It’s been a bad year for UK equities, with the All-Share index falling more than 20 per cent. You’d expect that in this environment defensive sectors would have done relatively well.
And you’d be wrong. My chart shows the point. It plots the performance of 25 main FTSE sectors this year against their betas, based on monthly returns in the five years to December 2019. It’s clear that there is no correlation between the two. Yes, the high-beta oil sector has had a horrible year. But two other high-beta sectors, mining and IT, have done relatively well – better in fact than low-beta sectors such as beverages, life insurance, utilities and healthcare.
Betas, then, have been utterly useless as a guide to this year’s returns.
You might think this is no surprise. We’ve known for years that the capital asset pricing model (CAPM) – the theory that predicts that low-beta stocks should outperform a falling market – is wrong.
But it is wrong because low-beta stocks on average do better than they should. This fact also predicts that we should have seen low-beta sectors do relatively well this year.
Which is exactly what we haven’t seen.
Why?
Over short periods, betas can and do change because the market falls for different reasons. A bear market caused by a recession will produce different winners and losers from one caused by stocks having been overvalued. And one caused by a financial crisis will have different winners and losers again. For this reason, past betas will always be an imperfect guide to future returns, even in a bear market, because they are an imperfect guide to future betas.
This is especially true now. The pandemic has scrambled up past patterns of behaviour. Banks and oil stocks have slumped as investors have been shocked by dividend cuts. Food retailers and IT have held up as investors believe these will benefit from more people working from home. And miners have done okay as investors hope that the first country into the pandemic, China, will be the first to see a decent recovery.
The winners and losers from this scrambling up have been largely unrelated to past betas. Hence the lack of correlation between those betas and returns.
Does this mean that the concept of beta is useless?
No. One of the best-attested facts in finance is that over the long run low-beta stocks do better than they should. This is true for different times and places (and even different assets) and is robust to different ways of measuring beta. And, in fact, it has been true in one sense this year: my no-thought portfolio of low-beta stocks has hugely outperformed the market this year. Extremely low-beta shares have done as they should. It is more normal betas that have been no use to investors.
Indeed, it’s possible that this year’s events explain the puzzle of why low-beta shares have traditionally done so well. Such stocks are vulnerable to tail risk – the danger that extreme, rare events will increase uncertainty about previously stable companies, perhaps by forcing them to cut dividends. This would cause them to cease to be low-beta stocks. The threat of such events means that a low past beta is no guarantee that a stock will remain relatively safe in future. Fearing this, investors might reasonably underweight such shares, even though they know that low-beta stocks generally do well. In this sense, good returns on low-beta shares are compensation for a form of tail risk. This might not explain all of the outperformance of low-beta stocks – but perhaps some of it.
But what about high-beta stocks? If we the market does recover, might these outperform as the CAPM predicts?
Not necessarily. One reason for the All-Share’s woes this year is that two huge sectors – banks and oil – have done terribly. If these recover their weight in the index is so great that we might see most sectors underperform the market simply as the latter will be pushed up by a minority of stocks.
If this happens, it might be some time before we see a return to normal market conditions.
