We’ve talked a bit before about how there’s a booming academic business in papers finding that investment managers do or do not add value versus non-managed alternatives like passive indexing or keeping your money under your pillow and just burning a constant percentage of it every month. Part of why that’s a thing is that the data can be prodded, smooshed, or cherry-picked to say many different things, and so they are. I enjoyed this paper about mutual funds by Stanford GSB profs Jonathan Berk and Jules Van Binsbergen (NBER today here, SSRN in April here) in part for its discussion of data problems, which starts with the fact that they used the industry-standard (in the academic-papers-about-mutual-funds industry) CRSP database and compared it to Morningstar data because “even a casual perusal of the returns on CRSP is enough to reveal that some of the reported returns are suspect.” Suspect like:
We then compared the returns reported on CRSP to what was reported on Morningstar. Somewhat surprisingly, 3.3% of return observations di ffered. Even if we restrict attention to returns that di ffer by more than 10 b.p., 1.3% of the data is inconsistent. An example of this is when a 10% return is accidentally reported as “10.0″ instead of “0.10″.
That is one way to get alpha. Anyway they look at the data using a (strangely) unusual metric of dollar value added, which is roughly alpha (gross excess return over some investable benchmark, in this case a Vanguard index fund) and multiplying it by assets under management, the intuition being that making 1% excess return on a $10bn portfolio is more impressive than doubling your $10 bet at the craps table. And they find that mutual fund managers are better than controlled money burning by the thinnest of margins: Read more »