News and Insights

Is momentum expensive?

Article | 15 Sep 2018

Growth, quality and momentum have performed well over the past few years. Investors have been happy to buy companies that have earnings certainty and produce healthy returns on equity. Cheap stocks, apart from a strong resurgence in the final quarter of 2016, have done less well, underperforming more expensive, faster-growing, stocks. After any period of outperformance of a particular style or strategy, it is natural to ask when the performance will revert. Asset allocators are now beginning to ask exactly this: is it time to shift away from growth, quality and momentum and towards value?

From a systematic perspective, significant factor rotations are too few and far between to be able to make statistical inferences regarding timing. No question, timing style shifts is as difficult as timing the market itself. It requires investors to take many different sources of information into account: the extent of the growth rally, capital market volatility, the economic cycle, and the positioning of other market participants, among others. However, as often is the case, one simple solution seems to work well.

Before we discuss this, though, we take a look at where we are in the cycle. To help us with this, we use the rolling 12 month average of the factor information coefficients (ICs). Defined as the correlation between next month’s returns, and this month’s factor rank, ICs are a way to visualise how well a factor is predicting returns.

From the perspective of Australian equites, the charts do not imply that investors are indiscriminately buying quality, momentum and growth. They are doing well, but the future monthly return of quality, momentum and growth is not guaranteed at the moment – we are not in a market in which simply allocating to these factors on an equal weighted basis will make money in the short term.

What is clear from the above is that if, and it’s a big if, an investor can time their exposure to factors, then their returns will be improved more than if they invested in just one factor. A profitable strategy would have bought value three times, holding for about 9-12 months each time, and sold growth twice, holding growth for the rest of the period. The problem, however, is that this is very difficult to do at the right point. No one knew how far the market would fall in 2009, or how much value stocks would rise in the second quarter of 2016, or how far tech stocks would run in the dot com era.

Correlation

No investor puts all their eggs in one basket. Even the most concentrated portfolios have more than one stock, and for good reason: the future is unknown. When constructing portfolios, or pairing managers, it may not be worth investing in two managers that both have very strong returns for the simple reason that the returns may be correlated. If the returns from both managers rise and fall together, then all an investor gains from buying two managers is more paperwork. However, if the returns from each manager differ in different environments, when one manager outperforms, the other will probably underperform. The aim is that over time, the outperformance of one manager will more than compensate for any underperforming periods of the other. This is exactly what we see with momentum and growth, which is negatively correlated with value. We plot the 12 month rolling returns from three long-short strategies made from momentum, growth and value.

When value works well, momentum and growth do less well, and vice versa. The negative correlation means that an investor portfolio that consists of both value and momentum will have less exposure to the cyclicality of styles within equities. Right now, it means that there is less need to ask whether growth will outperform, or whether the technology names have run too hard for their earnings. An investor with both growth and momentum, and value in their portfolios will have hedged their downside, and be more prepared for a factor rotation that may occur in the future.

What is happening?

We think that part of the reason the returns to momentum and value portfolios are negatively correlated is to do with timescales. On average, investors make money from buying a value stock about 12 months after purchase. With momentum (and from a factor perspective, growth to some extent), investors make money within the first 3 months. These timing differences allow investors to profit from both momentum and value in the same portfolio.

Another reason for the negative correlation might be in the underlying drivers of the factor returns. Momentum returns are likely to be driven by investor behaviours. To name a few (in no particular order):

  • Anchoring bias: when making decisions, individuals place too much weight on the first piece of information they get. In markets, this can lead to investors placing too much weight on the latest stock price.
  • Herding behaviour: following the actions of a group is easier than standing out from the crowd. Behaviour like this is particularly evident at the top of bull markets or the bottom of bear markets.
  • Gamblers fallacy: the mistaken belief that if something happens more frequently than normal, it will happen less frequently in the future. Stock returns are one example here. Just because a stock has had positive returns a few months in a row does not mean that it is more likely to have a string of negative months.
  • Loss aversion: people prefer avoiding losses to acquiring gains of the same size. Investors are more likely to hold onto losing positions that really should be sold.
  • Overconfidence: belief in excessive accuracy of one’s assumptions can lead investors to make portfolio bets that are outsized and made on information that has not been fully stress tested.

Returns from a value strategy, meanwhile, are more likely to be a compensation for risk. Cheap stocks are being valued as cheap because investors on average feel there is something wrong with the business, and there is no obvious fix that will lead to future growth. When grouped together, a basket of value stocks should compensate investors for this risk. Often, it is only a handful of the cheap stocks that go on to perform well, but that is all that is required for a portfolio based on value stocks to outperform.

So, is momentum expensive at the moment?

We calculate the median PE of four factor portfolios in Australian equities: momentum, value, quality and growth. The portfolios are constructed using the top 20% of stocks, equal weighted. We use median PE so that very large or small outliers do not affect the results.

Using this measure, both quality and momentum look expensive, but so does value. Growth (measured using forward EPS) looks less expensive, but is still not cheap relative to history. This result is concerning for investors thinking about allocating to factor portfolios, although if an investor is going to pair value with momentum, timing is not everything. Investors should be careful if they try to time styles.

From our perspective, implementation of factor portfolios is as important as choosing the right factors. The process for the Systematic Growth Fund has risk control at each stage, from fundamental factors, to smoothing, to optimisation. All these combined lead to a better risk reward outcome for investors. The effect of all these can be seen in the Systematic portfolio PE, which is not nearly as stretched as that seen in naïve quality or momentum factor portfolios:

Momentum, quality and growth do better than value in the long run in Australia. However, investors can improve their outcomes with these two steps:

  • Investors can hedge against large factor rotations by pairing momentum, growth and quality with a pure value exposure
  • When buying any factor, implementation is as important as factor choice, and thoughtful portfolio construction can dramatically improve investor outcomes

For more information please contact

Gary Adamson

Gary Adamson

Chief Executive Officer – Platypus Asset Management

gadamson@platypus.com.au