How much should companies invest in their employees? Probably a lot.
To what extent does a company’s investment in their people produce better company and stock results? Can we earn excess returns by buying stocks of companies that invest heavily in enhancing their “human capital”? If we can, the strategy should work, because it’s not widely utilized and it’s hard to do.
But answering that question is an empirical and philosophical challenge. The Harvard Business School professor Ethan Rouen thinks that we can invest profitably using this information, and (with co-author Matthias Regier) has written a paper that provides strong support for his hypothesis. (Rouen is a professor of accounting, not finance, so he brings a fresh perspective to investment questions.)
Very large excess returns
I recently spent some time with Ethan and discussed some of his findings. His research suggests that:
A long-short investment strategy, going long the firms that invest the most in their employees, and short the firms that invest the least, would generate annualized excess returns north of 7%. The strategy worked great.
An excess return of 7% per year is a huge number but, as Rouen graciously acknowledges, this is a backtest: “We're cheating because we're using historical data.” The strategy would have generated that result over 1991-2018 if a manager had selected stocks from a universe of 11,569 names in 30 European countries.[1]
Will the strategy work in the future? It will if current market participants have not already discovered and exploited the effect. Rouen argues that they have not: despite decades of warnings from academics and others, most companies still take a very short-term strategic view by focusing on current cash flows and earnings as opposed to the longer-term benefits of investing in intangibles, such as human capital, that do not show up on the balance sheet and do not generate (visible) income in the current quarter or year. In fact, more often than not, companies and the analysts that follow them adopt the opposite viewpoint says Rouen.
He said…..
“Other market participants don’t seem to be getting this. We find that the more companies invest in the future value of employees, the more pessimistic analysts become. They’re not realizing that [investing in employees’ human capital] reflects a future investment as opposed to an immediate expense.”
Because there is no line item in an income statement called “investment in employee human capital,” teasing that variable out of available data is a neat trick, which Rouen explains at great length in the paper. I won’t attempt to summarize the explanation, but the end result is that the authors divide personnel expenditure into two parts which I’ll call consumption and investment. The consumption part of personnel expenditure is the bare minimum amount necessary to attract and retain workers in the current time frame and does not enhance human capital (and thus creates no excess returns if Rouen’s basic thesis is valid).
The investment part is the rest of personnel expenditure. If a company invests in its people efficiently, and that information is not already in the stock price, then the stock is likely to generate excess returns. It also just feels good to invest in companies that are taking good care of their people. How could that be bad? It can’t be – all else equal. But things never are equal. A company has only so much capital and investing in one’s employees has to be weighed against investing in other necessities.
The key to investing in people is doing so “efficiently.” Inefficiency can take many forms. A company may be training employees for their next job instead of the current ones, or trying to corner the market in a particular type of employee by overpaying them. (We sometimes see this in tech companies.) The risk of inefficient investments in people may be the source of analysts’ pessimism about such firms. It’s like research and development: – there’s R&D that adds value and there’s R&D that isn’t worth the money. And companies can’t know in advance which is which. But happily, that’s not true about most investments in human capital: with enough attention to company strategies and human psychology, management can usually identify efficient ways of spending their money on their people.
At any rate, Rouen shows that, when you aggregate across all these companies, non-U.S. in his case but almost certainly in the U.S. as well, the payoff to investing in one’s employees is positive and reflected in future excess returns.
What is human capital and who gets the “return”?
Human capital as a concept is three centuries old. Adam Smith, without using the term “human capital,” described it as a factor of production consisting of the “acquired and useful abilities of all the inhabitants or members of the society.” In the early years of the 20th century, the English economist A. C. Pigou connected it to investment in a 1928 book, writing “There is such a thing as investment in human capital as well as investment in material capital.” American economist Gary Becker, in the 1960s, popularized the term and made it part of ordinary discourse.
But what exactly is it? What aspect of humans is their “capital”? If it’s literally all of someone’s “useful abilities,” which is the currently accepted definition, the “return” on that capital gets split among the capital owner (the employee, as wages), the employer (shareholders, as profits), and the government (as tax revenue). Setting aside the government share, the split among employer, employee and shareholder is determined by the usual labor-economics factors: scarcity of a given skill, the level of unemployment, labor mobility, and so forth. But we can be certain of one thing: the employee cannot get it all, because the employer would not hire him or her if there were no profit to be made from that decision.
The company in the awkward position of wanting to enhance the employee’s human capital so it can get more out of him or her but also knowing that this investment will also result in the employee becoming more attractive to competing employers. And, under current laws, it is impossible to keep an employee from quitting and difficult to keep him or her from going to work for a competitor. A company must decide how much to invest in its employees based on these conditions and risks.
As noted earlier, Rouen finds that stock analysts disfavor a large amount of employer investment in human capital – but the stock market favors it! Therein lies the arbitrage behind Rouen’s 7% long-short excess returns.
A garden of lemons
I’ll expand on this tension between employer and employee by recalling what another Harvard professor, Boris Groysberg, a very close friend of mine, said some years ago. I told him I was investing a lot of money in the people who work with me. At the time I was leading a small, fast-growing company and investing in people felt like the right thing to do. However, it was frustrating that sometimes my well-trained employees would leave three or four years later. I tried my best to keep people as long as I could, but they would come and go. (And some of them came back.)
Groysberg responded, “what if you don’t invest and they stay?” Touché. If you don’t invest in your people, you may have cultivated a garden of lemons, who ruin your reputation and eventually ruin your business. He concluded by saying that “When people leave, the part that stays, which is the culture that has been built with this person – a culture of learning – is equally if not more valuable, but also much harder to measure.” However, we have to measure it if we are to build stock portfolios based on this information.
Accounting principles don’t help much here. Ethan said, “Trying to value a modern company using its financial statements is like trying to repair a Tesla Model Y using a manual for a Ford Model T.” Academics and accounting professionals are looking into new ways of valuing today’s human capital-intensive firms. They focus on people, culture, reputation, and brand awareness, and they talk about things like “moats” that Ben Graham and David Dodd (fathers of value investing) might not recognize.
Confounding variables
What makes it especially difficult to root out the profits or equity returns produced by investing in people is the presence of confounding variables. These are variables that are not directly observed but that affect both the variables you are trying to explain and the ones you are trying to explain them with. This makes it hard to determine the trajectory of causation in any relationship.
In the current example, companies with large profits may be profitable because they invest in their people – or they may have extra money to invest in their people because they are profitable.
If we were going to invest client money in a human-capital strategy, we’d want to work out those relationships carefully first.
Listen to our full conversation here: Ethan Rouen: Quantifying Human Capital
[1] The EU 27 (as of early 2023), plus Bulgaria, Croatia, Cyprus, Estonia, Latvia, Lithuania, Montenegro, Romania, Slovenia, Slovakia and the UK.