The Need to Know “Why” and an Update on Trend Following
One of my working hypotheses is that most investors have a strong desire to have an explanation for “why” the market is moving as it is. An explanation that makes sense gives investors an illusion of control that is calming.
I say it is an illusion because there is little reason to think that an explanation is true just because it makes sense. Experiments show that people are equally able to produce an explanation for events that actually happened as for events that didn’t happen. A group of people who were told that the false outcome actually occurred are not only equally able to come up with an explanation but also are equally convinced that it was obvious, in retrospect, that this would occur.
Not only is an idea that “makes sense” not necessarily true, it is also not necessarily the case that knowing the true explanation for events will allow us to outperform going forward. Circumstances can change without notice, and there is no reason to think that we will spot a change in the underlying forces before the market does, even if we do understand them.
Nonetheless, having an explanation for market moves that makes sense is something that most investors strongly desire. It’s hard to see another reason for why CNBC and Bloomberg TV are constantly serving up new and different explanations for recent price changes. The fact that so many investors stay glued to these stations suggests that they are producing a service (explanation generation) that is wanted.
I don’t disagree that this is valuable. As the Decision Lab wrote, “believing that we are in control of what happens to us, even if it is not always completely true, is actually an important part of our mental health”. They report on experiments showing that participants going through an unpleasant experience (an annoying buzzer sound) are significantly happier if they believe they have the ability to stop it, even when they do not choose to do so. In investing, feeling like we are in control can help us stay invested, which will lead to higher long-term returns, even that belief is really an illusion.
That said, my belief is that, if the average investor has a strong psychological need to know “why” the market is acting as it is, then there is alpha available to those who can take the other side, making trades without any explanation for why those trades make sense.
Applications of machine learning to investing, the focus of my research for OSAM, are examples of strategies that can provide such alpha. Complex machine learning models make predictions that do not come with simple explanations for why they make sense. Indeed, in most cases, their predictions do not have simple explanations, and yet, they achieve higher accuracy than simpler models that do have them.
Compared to building complex machine learning models, there are strategies for capturing the same alpha that are simpler. One of those is trend following, a strategy that I have written about previously. If you were wondering why a machine learning researcher was writing about trend following, now you know: machine learning and trend following both provide buy and sell recommendations without any explanation for why they make sense, and hence, aim to capture one of the same sources of alpha.
In the remainder of this article, I want to give an update on the trend following approach that I described in my previous article and add another.
Growth vs Value
For at least the past three decades, growth and value stocks have taken turns outperforming, typically for years at a time. While growth and value investors spend their time arguing with each other about which approach is better, you could have stayed on the right side for most of that time with the simple trend following approach described in the post mentioned above.
In the original article, written in the middle of 2020, the trend was still in favor of growth stocks, so coming out of the pandemic, it would have recommended sticking with Zoom, Shopify, and the other “covid winners”. In 2021, the trend whipsawed between the two approaches, but in late December of 2021, it switched back toward value and has remainder there since then:
Following this strategy would have kept you out of the destruction in growth names that has occurred this year, with both Zoom and Shopify, for example, down over 70% from all time highs.
To be fair, when value outperforms growth, there is an obvious “why”: value stocks are cheaper. However, we are still left with the question of “why now”. Value stocks were cheaper for most of the past decade, yet growth outperformed for most of that time.
Today, a consensus has formed that value is outperforming because rates are rising. If decreasing rates are “obviously” good for growth stocks, then it should have been clear a decade ago that we should buy them as rates dropped. As my colleague Jesse Livermore points out, that did not happen:
“Growth stocks benefit from low rates” is the consensus explanation today, but it’s not clear to me at least that it is really what is going on. The 10-year yield currently sits at 1.7%, suggesting that the market thinks rates will remain below their 2018 levels for much of the next decade.
What is clear is that investors are dumping growth stocks at the moment, and it would not be wise to try to catch that falling knife even if you think you know the reason why the market is moving as it is. It’s much easier to simply stick with the trend toward value stocks until it reverses. Trend following will let you profit from these shifts if you can set aside the need to know “why”.
Developed vs Emerging Markets
In his book “Beyond Diversification”, Sébastian Page states that Developed vs Emerging Markets also shows clear trending behavior. Checking this out for myself, I found that this does appear to be the case. Here is the same chart as above but replacing value vs growth with emerging vs developed markets:
It is clear from the chart that EM and DM each tend to have long periods of outperformance. This remains true if we include the US in the Developed Markets, as above, or exclude the US as shown here:
The charts above were formed using data from the Fama/French data library. We see the same behavior using highly liquid ETFs, which have been available since at least 2004:
In the Fama/French data, both DM and EM have returned just over 8% annualized since 1990, with Sharpe ratios of 0.13 and 0.11, respectively. The simple approach of owning DM when the ratio of DM / EM is above its 10-month moving average would have produced returns over 11% annualized since 1990, with a Sharpe ratio above 0.16. You get an extra 3% per year with little work required other than the willingness to switch sides of the trade without an explanation for why it makes sense to do so.
Conclusions
With Growth vs Value and Developed vs Emerging Markets, trend following has a solid track record of producing alpha. What it doesn’t do is provide the calming effect that comes from knowing “why” one side should outperform the other. However, if you believe that emotional comfort should come at a cost to investors, then the outperformance of these strategies is no surprise.
