Friday, February 21st, 2020

In Search of Fresh Inspiration

This week we ask whether the so-called “Boris trade” has further to run or whether there will soon be a period of mean reversion. The Boris trade is shorthand for the outperformance of housebuilders, utilities and dollar-sensitive industrials, which started in Q3 2019, as it became inevitable that there would be a General Election, which the opinion polls predicted the Conservative Part would win.

Our basic stock selection technique assumes that asset returns are serially correlated in the short run, – i.e. momentum will continue until it meets some sort of resistance. However, we are only interested in exploitable momentum, which is defined as returns which are high enough to justify the extra risk of being invested in a stock rather than the index. We could, if we wanted, make the opposite assumption, which is that mean reversion is dominant in the short run, and that a period of recent outperformance increases the probability of underperformance in the future. Even so, we would still need to account for excess volatility. Thus, a stock must be classified in one of three categories: exploitable momentum, exploitable mean reversion, no clear signal.

We can also model whether our momentum strategy outperforms our mean reversion strategy and you don’t need a PhD in econometrics to predict that stocks which underperform or outperform the index by a large amount also tend to have a high level of exploitable momentum. But the data vary a lot from week to and there is useful information in the shape of the curve and the goodness of fit. It is also helpful to split the chart into four quadrants to see whether momentum strategies are more successful for stocks which outperform or underperform.

The chart for the last 52 weeks shows a curve with a good fit (R-squared of 0.61 compared with a typical value of around 0.45) and a higher than average number of constituents in the top half of the chart. This suggests that the exploitable momentum effect has been more powerful than usual and/or that the effect of mean reversion is confined to a group of stocks which have performed more or less in line with the index.

We can also compare the number of stocks in each quadrant with the average for the last four years. The top right quadrant (dominant momentum & stock outperformance) has more constituents than normal, but the top left (dominant momentum and stock underperformance) is in line with the average. This indicates that the Boris trade is a collection of “long” ideas, which is not matched by an above-average number of “short” ideas. The space (or the funding) for these long ideas comes from a significant reduction in the number of mean-reverting stocks which are underperforming the index (bottom left quadrant). This reflects the fact that there are some well-known mega-caps from the Energy and Telecom sectors, which are underperforming, but not by enough to have exploitable momentum.

Looking at the last four years we can see that the number of stocks in the top left quadrant is close to its maximum – certainly post-Brexit – and that the gap between top left and top right (exploitable longs vs exploitable shorts) is well above average. Our conclusion is that the Boris trade is close to its peak and the only way in which it could get stronger is if investors reached a consensus on the stocks they hate as well as the stocks they love. Anything is possible, but we will need a new narrative of Brexit-losers to facilitate this.

Leave a Reply

Full Blog Archive:

Other Recent Posts