Workforce Diversity, Stock Returns, and the Corporate Thesaurus

Does workforce diversity predict stock returns? A closer look at the data.

Investors have long sought to understand the relationship between workforce diversity and corporate financial performance. Numerous studies have found positive associations, particularly between management diversity and measures of business performance.

But does workforce composition also help explain subsequent stock returns?

PeopleReturn has spent years assembling historical EEO-1 disclosures: more than 13,000 reports covering 2,168 companies. We recently used these data to investigate that question across the S&P 1500.

The initial results were surprising: companies with higher female management representation subsequently underperformed on a risk-adjusted basis. But the relationship weakened under several robustness tests, raising questions about what the measure actually captures. 

First, an important caveat: This is not a backtest of a tradeable investment strategy.

Most of our 2016โ€“2020 data became public through FOIA releases beginning in March 2023, while company disclosures typically lagged certification by 6โ€“18 months. This fragmented and inconsistent availability makes the analysis unsuitable as a historical trading strategy. The window is also short: nine reporting cohorts (2016โ€“2024), followed by stock returns through 2025.

What we analyzed

Our S&P 1500 panel contains approximately 6,400 company-year observations. We examined four measures: women and people of color (POC) as percentages of the overall workforce and of management.

For this analysis, management combines the EEO-1 executive/senior and first/mid-level management categories. Companies have discretion in how they classify positions between these levels, so combining them reduces one source of inconsistency across companies and over time.

Each year, we ranked companies into quintiles within their GICS sectors and compared the following calendar year’s returns for the highest-representation quintile against the lowest, giving sectors equal weight. We then adjusted returns for the Famaโ€“French five factors plus momentum.

What we found

MeasureAnnualized risk-adjusted
Q5โ€“Q1 return
p-value
Women in managementโˆ’4.1%0.027
Women in workforceโˆ’4.4%0.075
POC in management+0.0%0.988
POC in workforce+0.2%0.914

Equal-weighted, sector-balanced portfolios, 2017โ€“2025. Negative values indicate lower subsequent returns among companies with higher representation. Historical associations, not investable strategy returns.

Female management representation showed a statistically significant negative association with subsequent risk-adjusted returns in this specification. Neither POC measure showed a statistically significant association.

The raw return spread for female management representation was negative in all 11 GICS sectors. Five sectors (Industrials, Financials, Information Technology, Consumer Discretionary, and Health Care) had at least 10 companies in both extreme quintiles every year, and all five showed negative spreads.

That consistency warranted further investigation, but not a conclusion.

How robust was the result?

Time period. The relationship was concentrated in 2021โ€“2025, at โˆ’5.1% (p=0.012). The 2017โ€“2020 estimate was โˆ’0.9% and statistically insignificant.

Portfolio concentration. In a market-cap-weighted portfolio restricted to sector quintiles with at least 10 stocks, capping any single stock at 10% reduced the 2021โ€“2025 alpha from โˆ’8.4% to โˆ’3.7%, and the result lost statistical significance.

Disclosure continuity. Our panel fell from 894 companies in 2020 to 562 in 2021. Restricting the analysis to 522 companies with observations in both reporting periods and eligible extreme-quintile returns, the 2021โ€“2025 equal-weighted estimate remained negative at โˆ’5.2% (p=0.043). The market-cap-weighted estimate was โˆ’4.8% and not statistically significant (p=0.125).

Multiple testing. We examined dozens of specifications across measures, weighting approaches, samples, and periods. Any single p-value should be interpreted in that context.

We do not regard the result as a validated investment signal.

Does the definition of management diversity matter?

A company with 60% women in management and 70% in its workforce ranks above one with 40% and 30%. Relative to its workforce, though, the second company has greater female management representation.

We therefore tested the management-to-workforce ratio and the percentage-point gap for both women and POC.

None of these four relative measures produced statistically significant risk-adjusted returns. Full-period estimates ranged from approximately +0.0% to +1.5%, with p-values between 0.35 and 0.99.

That doesn’t prove workforce composition explains the original result. The samples differ, and some recent-period estimates remained negative. But it raises an important question: how much of the absolute-representation result reflects occupational mix, business models, or other company characteristics rather than management diversity itself?

How does this compare with previous research?

PeopleReturn, formerly DiversIQ, contributed EEO-1 data to As You Sow’s 2022, 2023, and 2026 studies. Those studies identified positive associations between management diversity, particularly racial and ethnic diversity, and financial outcomes including revenue growth, profitability, return on equity, return on invested capital, and long-term share-price appreciation.

The two efforts address different questions. As You Sow examined a broad range of operating and financial outcomes, including longer-term share-price appreciation. Our analysis focuses specifically on subsequent one-year stock returns after accounting for sector exposure and established risk factors.

These findings are not necessarily contradictory. A characteristic associated with stronger operating performance need not produce higher subsequent risk-adjusted returns, particularly if its expected benefits are already reflected in market valuations.

Our POC measures showed no statistically significant association with subsequent risk-adjusted returns in either direction.

For investors, the lesson is straightforward: the availability of workforce data is only the beginning. How we define, contextualize, and test those data matters just as much.

Get PeopleReturn insights in your inbox

Layoffs? We prefer โ€œcareer transitions.โ€

Corporate America has found a new way to reduce headcount: stop calling it layoffs.

This week, Microsoft offered voluntary buyouts to roughly 60,000 employees, about 27% of its 223,000-person workforce. The actual number leaving will be considerably smaller. About 30% of eligible employees took the companyโ€™s spring retirement offer, and Microsoft expects a significantly lower rate this time. Even at 15%, thatโ€™s 9,000 departures, or 4% of the workforce.

The program is called the โ€œVoluntary Exit Offer,โ€ and Microsoft says it gives employees more ways to โ€œshape their next chapter.โ€ The package matches Microsoftโ€™s standard severance. The main difference from a layoff is who makes the call.

At 9,000 departures, that’s roughly $1.9 billion in annual compensation using Microsoft’s disclosed median pay. For comparison, Microsoft reported $45.9 billion in operating income and $45.2 billion in capital expenditures in the latest quarter alone.

It follows a spring Voluntary Retirement Program and 4,800 job cuts in July, all while Microsoft plans $175 billion in capital spending this year. Retirement in the spring, Exit in the fall. We look forward to the winter collection.

Meanwhile, just three weeks after closing its roughly $110 billion acquisition of Warner Bros. Discovery, Skydance offered some employees a one-time buyout and announced a five-day return-to-office requirement. โ€œNot everyone will see their future here,โ€ the co-CEOs wrote, โ€œand we respect that.โ€

When the merger was announced, the companies said their film strategy would drive โ€œlong-term job growth across the film and creative industries.โ€ The combined company also carries roughly $87.5 billion in debt,  with nearly $6 billion in projected annual interest expense, roughly what its cost-savings target is meant to save. The job growth, presumably, comes later.

Theyโ€™re hardly alone this year. American Express is cutting 3,500 jobs to become โ€œa more agile and efficient organization.โ€ Amazon eliminated 16,000 positions while โ€œreducing layersโ€ and โ€œremoving bureaucracy.โ€ Intuit cut 17% of its workforce in the name of โ€œvelocity.โ€

Thereโ€™s a catch to letting employees decide who leaves: the people most willing to go may be the ones you most want to keep. A University of Pittsburgh study of more than 3 million LinkedIn profiles found turnover rose 14% after S&P 500 tech and finance companies introduced office mandates, with senior, skilled and female employees leaving at higher rates. Buyouts and office mandates are different tools with the same risk.

To be fair, voluntary departures are preferable to involuntary layoffs for many employees, and companies have legitimate reasons to restructure. Whatโ€™s increasingly creative is the vocabulary.

Whatever the terminology, the objective is familiar: fewer employees on the payroll. The corporate thesaurus, unlike the workforce, continues to grow.


In other human capital news…

  • C.E.O.s Say the A.I. Rollout Is Going Well. Employees Feel Differently. Corporate leaders, increasingly in earnings calls and interviews, tout productivity gains and say they are doing right by their workers as they roll out the technology. But workers are much more concerned about being supplanted, and they donโ€™t trust their companyโ€™s executives. (New York Times)
  • Visibility Counts 2026: The Rollback and the Road Ahead Nearly nine in ten companies that named LGBTQ+ people in 2024 removed it in a single filing season. The business case hasnโ€™t changed. The pressure has. (Out Leadership)
  • U.S. Employee Engagement Below 30% for First Time in 15 Years U.S. employee engagement fell to 29% in the third quarter of 2026, its lowest level since 2011 and down two percentage points from the previous quarter. Hybrid and remote employees in organizations adopting AI experienced the sharpest declines. (Gallup)
  • The New Question for C-Suite Bosses: Whoโ€™s in Charge of AI? AI stands to transform businesses, and the executives who spearhead that change put themselves at the center of a companyโ€™s success story, and set themselves up for higher-profile roles down the road, while those who fumble an expensive AI rollout could get the boot (Bloomberg)
  • Too much disruption can affect workplace culture Companies that said their top challenge was โ€œtoo many simultaneous changesโ€ were 24% less likely to believe they were good at building culture compared to firms that cited other obstacles (HR Dive)
  • Empower Your Most Curious Employees to Drive Your Digital Transformation Leaders seeking to turn AIโ€™s promise into business value often overlook employees already experimenting with emerging technologies. These โ€œdigital wayfindersโ€โ€”frequently younger, junior employees but defined by their exploratory approach rather than their ageโ€”can reveal new possibilities, challenge established assumptions, and accelerate digital transformation. (Harvard Business Review)
  • How Trade Desk Rescued Executivesโ€™ Stock Options From a Dive in the Share Price Trade Desk shareholders have contended with a 66% slump in the stock price so far this year. This week, Chief Executive Jeff Green gave himselfโ€”and many of his employeesโ€”a do-over.ย  (Wall Street Journal)
  • Some 75% of Jewish employees find workplace DEI approaches inconsistent on antisemitism A new workplace study published on Tuesday has revealed that 75.5% of Jewish employees feel that Diversity, Equity, and Inclusion (DEI) approaches within their organizations fail to respond consistently to antisemitism. (Jerusalem Post)

Need human capital data for your planning?