Baseball and Exchange-Trade-Funds... do they have anything in common?

Baseball opening day is almost here, and for now the A’s are still playing their home games in Oakland. Like many of my readers, I am hopeful that this season will go a little better than last year’s debacle. Win or lose, my kids are excited that we’ve already booked tickets to fireworks night and the newest post-game attraction, the laser drone show. That’s right—after the game there will be drones with lasers!

I have been fascinated by the evolution of batting average within the baseball community and think there are real parallels to the ETF ecosystem. For example, consider that batting averages have tended to be the main mechanism for gauging a batter’s potential success, even though that statistic does not factor in other important offensive elements such as walks, speed and power. Similarly, ETF trading statistics, like average daily volume, do not contemplate the liquidity of trading the underlying basket of securities.

Apologies to my readers who have no interest in baseball, but today I wanted to spend a little more time discussing batting averages. I think that the reason this statistic has, for more than 100 years, been considered the foundation of measuring a hitter’s ability lies in its perceived simplicity. Take the number of hits and divide by the number of times the hitter went to the plate.

                                            hits

Batting average =   –––––––––

                                    times at bat

Except that is NOT the formula for batting average. The numerator implies hitting the ball in play and ending up on base—except if that happens when there is a runner on first base who is forced out at second, then that is a fielder’s choice and not a hit. Similarly, if the scorekeeper determines one of the fielders made an error on the play, then that also would not count as a hit.

The denominator implies the number of times the player comes up to the plate to hit, but that is not what constitutes an official “at bat.” At some point in the history of baseball, certain outcomes such as walks, sacrifices and being struck by a pitch were not considered part of an official “at bat.”

Given those nuances, here is a more accurate formula for batting average:

                                                   hits – fielder’s choice – errors

Batting average =        –––––––––––––––––––––––––––––––

                                        plate appearances – sacrifices – walks – hit by pitch

There is a lot of complexity to the seemingly simple concept of a batting average, and most likely a lack of appreciation for the decision-making that determined errors should not count as hits or sacrifice bunts should not count as an official at bat.

I find this concept very analogous to recent conversations I’ve had about index methodology and construction. Often when discussing any of our index funds, clients will ask how they compare to the benchmark. One of the great financial advancements ETFs have provided over the past three decades is the ease with which investors can get tax-efficient access to popular benchmarks, for example, the S&P 500 Index for large-capitalization US equities.

The unintended consequence of that ease of access is that the ETFs that track popular benchmarks become the benchmark in the eyes of investors. If A=B and B=C, then A=C.

Expecting ETFs that track major benchmark indexes to be supplanted with ETFs that track a different index within the same asset class is unrealistic, given how entrenched those benchmarks are within financial markets.

However, a couple points are worth highlighting. First, if the decision to use a benchmark index ETF is based on its perceived simplicity, investors should appreciate that the actual index methodology is often quite complex—just as we saw with the batting average formula. For example, the S&P US Indices Methodology manual is 58 pages long and outlines the decisions that were made over time on eligibility criteria, market-cap thresholds, rebalance schedules, weighting caps and buffer percentages.

To choose an ETF that tracks a major benchmark index solely for the perceived simplicity of its methodology would be misguided. 

Second, if you are selecting a benchmark ETF simply because it is the benchmark, that would beg the question of why one particular set of complicated rules should have an outsized influence on asset allocation.

This leads to my final point on understanding the actual rules of index construction. Many fans of baseball like to use batting average as their means of assessing a batter’s worth. Similarly, many investors might like the rules of the major benchmark indexes of the world. However, we have seen a shift in baseball where now fans prefer different formulas when evaluating hitters, for example OPS+ (on-base plus slugging plus) or WAR (wins above replacement). In this same vein, I am seeing more investors ask about alternatives to benchmark indexes, whether that be subtle tweaks while still maintaining a low tracking error or a completely different set of rules (multifactor) designed to provide a specific investor outcome.

Over time, baseball fans have better understood the complexity and decision-making that created the batting average statistic, and are now evolving to think of new and better ways to measure the worth of a hitter. A similar awareness is happening for investors who are choosing index ETFs outside of those that track major benchmarks.