Join our community of smart investors

How trading fails

The failure of risk-parity trades this years shows us how hard it is to exploit market inefficiencies.
April 23, 2020

A month ago the Vix index – a measure of implied volatility on S&P 500 options which is often called the fear gauge – was at an all-time high. Since then, the S&P 500 has risen almost 20 per cent. You might think this is normal; high risk should mean high return. Instead, it tells us a more interesting story about how financial markets change.

The curious thing is that while you might expect high implied volatility to lead to good returns, the opposite has often been the case. Since 1990 the correlation between the Vix and subsequent monthly changes in the S&P 500 has actually been negative on average. It is lower than average risk, not higher risk, that has tended to lead to good returns. For example, volatility was generally low during most of the S&P 500’s great run up between 2013 and 2019, but high volatility in 2008 led to the market dropping.

Back in 2015 economists such as Andrew Lo, Tyler Muir and Alan Moreira pointed this fact out. They showed that it meant that you could earn good risk-adjusted returns by cutting your equity exposure when volatility was high and raising it when volatility was low – because low volatility was more likely to lead to better returns.

Traders set up strategies designed to do just this. They are called risk-parity trades because they aim to keep a steady level of volatility on their books by changing the proportions of cash and equity as volatility changes, cutting equities when volatility is high, but buying them when it is low.

Such trades, however, have done terribly recently. According to Hedge Fund Research, they have lost more than 12 per cent so far this year. Thanks to this, they’ve made only 2 per cent per year over the past five years. On both measures, you would have been better off simply leaving your money in the S&P 500.

There’s a simple reason for this year’s losses. Volatility was low in February, so such trades were heavily invested in equities just before they crashed. And as volatility fell, they cut equity exposure just before shares bounced back.

 

In truth, though, the link between the Vix and subsequent returns on the S&P 500 was weakening even before the last few weeks. My chart shows this. Each point on the line shows the correlation between the Vix and the subsequent monthly change in the S&P. You can see that this was very negative for most of the period since 1990, implying that risk-parity trades would have worked nicely. After 2015, however, the correlation turned positive. Which means that such trades have done badly.

What we have here is a common problem in the social sciences – that what is rational for one person to do can be self-defeating if enough people do it. For example, if one farmer grazes his sheep on common land, he'll have well-fed sheep. But if all try to do so the land will become barren. Or if everybody at a football match stands up to get a better view, many end up with a worse one.

The same thing is happening here. If enough traders sell shares when volatility rises prices will fall – to a level so low that they’ll subsequently rise. The result is that a high Vix will lead to prices rising. The correlation between the Vix and subsequent equity returns will therefore become positive.

When traders learn about an anomaly – in this case the apparent perverse relationship between risk and returns – they therefore eliminate it by moving prices against themselves. This is common. John Cotter and Niall McGeever at University College Dublin have shown that anomalies that existed in UK equities in the 1990s were subsequently traded away. That echoes a finding for the US market by David McLean and Jeffrey Pontiff (though oddly, the same doesn’t seem to be true for European markets).

This does not, however, mean the history of equity investing is one of a steady approach to the textbook ideal of perfectly efficient markets. As traders lose money, they abandon their strategies and so the anomalies can re-emerge. As Professor Lo has shown, the success of any investment strategy will therefore wax and wane.

A nice example of this is the history of UK small cap stocks. In the 1980s economists showed that these had for years done better than bigger shares. Investors then piled into them. But this drove their prices up so high that a decade of under-performance followed in the 90s. In the late 90s, Elroy Dimson and Paul Marsh pointed this fact out in a paper called “Murphy’s law and market anomalies.” In a great example of that law, small caps then began to outperform again.

All this helps explain why hedge funds perform so badly: HFR estimates that they have on average made only 1.3 per cent per year in the past five years. It is not enough to know that the market has been inefficient. You must have exclusive knowledge of that inefficiency. And for most of us, this is unattainable.