Avoiding momentum crashes

In this week's Stockopedia Strategy Map article, Ed introduced the concept of Momentum. Momentum is the tendency for medium-term winners to keep doing well. A typical Momentum strategy looks at the share price performance from 12 months to one month ago. (In the very short term, share prices tend to mean revert rather than trend, so the last month is ignored.) A market-neutral strategy buys the top performers and shorts the worst performers, but a Momentum strategy can be implemented successfully on the long side only. Momentum is pervasive and has been shown to work across multiple time periods, geographies and asset classes.

It is generally accepted that Momentum generates excess returns because investors under-react to fundamental changes in company performance. This may be due to a psychological bias where investors dislike buying something they could have bought more cheaply earlier. I certainly suffer from this bias. However, the momentum effect may also be due to the under-reaction of company management to positive changes in their business performance.

These effects are relatively short-lived. As this chart from a study called Factors from Scratch shows, high momentum stocks tend to exhibit higher than average EPS growth over the next year:

However, that the EPS growth ends up significantly below market averages after a year. Momentum strategies require regular rebalancing into the most recent winners.

In academic studies, Momentum has been shown to generate higher risk-adjusted returns than either Value or Quality, making it a key component of the Stockopedia StockRank. Unfortunately, there is a trade-off for this superior performance. The strategy undergoes periods of very rapid underperformance, so rapid they are known as Momentum Crashes.

Momentum Crashes

Momentum crashes tend to occur at the end of a bear market, when trends reverse rapidly, catching out those who rely on the recent price action to make investment decisions. As the 2022 study Isolating Momentum Crashes by Maik Dierkes and Jan Krupski, notes:

Since 1926, there have been several momentum crashes that feature short but persistent periods of highly negative returns. From June to August 1932 the momentum portfolio lost about 91%, followed by a second drawdown in April to July 1933. Another prominent crash took place in 2009 when Momentum lost more than 73% within a period of three months.

A 2016 paper called Momentum Crashes by Kent Daniel and Tobias J. Moskowitz listed the most significant monthly percentage drawdowns (WMLt) for a momentum strategy:

These types of rapid reversals are far too severe for most investors to bear without abandoning a strategy. Anyone in any doubt about this should imagine their current portfolio reducing to a quarter of its value in a month and needing to deliver a 4x return just to break even. While recent difficult markets have presented challenges for almost all types of investors, this drawdown is on another level.

While not as rapid or deep as the 2009 fall, the Stockopedia pure long-only UK momentum screen has underperformed the FTSE All share by 40% over the last two years, causing significant pain for those who follow this strategy:

This shows that while momentum is a factor that all investors should probably consider in their investing decisions, it is certainly not a panacea for outperforming the market. If investors could avoid the momentum crashes, however, it may make it a much easier strategy to follow.

How to avoid Momentum Crashes

Be a long-only investor

The momentum factor detailed in academic studies goes long stocks with the strongest medium-term price increase, and shorts stocks with the largest medium-term price decrease. A long-only momentum strategy that doesn't short the losers delivers almost all of the long-term returns with added simplicity. A long-only strategy may also avoid some of the worst drawdowns of a momentum strategy, which typically occur as the general market trend reverses. For example, those who were short losing stocks when the market bottomed in March 2009 faced a period of significant losses since the weakest stocks rebounded most strongly.

Of course, this means that the momentum strategy is no longer market-neutral and may face drawdowns with the general market. However, these types of drawdowns may be easier to manage, either through hedging or just psychologically (the losses will occur when all other investors are also facing drawdowns).

Momentum Strategy Timing

It turns out that the returns to the momentum factor vary significantly over time, as momentum crashes tend to occur in certain types of markets. Typically, at the end of bear markets. Of course, if one could accurately predict the start and end of bear markets, investors would have little need for any other strategy. However, some recent studies have identified some specific factors that help avoid momentum crashes.

For example, in a 2015 study called Momentum has its Moments, Pedro Barroso and Pedro Santa-Clara found a negative relationship between momentum volatility and subsequent returns—momentum volatility increases in times of momentum crashes. The authors suggest a strategy of volatility targeting to improve the Sharpe Ratio. This isn't an implementable strategy for the individual investor. However, if a momentum investor notices that the daily movements in their portfolio returns appear unusually large, this may be a time for them to step aside for a while.

In the Daniel and Moskowitz study, the full table of the worst monthly momentum performances also includes details of the market conditions in the run-up to the momentum crash. The fourth column shows the market returns versus two years previously. The fifth column shows the market return in the month that the momentum crash occurs. Almost exclusively, they show that the market was weak in the run-up to the crash but rebounded strongly in the month of the crash:

The authors developed a regression that predicted the returns to a momentum strategy based on a bear market indicator and the variance of the daily returns of the market over the 126 days prior. Times of high market volatility and being in a bear market were negative for the returns to momentum. When market volatility increases in a bear market, the chances of a momentum crash increase significantly.

The third notable paper is the one from Dierkes and Krupskimes, who developed a crash indicator based on three factors:

• Their bear market indicator is based on if the two-year market return preceding the start of the month being negative.
• A rebound indicator is based on markets displaying positive one-month returns.
• Finally, if markets had already witnessed negative returns to momentum in a bear market which had rebounded, the risk of a momentum crash is high.

Again, if a momentum investor starts to see negative returns to their strategy during a rebound in a bear market, they should exit a momentum strategy with haste.

Finally, a recent paper found a link between Momentum crashes and 52-week high effects. The authors, Suk-Joon Byun and Byounghyun Jeon, say:

We find that a surge of investor speculation toward stocks far from their 52-week highs can partially explain the momentum crashes. If a momentum strategy is revised to be neutral on a 52-week high effect, momentum crashes are significantly attenuated and the revised strategy does not exhibit procyclical returns. Furthermore, the revised strategy generates a higher Sharpe ratio in different sub-periods and international stock markets.

This means that a momentum strategy may underperform when the strategy of buying stocks at 52-week highs underperforms. Again, the root cause appears to be the same: a rapid rebound after a prolonged bear market, but this finding may give investors another way of identifying these risky periods.

Use more than just price momentum

Quant fund Robeco uses an alternative version of momentum that uses analysts' revisions and news sentiment in addition to price measures. They find that this alternative definition reduces the drawdowns. As demonstrated by this chart graphing a particularly tricky period for momentum: November 2022 to February 2023:

This is what the Stockopedia Momentum Rank aims to do by including earnings momentum measures as well as pure price momentum. Full details of how the Momentum Rank is formed can be found here. However, as the recent performance of Robeco's Alternative Momentum and the Stockopedia Pure Momentum screen over the last two years shows, investors may be better off simply exiting (or even reversing) a momentum strategy when the chances of positive returns are low.

Implications for today

Momentum is a strategy that is pervasive across geographies and asset classes. So, while one market may be prone to a momentum crash, another will be trending nicely. The warning signs of a momentum crash appear to be the same: negative two-year returns, positive one-month returns and increased volatility of either markets or momentum strategies. As such, the UK small and mid-cap markets seem particularly prone to a rapid bear market reversal and hence a momentum crash, especially if volatility picks up. UK and US large caps appear less dangerous for momentum investors to engage with:

However, as always, when their own money is at risk, momentum investors should run their own numbers to look for these warning signs.