Is There a Cost to Transparency?

Abstract

Conducting the first direct tests for the cost of private transparency, the authors examined whether a willingness to offer transparency to investors is beneficial or costly in terms of hedge fund returns. Transparency is implicit when a fund accepts managed accounts because such accounts are directly controlled by investors. Overall, the authors found no evidence that transparency harms fund returns. They also found no support for concerns that managers offering transparency suffer from selection bias.

As a result of the 2008 financial crisis, investors have been pushing for more transparency for their hedge fund investments. Position-level transparency requires full disclosure of all fund positions to the investor on a regular basis. Such transparency is useful for several purposes. First, it makes it easier to monitor the underlying financial risks and the actions of the manager, which should help alleviate agency problems between investors and portfolio managers. Second, it allows risk measurement and aggregation across the entire portfolio. Third, it can also be used to optimize the portfolio management process.

Many hedge fund managers, however, fiercely resist offering transparency for a number of reasons, ranging from the potential loss of a competitive advantage by the reverse engineering of their strategy to the possibility of others’ front running their portfolio. In addition, it is sometimes asserted that the very best hedge fund managers have sufficient assets and thus do not need to offer transparency. Hence, requiring transparency could create a “selection bias” among hedge fund managers. Both arguments imply that transparency creates a cost to the investor in terms of lower hedge fund returns.

This hypothesis, however, has never been directly tested. Ours is the first study that directly tests for the costs of private transparency. We measured a fund’s willingness to offer transparency by whether it accepts managed accounts (MACs) for specific strategies. Because MACs are generally run pari passu with the main commingled fund, they indirectly reveal positions in the main fund to the investors in the MACs. Therefore, hedge fund managers who accept MACs do offer transparency, at least to some investors. Such transparency can be called “private” because it is limited to investors in the fund, in contrast to “public” transparency, whereby fund positions are revealed to the general public.

We investigated whether this willingness to provide transparency is related to differences in the investment performance of hedge funds. Using the TASS database, we classified funds into a sample in which managers accept MACs and another in which managers do not accept MACs. Because the database does not report MAC returns, however, we analyzed only the returns for commingled funds. Thus, we compared the performances of the commingled funds of managers who provide transparency with the performances of those funds that do not. We found that the performances of these two groups cannot be distinguished from each other. This absence of significance holds for raw returns, abnormal returns, and alphas from a multiple-factor model, as well as across different hedge fund sectors. This result also holds regardless of whether the fund is open or closed to new investment.

We also examined whether managers who provide transparency are more or less likely to be involved with fraud later on. We would expect that managers who accept the additional monitoring made possible by transparency would be less likely to commit fraud. We could not confirm this conjecture in the data, however, owing to the small sample of fraud cases. Nevertheless, we found that the duration of the fraud is longer by more than 14 months, on average, for managers who do not provide transparency. This result is consistent with the fact that transparency provides faster detection of fraud, which is opportunistic rather than premeditated for managers who provide transparency.

Overall, we found no evidence that transparency harms fund returns. In addition, we found no empirical support for concerns that managers who offer transparency suffer from selection bias.