Hopkins University has shown that economic expectations spread in the same way that a disease does, so we can be infected by others’ pessimism. Tony Yates, an economic consultant, says that economic ideas are like memes: some can spread rapidly. Hans Hvide at the University of Aberdeen and Per Ostberg at Zurich University, and Tilburg University’s Ben Jacobsen, have shown that investment decisions are influenced by word-of-mouth and peer effects. And Andrei Shleifer and Brock Mendel at Harvard University show that investors sometimes “chase noise”: they trade in the mistaken belief that others’ trades are well informed, which can cause prices to rise or fall too much.

A variation on the information cascade is the Keynesian beauty contest. Back in 1936 economist Maynard Keynes noted that trading is about trying to predict others’ opinions: the profits go to those who buy just before others turn bullish and sell just before they turn bearish. If small falls in prices cause traders to bet that other traders will soon become more pessimistic, they will sell with the result that price falls will trigger more falls.

All these mechanisms can be amplified by forced selling. Leveraged funds sometimes need to raise cash quickly, so they liquidate positions thus causing dips in prices to become slumps.

In fact, the mere possibility that these processes will operate can become self-fulfilling. If enough investors merely fear that a positive feedback loop is possible some will sell and those who might have bought on dips will stay on the sidelines. This will make such loops more likely.

Just how common are they though? Most of you will have bought a share believing it to be cheap only to see it fall further. Such experiences don’t always mean your assessment of the company was wrong: they can instead mean you were the victim of such positive feedback.

Three big facts tell us that such feedback loops have been significant for the aggregate market:

  • As Yale University’s Robert Shiller pointed out back in 1981, share prices are many times more volatile than dividends.
  • We often see spikes in volatility, as measured by the Chicago Board Options Exchange’s (CBOE) Vix index. Markets can be stable for some time, only for volatility to shoot up suddenly when small price falls trigger large ones.
  • There’s sufficient momentum in stock markets for the 10-month average rule proposed by Cambria Investment Management’s Meb Faber to work on average: it pays investors to sell when prices dip below their 10-month average, because such falls can lead to further falls. Investors who use this rule, of course, themselves contribute to the positive feedback loop.

But there’s another big fact here. This is that the dividend yield on the All-Share index has in the past been a fantastic predictor of longer-term returns: since 1986 the correlation between the yield and subsequent five-year total returns has been a whopping 0.82. This tells us that in the long run equity markets have been stabilising: cheap markets have recovered. Yes, falls can lead to falls in the short term, but in the longer term they have led to rises because eventually buyers do step in. The problem is, though, that we can never tell for sure when they will do so.