Stock Investors’ Returns are Exaggerated

Long-run buy-and-hold stock market returns with dividend reinvestment (“total shareholder returns” or “TSR”) are significantly higher than risk-free returns, leading to an annual equity risk premium of about 6%. Stocks are thus pitched as a key component of wealth accumulation strategies and have become an important savings vehicle for American families. The high equity premium has also led policymakers to consider investing Social Security funds in the stock market.

In a paper recently posted on SSRN, Stock Investors’ Returns are Exaggerated, we explain that the stock market generates much less wealth for investors as a group than it appears. While individual investors who hold stock and reinvest dividends (“TSR investors”) earn the equity premium, investors as a group cannot. The main reason: TSR requires dividend reinvestment in shares previously held by other investors, which is by construction impossible for all investors to achieve. As a group, investors cannot plow dividends back into public firms, and must invest them in lower-yielding assets. Thus, every TSR investor who earns the equity premium necessarily involves another investor not earning that premium. The same holds for cash distributed via net repurchases (stock buybacks less issuances), which now exceed dividends. In addition, TSR is boosted by well-timed repurchases and equity issuances that merely transfer value from trading to continuing stockholders. The returns for stock investors collectively therefore must be lower than that implied by TSR, as public firms distribute considerable amounts of cash and engage in market-timed equity transactions with their own shareholders.

To measure aggregate shareholder returns and the effective premium enjoyed by the typical stock investor, we propose and implement a new approach: “all-shareholder return” or “ASR.” Using this ASR approach, we estimate that TSR overstates the effective premium for stock investors (the “ASR premium”) by between 21% and 268%, depending on alternative investment options. Most investors cannot, and do not, earn the 6% equity premium available to TSR investors.

While we are not the first to stress the importance of measuring aggregate shareholder returns, prior research has employed an internal rate of return (IRR) methodology that has well-known and significant conceptual and computational drawbacks. To begin, IRR is generally not a valid measure for the effective rate of returns earned by investors unless net cash distributions are reinvested at the IRR rate, a restrictive condition that cannot be satisfied for stock investors as a group. In addition, IRR can produce multiple or no real solutions, be sensitive to measurement windows, or be subject to “hindsight bias.” Thus, the actual return obtained by all stock market investors remains an open question.

Our ASR approach avoids the problem with IRR by explicitly taking into account actual reinvestment possibilities for net cash distributions to investors. As a first and partial step, we examine the effect of relaxing the dividend-reinvestment assumption embedded in TSR, and assume that buy-and-hold investors invest dividends in alternative assets—treasuries of different maturities, corporate bonds, and housing—instead of the stock market. On this modified TSR approach, the premium earned by market investors is between 14% (assuming dividends reinvestment into housing) and 51% (assuming dividends reinvestment into 1-month treasuries) lower than the TSR-implied equity premium.

We then integrate into the analysis repurchases and equity issuances, using a modified IRR (mIRR) approach that circumvents IRR’s shortcomings and that has been applied to measure returns of private equity. We find that the ASR premium is 17%-73% lower than the TSR-implied equity premium or, equivalently, that TSR overstates the effective premium for stock investors by 21% to 268%. These findings are robust in the three 30-year sub-periods in our sample.

Finally, we analyze the sources of the ASR-TSR gap. The main reason why ASR is significantly lower than TSR is due to the reinvestment effect: large cash payouts by public firms that, from shareholders’ collective perspective, can only be invested in other assets (such as bonds) that tend to yield significantly lower returns. We estimate that 70-92% of the ASR-TSR gap is due to this reinvestment effect. We also find that the timing of cash flows, either with respect to TSR or alternative investment asset returns, explains between 8-32% of the ASR-TSR gap. And, over time, the reinvestment effect’s contribution to the ASR-TSR gap declines while the market-timing effect becomes more important, consistent with the increasing prevalence of stock issuances and buybacks. Nevertheless, the reinvestment effect is the predominant effect in each 30-year sub-period in our sample.

Our findings contribute to the literature on the returns of stock market investors as a group. Like prior work, which uses an IRR approach, we find that aggregate shareholder returns are significantly lower than TSR suggests. But we argue that our mIRR approach is more appropriate and robust measure of value creation by the stock market. In addition, while prior work suggests that the difference between aggregate shareholder returns and TSR is likely driven by the market timing of equity issuances and buybacks, which tends to shift value from trading investors to buy-and-hold investors (including TSR investors), we show instead that the main driver is non-TSR investors’ need to reinvest the cash in lower-returning assets.

Our work is also connected to the literature on market timing by firms issuing or buying shares. Prior work estimates that the wealth-transfer effect of repurchases and equity issuances each year is about 40% of net income, transfers that can be expected to accrue to the benefit of long-term investors. Our work shows that such wealth transfers among investors do in fact boost returns for TSR investors and contribute to the gap between TSR and ASR.

Finally, our findings contribute to various policy discussions about the role of the stock market as a savings and wealth creation vehicle. First, while wealthier families are more likely to invest in equities, our work suggests that the stock market does not contribute as much to income and wealth inequality as TSR returns may suggest. Second, our work has implications for the debate over the potential role of the stock market in buttressing the Social Security Trust Fund. It suggests that returns will not be as great as might be expected unless the Trust Fund sticks to a TSR strategy; but if it does so, its outsize returns will simply come at the expense of other investors.

The full paper is available here.