Going against your gut? Volatility in equities
Lately, it has been feeling as though individual stocks are moving about more than usual – the dispersion of returns between different stocks has been high. Given recent geopolitical events and financial results from several companies, perhaps this is not surprising. Yet, as we often comment in these pages, feelings about financial markets are often at odds with the facts.
As a reality check, we have analysed the monthly returns of single stocks over the past several years to see whether the differences in returns over the past few months have been especially high. More specifically, we have calculated the variance (a statistical measure of dispersion) of the one-month total returns for both US and UK markets (figure 1).
In the US market, variance in 2016 has been fairly mild by historical standards. After a slight uptick towards the end of 2015 and into the start of 2016, dispersion quickly trailed off again. It seems that our feelings had been getting the better of us. Now, though, since Donald Trump’s victory in the race for the White House, variance in November is showing signs of picking up again. Our feelings, this time, were perhaps prescient.
Looking at the UK, we can see a similar story but with the notable exception of the month of June, where variance was abnormally high. Given the vote to the leave the EU (“Brexit”), this makes sense. In the months since the vote, the UK market has settled down somewhat, but – like the US market – it is now heating up again.
Is this the beginning of something more or merely short-term reactions to political events? Time will tell, but perhaps there is something we can learn from changes in levels of dispersion. If dispersion becomes abnormally high, for example, does that presage lower returns from equity markets?
To answer this question, we can look back through history and see how levels of dispersion have tended to correlate with the return from the market over the following months. For our analysis, we have divided history into quintiles (five roughly equally sized buckets) based on lowest to highest levels of one-month variance in stock returns.
We then calculated the average return of the wider market over the following three months. Essentially, we wanted to see whether higher levels of variability in individual stock returns were a sign that markets were about to perform somewhat poorly (see table in figure 1).
As it turns out, higher levels of dispersion do appear to show subtle signs of predicting lower returns from the market over subsequent months. The table in figure 1 shows average three-month returns from the US and UK markets by quintile of preceding variance. In the US, for example, when dispersion has been at its highest, returns over the subsequent months averaged just 1.0%, compared to nearer 3% after periods of low dispersion.
Admittedly, the relationship does not appear to be strong, and many factors other than dispersion influence future returns from markets. Still, should dispersion continue to pick up, it might be time to reduce our allocations to equities.
Past performance does not guarantee future performance.