Machado and Harper haven't signed because baseball teams are now run like Wall Street 'quant funds'

This post was originally published on this site

The market for pro baseball talent has been ice cold this winter.

With spring-training camps opening this week, some 100 Major League free agents remain unsigned — including elite young hitters Bryce Harper and Manny Machado, who each were projected at the end of the 2018 season to be headed for ten-year contracts worth $300 million or more.

There are almost as many factors being blamed for the ballplayer bear market as there are jobless free agents. Some players and commentators accuse team owners of active or tacit collusion, withholding offers to depress salaries. Houston ace pitcher Justin Verlander this week ripped owners’ lack of interest in Harper and Machado, both just 26 years old.

Another camp points to the bounty of national media money guaranteed to each team, which makes ticket sales less crucial and dulls demand for fan-pleasing marquee signings. “Tactical tanking” has become more popular, too, with teams choosing to endure a few years of bad records to rebuild with draft picks, following the pattern of recent World Series winners the Royals, Cubs and Astros.

And then there is baseball’s “analytics revolution,” which is following a similar path to Wall Street’s embrace of “quant” investing in recent decades.

At the core of both movements is an ability to gather and slice and crunch data to isolate key performance factors. Quant investors seek to create a portfolio using models blending several factors (value, momentum, quality, analyst estimate revisions) that leaves them agnostic on the stocks themselves and typically has them underweighting the largest index components and hottest glamour stocks — which might be great companies but whose valuation already builds in years of great future fundamental performance.

The new breed of baseball general managers and their extensive analytics departments do the same. They view a team roster as a portfolio, and each player the sum of several performance factors. These are physical skills and statistical tendencies that can be blended countless ways to create a team “portfolio” with a certain “expected return,” often without paying top dollar for a “sure thing” or a big name.

Quant investors are mindful of “dispersion” of price returns and some funds actively bet on such divergences evening out over time. Bradford Doolittle at points out that researchers have studied the dispersion of salaries on a baseball team: “The core question is whether it’s better to have a roster where player salaries are all kept within a certain range (low dispersion) or if teams are better off concentrating salary on one, or a handful, of star players atop the roster (high dispersion — the ‘stars-and-scrubs’ approach). Pretty much all of those kinds of studies I have seen suggest that low-dispersion teams tend to outperform high-dispersion teams.”

Former all-star first baseman Mark Teixera, now an ESPN analyst, recently cited “analytics” as the reason teams have come up short of Harper and Machado’s asking prices. The statistical probabilities say players peak around age 32, so teams are hesitant to commit to, say, $30 million a year through age 36 for these guys.

Of course, team owners always knew there was risk in long-term contracts and that players later in a deal were extracting above-market value, but they viewed it as a price worth paying for putting the best available stars in their uniform.

The difference these days: A fixation on the analytical models gets in the way of such ego-driven or crowd-pleasing impulses.

A history of quantitative investing by 361 Capital notes. “With factor investing as the basis for quantitative investing, these strategies are built with the goal of helping investors avoid common behavioral pitfalls; designed to remove the emotional input from the investing process. Leveraging a repeatable process, quantitative approaches can be used to improve investment decisions by producing objective analysis which illuminates elusive, but repeating, historical patterns.”

To a quant, Harper and Machado are like, say, Netflix or Amazon or PayPal – great growth properties that should do well for years to come – but whose current price already builds in all those years of promised (but not wholly certain) future performance.

Baseball’s shift to a more hard-headed, numbers-based approach to talent evaluation was famously chronicled in Michael Lewis’s 2003 book “Moneyball,” the story of how GM Billy Beane exploited other executives’ behavioral blind spots to find undervalued players who were good at things competing teams didn’t properly appreciate.

But the big-data era of baseball quant investing is now in full swing, enabled by a system of radar and high-powered cameras that track every pitch and player movement with astonishing precision. No longer does a pitcher simply have “a lively arm with good late movement.”

Statcast tells us the spin rate of the ball exiting his hand and how much it moves laterally or vertically on the way to the plate.

Hitters don’t simply have “good bat speed and nice pop,” but Statcast yields “exit velocity” of a batted ball, its “launch angle” and the probability that it will result in a base hit or a home run.

In a similar way, quantitative-investing models set aside the traditional human analyst’s qualitative search for “great, well-managed companies with a strong growth record” in favor of relative-valuation scores, a stock’s volatility profile and a stress-tested forecast of margin resilience. Quant investing exploded once technology advanced to provide tick-by-tick price data and intense performance-attribution studies.

In general, strict quant investors will own fewer of the largest, most expensive stocks that dominate the big indexes and gravitate toward under-owned names; will frequently rebalance their holdings to keep their factor exposures consistent; and will cut losses ruthlessly through strict risk-management rules.

Baseball’s quants are following along a similar path. Each player’s statistical output is compressed into “wins above replacement” – a calculation of how many wins he accounted for, above a hypothetical marginal Major League prospect. Machado is near the top of FanGraphs’ projection for 2019 in WAR at between 5 and 6 wins. That’s a fabulous total, but also a target for a front office to try to approximate with other “portfolio components.”

Being able to identify the opportunity cost of not having Machado enables baseball quants to try to gather up pieces of that forgone 5 or 6 wins through younger, cheaper relief pitchers with a high swing-and-miss rate, say, or a couple of hard-contact, high-on-base-percentage hitters about to enter their prime and not yet eligible for free-agent riches.

In other words, when the performance factors are quantified and treated as fungible, players can be mixed and matched in a cost-effective way to maximize win probabilities.

Not exactly the romantic approach that prevailed when most old baseball fans fell in love with the game – and it’s no guarantee of on-field success. But perhaps an inevitable result when big data and big money collide in the big leagues.

Of course, investing and baseball are different in one crucial way. In markets, return on the money invested is literally the entire game. In baseball the point is to win the most games, not get the most wins for payroll dollar. Perhaps that point is being lost by some value-obsessed teams.

Add Comment