It may not tell you what you think it does
It is appropriate to prefer strategies that fully participate in up markets, and that provide protection in down markets, and investors frequently employ up-market capture ratio vs. down-market capture ratio to evaluate investment strategies. However, because of the way statistics are calculated, comparing a strategy's up-market capture to its down-market capture provides little, if any, information beyond knowing that the strategy outperformed its benchmark on a risk-adjusted basis.
It seems logical to assume that a portfolio manager whose strategy had greater up-market than down-market capture demonstrated either timing skill (higher beta in up than in down markets) or a defensive process (added more alpha in down than in up markets), but there is little evidence to support either assumption. Most of the time, the difference between strategies' up-market and down-market capture ratios offers little information about portfolios' market-conditional performance. Instead, the difference in capture ratios is simply another measure of unconditional alpha, which means it should not be viewed as an additional indicator of manager skill.
To understand market-conditional behavior, we recommend researchers look deeper: To evaluate a manager's timing skill, compare the portfolio's beta in up markets to its beta in down markets. To investigate if an investment process is "defensive," compare the portfolio's alpha (market-adjusted return) in down markets to its alpha in up markets.
Figure 1 illustrates three ways that an investment manager can achieve higher up-market than down-market capture. The first set of columns relates to a 5% down market. The second set describes a 15% up market.
- Market timing: The first row represents successful market timers, which have higher beta in up markets than in down markets. For this analysis, it does not matter whether the shift in beta is achieved by rotating the stocks in the portfolio or by using cash. Regardless of how it is achieved, timing skill will result in the portfolio returning more than the market in up markets and losing less than the market in down markets. In this example, the up-market capture is 110%, the down-market capture is 90% and the difference is 20%.
- Defensive strategy: The second row represents defensive strategies, which add more value in down markets than in up, and therefore have lower down-market capture than up-market. In this example, the strategy produces an alpha of 1.5% in the down market and an alpha of 0.25% in the up market, while having a beta of .90 in both. That results in a 32% difference between up-market and down-market capture ratios.
- Unconditional alpha: The third row of the table represents a manager that has skill that's unrelated to market direction. The strategy's beta is always 1.00 (no timing), and the alpha is the same in up and down markets (not a defensive strategy). Despite the lack of timing skill and the equal success of the strategy whether the market is up or down, its up-market capture is 27% more than its down-market capture. The reason is that the capture metrics are calculated as the portfolio's average return divided by the market's average return. Positive alpha results in less than 100% capture in down markets and more than 100% capture in up markets. As a result, the up-market capture is higher than down-market, even though the alpha is uncorrelated with market direction.
What does the difference between up-market and down-market capture tell us?
As described above, there are three conditions that can make up-market capture larger than down-market. To investigate which of the three were representative of actual portfolio performance, we used data for U.S. core equity strategies from the eVestment database. The statistics use 10 years of monthly portfolio returns ended Sept. 30, 2017, for 191 portfolios. With those data we calculated up-market and down-market capture ratios, up- and down-market betas, up- and down-market alphas and unconditional alphas. We then addressed the question: Was up-market minus down-market capture more related to the strategies' up-market minus down-market beta, up-market minus down-market alpha or unconditional alpha?
Timing skill did not explain the difference between up-market and down-market capture ratios
Figure 2 shows there was no relationship between the difference in portfolios' up- and down-market betas and the differences between up- and down-market capture ratios. The correlation between the two metrics was slightly negative, but statistically insignificant.
Strategies' defensiveness did not explain the difference between up- and down-market capture ratios
Figure 3 shows there was little relationship between the portfolios' defensiveness (the difference between up- and down-market alphas) and the differences between up- and down-market capture ratios. The correlation between the two metrics was only modestly positive and statistically insignificant.
Unconditional alpha did explain the difference between up- and down-market capture ratios
Figure 4 shows there was a strong relationship between the portfolios' unconditional alphas and the differences between up- and down-market capture ratios. The correlation between the two was 0.87 or an r-square of 76%, which means the variability in unconditional alphas explains 76% of the variability of the differences between the strategies' up-market and down-market capture ratios.
The difference between up-market capture and down-market capture ratios is highly correlated with unconditional alpha, and is not correlated with strategies' market timing performance or defensive properties. Therefore, investors should not consider the difference between up-market and down-market capture as an additional indicator of manager performance.
However, there are statistics that can help identify strategies that added value through timing and strategies that exhibited defensive properties.
Successful market timers should have larger betas in rising markets than in falling markets. Strategies with defensive characteristics should provide more alpha in falling markets than in rising.
Lastly, if your manager research process uses a screening process to help select timing or defensive strategies, it is important to evaluate the underlying investment process to ensure the value added from market timing or defensive characteristics was an intended outcome of the process, rather than a statistical artifact that is unlikely to repeat in the future.
Ralph Goldsticker is chief investment officer at Alan Biller and Associates, Menlo Park, Calif. This content represents the views of the author. It was submitted and edited under P&I guidelines but is not a product of P&I's editorial team.