ESOPs, Broad-Based Stock Options, and the Stock MarketThere is considerable evidence that broad-based employee ownership, especially when combined with participative management, improves corporate performance (see this site's article on "Employee Ownership and Corporate Performance" for details). But does employee ownership in public companies lead to improved stock prices? There are reasons to argue both ways. Much of the research on corporate performance is of ESOPs in closely held companies. These plans usually own substantial amounts of stock (generally more than 30% and often more than 50%); provide significant annual contributions to employees (the median is in the 8% of pay range); are almost always in addition to, rather than in place of, compensation; and are much more associated with participative management techniques than are other forms of ownership. By contrast, ESOPs in public companies tend to provide smaller percentages of ownership (15% or less) in most cases, often are used to replace a match to a 401(k) plan, and are not significantly more associated with participative management than non-ESOP ownership. Broad-based stock options in public companies are also not more associated with participative management, and the level of benefits, while often significant, is not as large as that provided by ESOPs in public companies (the annual present value of options per employee in large public companies is in the $1,000 to $2,000 range, which translates into about 5% of pay or less). Finally, even if employee ownership does improve corporate performance in public companies, despite these differences, that does not mean that this will be reflected in stock price, especially if the plans result in additional dilution to shareholders.
On the other hand, even if public companies generally do not take full advantage of employee ownership as a management technique in terms of employee owner involvement programs, ESOPs and broadly-granted options (option plans in which at least half the full-time work force meeting minimal service requirements receives options) could help attract and retain good people and could serve to motivate people to make more effort at work.
Over the last several years, several studies have looked at the impact of employee ownership plans on shareholder returns. The results are generally positive, although specific causal linkages are difficult to prove.
ESOPsWe will focus here on studies from 1990 onward because in the 1980s, ESOPs were often associated with anti-takeover strategies in public companies. Thus it was impossible to distinguish the ESOP effect on share price from investor concerns that an ESOP would make a takeover more difficult, dragging down stock prices. The largest study on ESOPs in public companies was performed in 1999 by Hamid Mehran, then of Northwestern University for Hewitt Associates. He found that ESOPs in 382 publicly-traded companies increased the return on assets (ROA) 2.7% over what would otherwise have been expected. The study looked at the companies' financial returns for two years prior to the plan's implementation and four years after. On return to shareholder issues, this study only looked at initial shareholder reaction to ESOP announcement. Over 60% of the companies experienced an increase in their stock price in the two-day period following public announcement of the ESOP, with the average increase for all companies at 1.6%. In 1992, Douglas Kruse and Joseph Blasi of Rutgers University, and Michael Conte of the University of Baltimore, created an "Employee Ownership Index" (EOI). The EOI tracked the average percentage increase in stock price of all publicly traded companies with a public record of 10% or more broad employee ownership (usually ESOPs) and more than $50 million in market value. The EOI was subsequently maintained for some time by American Capital Strategies, an investment bank based in Bethesda, MD, and was published quarterly in the NCEO's Journal of Employee Ownership Law and Finance. The EOI grew 193% from 1992 through 1997, while the Dow was up 145% and the S&P 500 140%. The authors did not attribute any causal relationship to these numbers, however.
Stock OptionsOverall, the data on stock options indicates that broadly granted plans improve stock returns. This is in contrast to a massive amount of data looking at grants focused on executives, which find generally neutral or negative results.
The first assessment of the impact of broad-based stock option plans on corporate performance was done in 2001 by Douglas Kruse, Joseph Blasi, Jim Sesil, and Maya Krumova of Rutgers University ("Broadly Granted Stock Options Improve Corporate Performance," published in Stock Options, Corporate Performance, and Organizational Change, NCEO, 2001). The study looked at 105 publicly traded companies with usable data from 1992-1997. For the corporate performance part of the study, companies had to provide options to at least 75% of full-time employees. The research showed that these option programs appear to increase productivity and return on assets. Total shareholder return, however, showed no statistically significant difference in the relative performance during the two periods, meaning any measured change could simply reflect random sampling error. The researchers thus argued that in a cause-and-effect sense, any value consequences of dilution caused by broad-based options seems counterbalanced by increased productivity.
However, when the researchers looked not at the difference between "before" and "after" the option plans, but just how a basket of investments in companies with broad options would perform compared to a basket of other investments, they found a different pattern. Here they looked at a broader group of companies with options, including any with more than 50% of non-management employees getting grants, those with less than 50% of non-management employees getting grants, and those with either no plans or plans just for management. Over the 1992-1997 period studied, they report that:
"All Compustat non-stock option firms had average individual company cumulative returns of 193.1% (81.8% at the median) while the Compustat 500 had returns of 275.0% (151.7% at the median). All broad-based stock option companies had average individual company cumulative returns of 303.2% (163.9% at the median) over the period. This return was statistically significantly greater than the return for all non- broad-based stock option companies. Broad-based stock option companies with more than 50% of non-management actually receiving grants had average individual company cumulative returns of 232.6% (108.9% at the median). And broad-based stock option companies with less than 50% of non-management actually receiving grants had average individual company cumulative returns of 318.9% (128% at the median). On a 1992-1997 cumulative basis using this measure, the average individual company cumulative return and the median individual company cumulative return of the broad-based stock option companies surpassed that of the non-broad based stock option firms."These results suggest that while broadly-granted grants cannot be proven to be the cause of improved shareholder returns, an investor would be wise to invest in companies with these plans nonetheless because having such a plan seems indicative of better performance for reasons that are as yet not clear.
In a 2003 study by Watson Wyatt ("Stock Option Accounting and Total Return to Shareholders: Has the Market Already Factored an Expense Into Today's Stock Price"), the researchers (who were not identified) looked at 800 companies drawn from the S&P 1500 that had footnoted options expenses. While the report focused on shareholder reactions to options expensing, the authors of the study also found that "over the last three years, for the options that are allocated to broad employee groups [this term was not defined by the study], an increase of one percentage point (i.e., from 78% of the options to 79% of the options) is associated with a nearly one percent increase in the firm's market value. This is equal to nearly $15 million in market capitalization for the median firm in our sample." As a result, the authors concluded, "companies with high levels of participation in stock option plans create greater shareholder value than companies with low participation."
Finally, the market seems to recognize this effect. In a 2003 study by Eric Hager of the University of British Columbia, shareholder reaction to announcements of stock options grants in Canada and the U.S. was analyzed to determine how shareholders reacted to announcements for broad-based options grants. Announcements were excluded that did not indicate employees were simply eligible for stock options, but were included if employees only or employees in addition to managers, executives, directors and/or consultants actually received grants. The study, "Do Employee Stock Option Grant Announcements Affect Shareholder Value?", was published in the summer 2003 issue of the NCEO's Journal of Employee Ownership Law and Finance.
For companies to be included in the study, there had to be sufficient trading data available for the period from 120 days prior to the announcement to one day after. There also had to be a clear announcement of actual grants being made, rather than the announcement of a stock option plan. Hager points out that the announcement of a plan can be misleading because a plan may make a certain number of employees eligible for a grant but actually only make grants to a much smaller number. So Hager instead looked at actual grant practices. A total of 54 announcement grants were suitable for inclusion in the event study in the U.S. The U.S. grants were from 1993-2002. Hager used a standard event-study methodology to extract returns greater or less than what would have been predicted in the day following the announcement based on how a model accounting for how other companies in the industry performed.
For U.S. companies, Hager found that returns were up 1.78% over what would have been expected for broad grants when more than 1% of the equity was granted, but there was no significant relationship when less was granted. The results were not significantly affected by looking at grants only after September 1, 2000. The findings for the broad-based grants were statistically significant (that is, not likely to be a result of random variation). The results for Canada were even stronger.