Trend Following in 2017: A Post-Mortem?

(Spoiler alert: There is (most likely) nothing wrong with trend following)

Many of you know us from our counter-trend models, which can act as a great complement to trend-following models, but this blog will be focused on trend following given the struggles they’ve had recently.

Using Société Générale’s (SG) Trend Indicator we will see if the reports of trend following being “dead” are true. The SG Trend Indicator is “designed to have a high and stable correlation to the returns of trend following CTA strategies.” They do this by “trading” a simple moving average cross-over model across 55 futures markets in the 4 major asset classes (equities, currencies, fixed income, and commodities); with equal risk across the 4 asset classes.

The SG Trend Indicator was down 15% for the year. And following the performance in 2016 (-0.19%) and 2015 (-8.80%) people have started questioning if trend following is still a viable strategy. We find it very difficult to say a strategy is “dead” because almost all strategies go in and out of favor. And without a fundamental reason for a change in market dynamics we cannot conclusively say a strategy will not work going forward.

So, what happened to trend following last year? SG has a great tool for viewing the attribution of its Trend Indicator. (We encourage anyone who is invested or interested in trend following to check it out.) Using their tool we can see that trend following on equities worked really well (contributing about +12% in attribution) while bonds (-9%), commodities (-9%), and currencies (-7.5%) struggled heavily.

Figure 1: Asset class attribution

Within equites the largest contributors were the S&P 500, Hang Seng, NASDAQ 100, and KOSPI. All of which were held long throughout the year. Figure 2 shows the S&P 500, the moving averages used by the SG Trend Indicator, and the S&P 500’s cumulative attribution. As you can see the S&P 500 was in a steady up trend in 2017 which created a long position throughout the year and provided about +2.1% in attribution.

Figure 2: S&P 500

On the other hand, Gold (within commodities) was a major detractor; contributing -4.8% in attribution. There was a lot of back-and-forth price movement (what we call “noise”) in Gold during the year which caused the trend-following model to get “whipsawed” between long and short positions. The pink shaded area in Figure 3 shows the trend-following model switching between long and short.

Figure 3: Gold

Trend following did not have a great year in 2017 because of bonds, commodities, and currencies; however, a simple trend-following model performed well on equities. If trend following were to stop working on every asset class for a longer period (5-10 years) it would be a cause for concern that maybe trend following really is “dead.” But because of the variability in performance of different asset classes over the last couple of years we cannot say trend following will not work going forward.

Now as a caveat to the above analysis, this analysis was done on SG Trend Indicator which uses a simple moving average crossover model. Many/most trend following funds have more sophisticated models that better capture trends and avoid getting “whipsawed.” But the SG Trend Indicator is a good proxy for what you can expect (at the minimum) for a trend follower.

While trend following has been out of favor recently and struggled during the markets’ ascent, managed futures overall still play an important role in a portfolio: as a true diversifier with little to no correlation to the broad markets. This protection becomes invaluable during times of choppiness in the market.

Read our last blog post, Alpha vs. Fees >