Saturday, March 05, 2005

Are you paying for high alpha or bad benchmarks?

Clumsy benchmarking and incomplete performance disclosures are probably two of the largest factors contributing to the sustained predominance of active investment management in small to mid sized pension plans and the retail environment.

Over the course of many institutional investment evaluation engagements I have discovered that much advertised or implied active management alpha can be better attributed to high benchmark misfit than high investment skill. Benchmark misfit refers to the asset class and style differences between the fund and the benchmark to which it is compared. An aspect of the asset management business which is troublesome thought not surprising is that benchmarking, full disclosure and performance evaluation is not a mission critical competency for the sell side. I haven't concluded whether this results from a deliberate intent to deceive or simply the fact that good benchmarking is too complicated and nuanced for often impatient and uneducated investors. Probably elements of both. .Suspiciously, the bias in benchmarking or information delivery always tends to accrue to the benefit of the manager or seller rather than the investor. Examples of misdirection abound. I was told recently by a large pension provider that if their client wanted to see portfolio level historical returns they could only get it by engaging one of the firms affiliates? Some asset managers/pension providers must feel that client specific cumulative historical returns and fair benchmarking can only serve as constant reminder that a client has underperformed. If any attention in these settings is paid to performance it is often only at the individual fund level. The benefit of this practice is that as soon as an underperforming fund is removed from the portfolio, the negative performance is gone as well. This why you may see many investment choices in a sponsored program with under 3 years of actual performance but 3 and 5 year composite numbers that look very strong. New funds with good historical results are used to attract new money. Fiduciaries should insist on seeing portfolio level cumulative actual performance against a reasonable portfolio level benchmark.

The S&P 500 is used quite often as a measure for US equity performance ( regardless of market capitalization or style in the actual portfolio). However, one rarely (make that never) see a Large Cap US equity strategy benchmarked to a diverse global benchmark where the asset class diversification and return characteristics favor the benchmark. I have often seen strong investment out-performance for products presented on a cumulative performance basis covering substantial time periods. What may not be presented are the annualized returns over the cumulative period. This information can clearly demonstrate the amazing leverage a single and possibly accidental annual outperformance can have on long term results. Another "trick" involves setting an investment policy cash target of 5% or so but never allocating more than 1% in cash for liquidity. Actual portfolio performance against a policy benchmark in this case automatically includes the tailwind of the long term risk premium of equity/bonds over cash. If investors do get paid to take systematic risks(over the long run) then, including assets with more systematic or market risk in a portfolio than in its benchmark should guarantee outperformance and possibly less risk over time ( other things being equal). This happens to be the kind of risk-free bet that many managers are getting paid very handsomely to disguise.

I agree with Ron Surz that benchmarking and peer universe generation is very prone to leading investors to wrong conclusions either through incompetent generation or purposely skillful manipulation. There are several large institutional pension investment programs that base their "fiduciary ratings" heavily on peer universe rankings. Unbelievably, in 80-pages of process and comparator information you will not find a detailed description of how the peers were selected and who they are. In lieu are vague references to some creditable firm which approved the reasonableness of the selection process.

Take care in looking at performance. The better it looks the longer you should look.


Post a Comment

Links to this post:

Create a Link

<< Home