I propose three methods: the compensation cost approach, the lifetime income approach and the most popular, but arguably the least insightful, indicators approach.
An interesting and a challenging part of “S” in ESG relates to the question of how to measure the human capital of a firm. Increasingly, the Western world has outsourced manufacturing to Asia. Hence, physical and tangible assets are now less important compared to human capital, especially in software heavy and digital businesses.
The pressing concern on most CEOs’ minds these days is how to attract and retain human capital but U.S. GAAP (generally accepted accounting principles) treats labor as a “human cost” or “human expense.” Human capital, defined as the knowledge, skills, competencies, and attributes of the workforce, that enable the firm to earn higher operating and stock-based returns, is perhaps the largest unrecognized asset on a modern firm’s balance sheet.
Which leads to the inevitable thorny question: how does one begin to place at least a range of values on that asset? It turns out that such measurement is routinely attempted at the sovereign or the country level in their national accounts. Inspired by that work, I can think of three approaches discussed below.
Compensation cost-based approach
In this approach, the value of corporate human capital is measured as the cost of compensating workers over their tenure with the firm. One has to be careful to include all aspects of compensation such as salary, bonus, pension contributions, health care spending, perquisites, and, most importantly, the cash realized by employees when they cash out their equity or option compensation. The analyst will also have to add training costs, costs to make the workplace safe from both bodily and increasingly emotional harm, and efforts to enable employees in their job search and mobility, both inside and outside the firm.
One can get finer in this thinking by perhaps treating part of the explicit compensation, medical, health, safety, and training costs as “maintenance” human capital spending and hence as an expense whereas the “investment” portion of such spending can be viewed as the human capital asset that increases future productive growth.
The necessary condition to enable such estimation is that the SEC requires U.S. firms disclose labor costs in detail, as I have argued before. In the previous administration, the SEC came out with a few principles on human capital disclosure. However, actual disclosure has been limited and closer to the indicators approach discussed later with inevitable issues related to comparability. I am not sure that the disclosure of compensation costs improved after that rule. Approximate estimates of compensation costs of companies today can be compiled as a starting point to move this conversation forward.
Lifetime income-based approach
The idea here is to aggregate the discounted values of incremental future income streams that employees expect to earn throughout their lifetimes on account of the time they spend with a specific firm. As one can readily appreciate, the compensation cost-based approach focuses on the inputs side of the equation whereas the lifetime income-based idea considers the output side. It also follows, perhaps, that the cost-based approach will yield lower estimates of human capital than the lifetime income-based approach.
To make these ideas a bit more concrete, assume that the compensation of an entering MBA student into the business school is $60,000 per annum. Consider the average compensation of that student when she graduates, at say $150,000 per annum. Because the average MBA student is roughly 28 years old when she graduates, the student has a 37-year career ahead of her at the very minimum. The lifetime income based approach would suggest that human capital added by the business school is the present value of $90,000 over the next 37 years, discounted at, say, the average interest rate paid to borrow the student loan (roughly 5% today) is $1.5 million. I have assumed no growth in that $90,000 value add over time to be conservative in our valuations.
How would this work for a firm? The lifetime-income approach is very data intensive. The analyst should be able to compile salary and demographic data on the employees that joined. Similarly, using Linked In and the like, one can compile data on the positions that exiting employees land when they leave the firm and the associated compensation that they likely earn with the new employer.
Using the sample of past hires and resignations, one can assess the value added by the firm. Applying that value-add number to the existing employees is the next step. I have deliberately ignored finer questions such as self-selection. For instance, the employees who quit were either under-valued by the existing firm or let go for poor performance. In stage 2 of the analyses, one could decide how to weight the estimates by such self-selection considerations. Moreover, the estimates will have to be adjusted for the probability that the worker will leave the firm and the yearly rate of income growth and a discount rate appropriate to the firm. It may also be useful to consider the following age buckets of employees: 18-33 years, 34-50 years, 50-65, and 65 and above as rough thresholds around which careers evolve into different stages of a worker’s life.
Clearly, the lifetime income approach requires a lot of external and internal data. Another downside is that the lifetime approach arguably captures human capital added to the employee, as opposed to the firm. The obvious comeback is that the firm has potentially benefited more than the value the company added to the employee. Hence, the lifetime value added method potentially captures the lower bound of the labor market-based estimates of human capital for a firm.
Indicators based approach
This is perhaps the most indirect and the most popular approach in practice as we try to estimate human capital based on non-financial indicators of employee well-being such as abnormal turnover of workers from the firm, or retention of the work force, the extent of diversity in the work force, either in terms and/or ethnicity, or age decomposition of the workforce, and somewhat harder to measure attributes such as employee engagement, net promoter score of the CEO as perceived by employees, inclusion initiatives or even a measurement of the firm’s corporate culture and so on.
The obvious upside to this approach is the wealth of non-financial data about the well-being or otherwise of the workforce. Absence of a dollar number in this approach opens the Pandora’s box associated with a lot of ESG measurement- non comparable quantitative metrics that cannot be easily translated to dollar values.
For instance, it is tempting to interpret high turnover as a negative signal about the firm’s human capital. However, firms that are good at managing institutional memory, even in the presence of high turnover, likely have higher human capital valuations.
What about depreciation or appreciation in human capital?
Depreciation is particularly a concern with the compensation cost approach. The two obvious drivers involve age of the workforce and the potential for economic obsolescence of skill levels. Age is complicated in that unlike a capital asset, experience or passage of time might make human capital more valuable. Obsolescence of skill levels, which may be correlated with age, is a more serious concern. How many 60 plus year olds really under cyber risk, Artificial Intelligence (AI) and the like? Revaluations of the lifetime income-based approach will potentially incorporate any depreciation or appreciation in human capital by simply marking the human capital to market every year. Dealing with appreciation or deprecation of human capital with the indicators-based approach is, of course, difficult.
How to compute return on human capital for the firm?
This is a question I hear often from investors. Here are two tentative approaches:
Approach 1. Compute the portion of the firm’s costs that do not relate to human capital. For instance, that a software firm just went public and assume that the costs and revenues of a software firm, as reported in their income statement, are $100 and $150 respectively. Further assume that $60 of the $100 in costs relate to compensation paid to coders and the salespeople hired to sell the software and $40 for utilities, depreciation of office space or rent paid for premises. The CFO has determined that all of the labor costs constitute an investment. Let us say that non-labor costs in the income statement, as a first cut, are recovered via revenues on 1:1 basis with no mark up. Hence, $110 of revenues represent a payoff on $60 of human capital assets.
Approach 2: How does one compute a return on human capital with the lifetime income-based approach? I suspect that the lifetime income idea works better when there is a clear exit strategy, either via an acquisition or in a private equity transaction.
The closest analogy I can think of are transactions categorized as “acqui-hires” such as Cisco buying say a company of 400 engineers at say $1 million per engineer. Cisco effectively bins the software acquired but views $400 million as a reasonable consideration for buying a team of 400 engineers who can work together effectively. The actual compensation paid to that engineer every year may be say $300,000. Cisco pays a multiple of 3.3X on the compensation cost hoping that the value added by the engineering team to Cisco is far higher than $400 million perhaps in the three and half years of effective life assumed in the valuation multiple they paid.
A related idea is that private equity firms are increasingly getting into the business of finding undervalued firms that do not use their human capital well. A comparison of the entry and exit prices for such investments of private equity firms relative to the costs they incur to retrain or optimize the acquired work force would also shed some light on the return to human capital.
How to get there?
My goal was to simply sketch a few conceptual ideas on how to measure corporate human capital. There are several practical and conceptual barriers. Do HR groups of firms even collect these data? How will investors get a hold of such data?
I do not expect accounting rule makers to mandate placing internally generated human capital in the near future because of the usual concern that valuing internally generated intangible assets is difficult and unreliable. Plus, accounting theorists have argued that the value of the employee to the firm shows up in the income statement and a higher ROIC (return on invested capital), or in the value of the shares, ideally reflecting an expected high ROIC.
I could counter that ROIC could be high for many reasons and the point of fundamental analysis is to identify the source of the high ROIC. My objective, however, is not to necessarily fight the recognition battle with rule makers. As we wait for regulatory action, the idea is to enable investors flag stocks with high or undervalued human capital. Hence, investors will have to rely on big data or alternate data to form their estimates.
A resourceful analyst can get to approximations of compensation costs a firm pays. Computing lifetime income-based values from outside the firm is tough unless one can get detailed granular payroll data by employee or at least for a class of employees. Assuming that we can get decent data, acqui-hires and private equity transactions to optimize human capital deserve to be studied in depth.
I am eager to hear your reactions on this important question: how indeed should we measure human capital?