Isn’t Performance Just Math?
If you have worked in the investment performance industry for a certain length of time, you are very likely to have come across the statement “performance is just math.” In a sense, it is math. Investment performance is taking values and inputting them into a formula which produces a result. So yes, it is math. Such conclusions lead many experienced investment professionals to become puzzled, perhaps dumbfounded, when confronted with varying returns for the same portfolio. Here are three reason why performance isn’t as simple as 1 + 2 = 3.
- Whose performance are you measuring? In other words, are you measuring the performance of a manager or a client? If the former, then the best practice is to utilize time-weighted return measures to reduce the impact of external cash flows. One exception is the case of private equity investments where the manager controls the cash flows (i.e. capital calls). If the latter, then a money weighted return measure would be more appropriate to capture the impact of cash flows as they impact the return of the investments from the client’s perspective. Unfortunately, it is very likely the two measures will yield differing results.
- How frequently are you measuring? One might think this shouldn’t matter. Isn’t the whole the sum of the parts? Well… not necessarily. The monthly modified dietz formula, which is an estimate of a money weighted return when measured for one month, but when geometrically linked over multiple time periods becomes a time weighted measure, is still used frequently. The modified dietz can be deployed and run between significant external flows (per GIPS requirement) or even on a daily basis. Generally the more frequent the measure, the more accurate the result. The result, again, likely differs based on return frequency.
- Data tells the story. Everyone has heard the old saying “garbage in, garbage out.” Well this holds true in investment performance. Market value, accruals and transactions are only as accurate as the underlying data that supports them. Accruals can’t be calculated without coupon rates, market values can’t be calculated with accurate pricing and transactions won’t create the right impact if they aren’t coded properly. Is the performance right? The answer should be “it’s right but the data tells the story.”
Some say that performance is more of an art than a science. I’d say they are right… based on the data.