Skill Ratio vs. Information Ratio
Following the introduction of the Skill Ratio, it’s worth exploring how it differs from the information ratio.
First, how are they calculated? The information ratio’s denominator can be computed in different ways, but it generally explores tracking error based on shorter time periods, usually monthly or annually. The Skill Ratio can also use varying time periods, but the original article applied a “tracking error” based on rolling five-year returns.
The distinction is this: As an investor, I should be willing to forgive a manager who has a bad month or even a bad year. I want to give a portfolio manager the benefit of the doubt and allow proper time for their investment theses to play out. Demanding consistent outperformance over such short time periods is unrealistic. Therefore, it is inappropriate to judge excess returns on a monthly or even annual basis.
What I have a much harder time overlooking is a bad five-year period.
Might five years seem arbitrary or even too short? Perhaps. But it doesn’t matter if an active manager can eventually outperform over a “full market cycle” or within seven years instead of five. Why should I wait an entire market cycle or seven years for an active manager when the Skill Ratio shows that there are cheaper options — multi-factor indexes — that have consistently delivered outperformance within a five-year window?
I’m not out to prove that active management doesn’t work, just that a more logical solution exists for those who wish to pursue alpha. I don’t have to outrun the bear, I just have to outrun you.
When comparing the information ratio and the Skill Ratio across all 490 US large-cap funds that have survived since 1999, there is an R2 value of 0.70. This indicates some degree of significance in the relationship between the information ratio (independent variable) and the Skill Ratio (dependent variable).
US Large-Cap Mutual Funds, 1999–2017
However, when you cluster the data into funds that outperformed and underperformed the S&P 500, a different story unfolds.
Funds That Underperformed the S&P 500, 1999–2017
The R2 value for underperformers shows a relatively high correlation (0.65) between the Skill Ratio and the information ratio.
Ultimately though, the Skill Ratio, like the information ratio, is intended to be used with managers who have delivered some level of outperformance over time. Whether or not it’s sensible, most people can, and do, rule out underperforming investments based simply on their lagging returns. No pompous ratio is necessary.
However, among the cluster of outperformers, the R2 value is 0.39, which does not indicate a convincingly strong relationship between the Skill Ratio and the information ratio.
Funds That Outperformed the S&P 500, 1999–2017
One of the two data points near the top of the preceding graph is the Natixis US Equity Opportunities, which has an information ratio of 0.32 and a Skill Ratio of 1.78. The large residual indicates that the fund had relatively volatile excess returns on a monthly basis but fairly consistent excess returns on a rolling five-year basis.
Isn’t it worthwhile to understand that while a manager’s monthly excess returns may be volatile, their excess returns across five-year time periods are relatively smooth?
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: ©Getty Images/erhui1979
Interesting and informative. Thanks.
Thought provoking post. I understand that the graphs showing the relationship between the Information Ratio and Skill Ratio are meant to be viewed from a wholistic stand-point, however, the lower correlation for the outperformers could be explained by the type strategy each fund runs. An extreme example: if you have an outperforming long-only fund that has an investment horizon of say 5 years for any given trade/investment, and you decide to look at the returns on a very short term basis (for arguments sake – seconds or minutes) then clearly the Information Ratio will be very low and the Skill Ratio will be high. I guess I think it makes sense to potentially categorise the funds into strategy type to compare apples with apples.