Over the next year or so, we are likely to see terms like ‘smart beta’ or ‘smart indexing’ with increasing regularity. There’s nothing wrong with that, however some may take exception to the perceived co-opting of the term ‘index’.
There is no rule or even universally accepted standard as to what constitutes an index in terms of measuring the performance of a population of stocks. By one account - the American Heritage dictionary, the noun ‘index’ can mean one of five things. The most relevant in our case is: ‘A number derived from a formula, used to characterize a set of data.' Even the term ‘benchmark’, which according to American Heritage dictionary means: ‘A standard by which something can be judged’ should not confer any increased sense of legitimacy to that which it’s applied.
Naturally, there are indices and benchmarks that are more appropriate than others in measuring the performance of a given population of stocks, but therein lies the rub; what population are you trying to measure?
Here are some thoughts on what to consider when selecting the proper index and/or benchmark:
- The population/universe of stocks whose performance one is attempting to track/measure
- The investment objective
For example, if an active manager is employing a sophisticated quantitative strategy for a U.S. portfolio, perhaps the S&P 500 is not the best bogey; it could be like shooting fish in a barrel. If the manager were compared to an index similar to and/or constructed based on the respective strategy, perhaps it would be revealed that they are underperforming or charging unwarranted fees. On the flip side, the manager with this quantitative mandate could be shortchanged if their portfolio outperforms the index based on the respective strategy yet underperforms the S&P 500.
Investments in index-based portfolios are nothing other than passive investing. Regardless of how complex or sophisticated the methodology, if it can be fully represented by a model which, theoretically, can be replicated by any manager, it’s an index and one could reasonably expect to get similar results across managers exclusive of portfolio trading and administration expenses.
A key characteristic of any index is that the composition is known prior to being effective. Based on the evolution of indexing, a better question might be, what constitutes beta? If a given index produces consistently higher returns with lower risk than a traditionally used index, perhaps there’s an argument that it represents a better measure of beta.
