Past hedge
returns: more than meets the eye
Depressed
stock markets have given rise to the mainstream popularity of hedge funds.
Slick marketing for such funds often includes a snapshot of how these funds
have performed over the past several years, compared to well known stock
indexes, like the S&P 500 and the S&P/TSX Composite (formerly TSE 300).
However, there is more than meets the eye of past hedge fund returns. Depending
on the source, hedge fund returns may be biased.
Clifford Asness,
Managing Principal of AQR Capital Management LLC in New York, recently wrote a
sober reminder of the biases present in hedge fund databases. In an article
entitled, “Do Hedge Funds Add Value?” (AIMR Conference Proceedings on Hedge
Fund Management, 2002) he warns of the following:
Since many
positions held by hedge fund managers are illiquid (i.e. tough to sell quickly
without driving down the price), actual market values for many securities are
unavailable. Recall that this same issue came up in our discussion of the
valuation risk with labour sponsored investment funds (http://www.sterlingmutuals.com/Telus_LSIF_I_22mar2002.htm).
However, unlike the laws governing LSIFs, which that mandate internal values be
approved by an independent valuator, hedge fund managers have more flexibility
on valuing illiquid positions.
A common
way to value such holdings is to simply use an amount equal to the lesser of
original cost or fair value. But the fuzziness comes into play in the
determination of “fair value”? In other words, it’s very subjective – and that’s
a potential risk. Consider the simple example of a stock that is purchased
today at $10. Over the next few weeks, the market falls ten percent but no
trades take place in that stock. But since the managers assess its value to be
more than $10, they carry it at $10 (the lesser amount) on the fund’s books.
It’s likely the stock’s price would have fallen somewhat just by virtue of the
market’s decline. Keeping it at $10 on the books may be unfair.
(Recall
that a stock’s quoted price often fluctuates more than its “true value”.
Today’s market environment is a perfect example of how good stocks can see
their prices get undeservedly beaten up. However, the true volatility of quoted
market prices still should be reflected, even if the “true value” doesn’t change.)
An
unrealistically high share price will keep the fund’s unit price unduly high –
thereby favouring sellers and punishing buyers.
A
by-product of such uncertain valuation policies is an artificial smoothing of
fund volatility (up and down price swings). This smoothing effect de-emphasizes
a fund’s downward price swings, making its risk/reward combination look better
than in reality.
The more
illiquid a manager’s holdings, the more potentially biased its performance.
When a hedge
fund winds up or ceases to exist, some hedge fund databases will completely
eliminate that fund’s current and historical data. This is a common criticism
of historical mutual fund performance. When mutual fund category averages are
quoted, most omit the funds that have “died” over the years. The assumption
underlying this bias is that funds that cease to exist have, as a group, poor
historical performance. So, taking this group of poor performers out of a
category average artificially boosts the category’s track record.
Funds with
some history that are not currently included in any hedge fund database may, at
some point, ask to be included. Since these funds wouldn’t be new (just new to
the database) these funds may be allowed to provide all of their historical
data from inception, even though they were not part of the database in previous
years. Managers requesting this are likely to have good performance, thereby
creating another upward bias to historical hedge returns. Think about it – a manager
is not likely to want to be included in a database (and compared against the
competition) unless fund performance looks favourable.
Some
managers simply choose not to be included in hedge fund databases. There are
two potential reasons for a lack of inclusion. Managers with poor performance
won’t want to look bad against better performing peers. Also, some hedge
managers may simply be at a point in their careers where they have their ideal
level of assets under management, and simply don’t want any more business.
Hence, this factor may be positive or negative depending on the circumstances.
Some hedge
fund databases claim to be free of some or all of these biases. However, when
faced with an impressive track record of a hedge manager (or a fund category),
an awareness of potential biases will help in your decision-making process.
Next
week: We’ll look at historical hedge
fund returns with these biases in mind and provide some tips on portfolio
inclusion.
Dan Hallett, B.Comm., CFP, CFA is the Senior
Investment Analyst with Sterling Mutuals Inc. He can be reached at dhallett@sterlingmutuals.com Sterling Mutuals Inc. is registered as a
mutual fund dealer in Ontario, British Columbia, Alberta, and Manitoba.