Ghost Returns
For a while, I ran Morningstar's hedge fund database. A grim task that, as hedge funds delight in gaming database providers. Consider one hedge fund that started up in the early 1990s. "Four months later the fund began reporting to a database, and a year after inception it reported assets under management (AUM) in the top quintile of all funds. In the mid 2000s, the fund experienced a troubled quarter and saw its AUM halve in value. It then ceased reporting AUM figures. The fund's performance recovered, and during the last quarter of 2008 it reported a particularly good double digit return, putting it in the top decile of funds. However, a few months later this high return was revised downward significantly, into a large negative return."
The example and quote come from "Change You Can Believe In? Hedge Fund Data Revisions," a working paper by professors at Oxford and Duke. As private entities, hedge funds can pick and choose when reporting to databases. They often report to some databases and not others; they release some data points and not others; and they appear and disappear from the databases on terms that the hedge funds find to be favourable. Worst of all, and the paper's subject: Hedge funds might not report correct performance.
In the four-year period from 2007 to 2011, the professors found that 49% of hedge funds later revised at least one of the monthly performance figures that they submitted to the database providers. That wouldn't be of particular concern if the revisions were tiny. In some cases they were--only a single basis point. However, 21% of the funds had one or more revisions that were at least a percentage point--a large amount indeed.
The professors found a number of disturbing tendencies with these performance revisions. More of them were down than up. Hedge funds that had at least one performance revision had lower future returns than funds that did not have revisions, with the gap between the two groups being a hefty 3.5 percentage points per annum on a risk-adjusted basis. In addition, funds that had a performance revision were likelier to cease operations and liquidate.
All this suggests that hedge fund databases provide a new data point: Fund has previously revised a performance number. Fortunately, it is not a data point that is necessary for mutual funds. In addition to having much lower expense ratios than hedge funds, mutual funds rarely revise their past performance figures, and when they do it's almost always a tiny adjustment. Receiving bad data is a rich man's disease.
Read all John Rekenthaler's missives here.
