Wednesday, December 13, 2006

Closet indexing or folding bad hands

I've mentioned my fascination with poker in this space before. While I don't play very much myself, I see so many parallels to successful trading and successful poker playing. One of these parallels is knowing which hands to bet, and which hands to lay down.

Let me back up and talk a little about the concept of benchmarking. Professional investment managers live and die by whether they beat an index. So much so that many in the profession decry certain managers as "closet indexers." That is, an investment manager who charges active management fees, but really is just trying to match the index.

Now, I'm not saying that some managers are acting as closet indexers in a cynical attempt to retain clients by staying around the benchmark. Most would agree that while outperformance is OK, underperformance gets you fired. Also, most IM's that I know get paid more on asset growth than on relative performance. So if producing returns that are around the benchmark is enough to bring in new assets, the manager may conclude that its not worth risking underperformance to deviate from the index at all. Obviously such a person is not serving their clients.

But many times the term "closet indexer" is over used. When large investors like pension funds or endowments hire investment managers, they do so in an attempt to fill various buckets. E.g., a large cap, small cap, venture capital, investment-grade bonds, high-yield, etc. These buckets are selected and weighted based on estimations of long-run return and cross correlation patterns. What do they use to estimate return/correlation figures? Indexes.

So if I'm hired to run a Lehman Aggregate strategy for a large endowment, but all I buy are MBS, then regardless of whether I'm outperforming or not, I am not serving my client. Then client asked me to manage a portfolio which would have a beta near 1 to the Lehman Aggregate. Same would be true if I held a very low duration or bought a bunch of high-yield bonds. The portfolio I was hired to create has to fit with various other portfolios, and if I ignore that then performance doesn't matter, I should be fired.

So let's say that I just have no view on rates right now. I've done my research and find arguments for both falling and rising rates to be compelling. But I'm running a bond portfolio, so obviously I'm going to have to make some kind of duration decision. No problem, I pick the index's duration. Am I being a closet indexer? No, because the client has asked me to run a portfolio with a beta to the Agg of around 1. So when I have no view on interest rates, I set my position such that my beta is near 1. It isn't a cop out, its good investment management.

I think of it like a poker hand. If I look at my cards and read all the other players and I just don't think I have a winning hand, I should fold. Too many investment managers want to make a bet just because they think its their job to make bets. Sometimes its just your job to wait for another hand or a different market environment.

2 comments:

Anonymous said...

What's the best way for a bond manager to find alpha for the portfolio ?

Accrued Interest said...

I think most managers would say credit analysis. For me its MBS analysis. Both would fall under the category of security selection.