We can get monthly performance data for all three factor premiums all the way back to 1963, the inception point of the profitability data. Since our focus in this piece is on the three major factor premiums we target in our portfolios, we used that as our starting point for this analysis. We calculated five-year rolling returns for each of the three factor premium returns since 1963. Five years is long enough to identify a trend in the factors but short enough to retain some variability in that trend over time. For each rolling five-year period we did a simple frequency count: how many of the factor premiums were positive during the five-year period? The table below shows the results.
In the table we see that 52.7% of the time, two of the three factor premiums we target were positive historically. Put another way, of the 603 five-year rolling periods we examined, two of the three factor premiums were positive in 318 periods. If we add in the number of times that all three factor premiums were positive, we see that 82.2% (52.7% + 29.5%) of the time historically two or more of the factors were positive.
Another point to note, and maybe the most impressive observation from the table, is that there has historically NEVER been a rolling five-year period when all three factor premiums were simultaneously negative, which is why we design our portfolios to have and maintain exposure to multiple factor premiums.
In a world where factor premiums can be cumulatively negative for multi-year periods, it is heartening to know that there is a good chance that one, or perhaps two other factors might be positive while that one factor is out of favor. It’s also encouraging to see that when we examine these factor premiums, the chances that at least TWO of the three factors will be positive has been historically very high. The diversifying benefits of having multiple factor premiums in our portfolios is real, and it is a trait we would expect to continue to help investors for the foreseeable future.
Special thanks to Bob Bannon for his help compiling the data for this blog.
Past performance is no guarantee of future results. Information is not indicative of any Loring Ward product or strategy.