Gavekal Capital: Who Says Tracking Error Isn't Predictive of Performance and Alpha?

Tuesday, March 10, 2015

Who Says Tracking Error Isn't Predictive of Performance and Alpha?

The extremely well greased PR machine promoting passively managed index funds as the best thing since sliced bread has gone to great effort to convince the investing public that money cannot be made through active strategies. Outperformance is illusory they say, not predictable, not repeatable, subject to change, and the search for outperformance is a risky endeavor. If only that were true everyone's investing lives would be much easier. We'd all buy the market capitalization weighted investable universe packaged into one nicely wrapped ETF charging 0.1% and our quest to be average would be complete.

Unfortunately for the efficient marketers - but fortunately for anyone who wants to be above average - there is plenty of evidence to suggest otherwise. We'll be hitting a lot more on this topic in the weeks ahead, but for now we simply present the following table showing the correlations between annualized performance and annualized alpha to tracking error (tracking error is the standard deviation of return differentials between the fund and a blended stock/bond index) for all 482 balanced funds available to US investors that have a four-year track record.

Efficient marketers want everyone to believe that buying funds with higher tracking error to one's benchmark (i.e. not holding the "investable universe") is super risky and unlikely to result in material performance differentials over time anyways. This empirical evidence shows otherwise. Tracking error doesn't guarantee outpeformance, but it is a prerequisite.