Pages

Saturday, May 21, 2016

benchmarks >> passive > active

Short-term note: Already shared by CFA Switzerland.  This article contains many citations to nice articles from them (that I wrote), as well as cites from Morningstar, New York Times, and Pensions&Investments publications.

Removing benchmarks from a recently uncertain topic (here, here), we can instead focus on a slightly different perspective: how the performance differs between actively and passively managed funds.  Arguably the analysis now focuses active funds on other funds of the same asset class category, who are also weighed down with some (though lower) frictions of trading costs, taxes, and those infuriating fees.  A Morningstar report in the past year helps shed some light on this, and we explore it across all exclusive asset classes and time horizons.  What we see, sampled in the chart below, is that active funds continue to be underdog performers relative to lower-fee passive funds.  Most of the data still shows a fabulously low, minority chance of selecting active fund managers who outperform.  Certainly the conclusions about the strength of benchmark outperformance doesn't change at all, relative to the collective body of investment and academic research.  Though here the comparisons against passively managed funds makes things a little easier.


While a number of people have engaged with our previous work (linked above) on this matter, an analyst at Mandarin Capital following this blog brought this particular nuance to our attention.  Here are several probability observations we see worth noting in the raw data.

We note that the portion of active fund managers who outperformed their passive fund peers is volatile per asset class.  There also tends to be persistence within each asset class.  While US Large capitalization managers have a tougher time, over any horizon, emerging markets and bond managers fared just a little better.

A simple average of all of these asset classes provides critical directional insights.  Over a one-year horizon, 41% of active fund managers outperformed their passive fund manager peers.  This is obviously reflecting a less than ½ chance of your active fund being "above average".  There is also a wide distribution of outcomes about this 41%, depending on the asset class.  For example, 2 of the 9 (22%) categories saw 50% or greater of their fund managers outperform.  This is real performance depreciation; results that have less than a ~5% chance of occurring from "bad luck" alone.

Of course outperformance continues to become more challenging, through any combination of luck or skill, as we extend our time horizon.  A reason Warren Buffett is choosing low-fee Vanguard index funds in his will.  Over a 10-year frame, only 37% of active managers are outperforming their passive management peers.  The standard deviation about this portion, depending on asset class, also collapses by nearly 1 percent (when going from a 1-year, to 10-year horizon).

The result is that over 10-years, only 1 of the 9 (11%) categories saw 50% of greater outperformance by their active fund managers.  While the 1-year winners were US Small Value and Emerging Markets categories, by the 10-year horizon it was instead the US Mid Value that was the winning asset class.

Generally speaking however, there is persistence among the asset class manager's performance.  A result of not making many independent investment decisions over time, but rather the lengthy longevity of investment decisions playing out over a business cycle(s).  We've seen this before with benchmark data and with Harvard's endowment returns (which previous president Larry Summers highly enjoyed). We show in the dashed line above how if investment decisions and performance were independent year-to-year (hence outright zero persistence), then the portion of actively managed funds outperforming would be significant lower through time.

The actual active manager performance results are generally greater than could be expected by this theory, as we evidence at the 10-year horizon.  Though we also evidence at this time frame, the entire active management 1-standard deviation expectation (distribution between the trend high and trend low on the chart above, and especially in America) is below the 50% chance of an individual selecting an outperforming (or above-average) fund to begin with.  And why not, since there is all electricity but no compelling insight offered.  A chance that becomes wildly risky to take on as well, given the progressively tighter active management performance, much of it well below the 50% level, over lengthy horizons.  Bottom line, when possible, choose a diverse set of low-fee passive index funds and don't waste money chasing one another for the mirage of outperformance.

1 comment:

  1. Thanks for the shout-out. Is there any dispersion on the passive funds? If I choose a low-cost emerging market fund for example, will they all perform in line with the average ? Or do I risk picking an underperforming fund?

    ReplyDelete