Author
Are stock prices based on financial fundamentals? Recent episodes of mania for “meme stocks” demonstrate extreme cases in which individual stocks soar and crash for reasons that seem far removed from the balance sheet. But for stocks in general, do prices primarily reflect a sober assessment of expected corporate earnings? This is the bottom line that should matter to any part-owner of a company with a long-term outlook.
Research over many decades, launched more than 20 years ago by Robert Shiller, supports a view that something else is the dominant force behind stock prices. Shiller observed in a 1981 paper that the large gyrations observed in historical stock prices display “excessive volatility” that cannot be justified by the subsequent stream of dividends that shareholders receive. If earnings can’t explain prices, this leaves a stew of other factors motivating investors’ actions. Herd mentality, fluctuating risk tolerance, and emotion might matter more to a stock price on any given day than a company’s financial statement and outlook.
Minneapolis Fed Monetary Advisors Jonathan Heathcote and Fabrizio Perri, with Bank consultant Andrew Atkeson of UCLA, take up this long-running debate on asset pricing and reach a different answer (Minneapolis Fed Staff Report 660: “There Is No Excess Volatility Puzzle”). They find plausible circumstances in which the expected stream of future dividends can, in fact, substantially explain the often-erratic movement in stock prices over the past century (Figure 1).
Among other insights, they find a rational basis for the “irrational exuberance” of the first dot-com boom, as dividends per share grew in the subsequent decades at rates that ultimately bore out the lofty stock prices of the year 2000. If this past is prologue, their conclusion also suggests recent record-high stock prices might be justified by expected profits in coming years.
The economists find support for their “dividends hypothesis” under two conditions: (1) Investors frequently receive unexpected information (“news shocks”) that affects their assessment of the long-run path of dividends, and (2) those investors have a low discount rate with regard to future dividends (that is, news about dividends far into the future still strongly affects the price that they are willing to pay for the stock today).
These conditions can lead to high volatility in an environment where future dividends are uncertain. There are multiple reasons for the uncertain path of dividends, including the overall path of corporate profits and unforeseen “corporate actions,” such as stock splits and buybacks. In such a setting, the economists find it would be natural for share prices to oscillate dramatically with news shocks—even as the observed path of dividends, after the fact, appears smooth by comparison.
Atkeson, Heathcote, and Perri develop forecasting regressions and test the dividend hypothesis using data from the CRSP Value-Weighted Total Market Index of stock prices from 1929 to 2023. The regressions adapt the basic approach of Shiller and co-author John Campbell, decomposing asset prices into a fundamental component (the discounted present value of the future stream of dividends) and an additive residual component. In this case, the economists posit that this residual term is constant and negative (meaning that it operates as a risk adjustment, depressing the price versus the risk-neutral present value of future dividends). In both linear and log-linear specifications of their model, they find that the residual component can remain constant while an underlying trend component of expected dividends—the measure of fundamentals—can largely account for prices-per-share during the past century of the U.S. stock market.1
The economists characterize their findings as far from the final word. They view their results as reopening the conversation about the role that financial fundamentals might, in fact, play in determining stock prices. “The answer one gives to the question of whether the stock market moves too much to be justified by subsequent movements in dividends really depends on parameters,” they write, homing in on the discount rate and the nature of the risk adjustment that investors apply to expected returns. Pending future research to pin down the values of these parameters, Shiller’s “excess volatility puzzle”—long a central paradox of financial economics—may prove less puzzling than thought.
Read the Minneapolis Fed staff report: There Is No Excess Volatility Puzzle
Endnotes
1 In recent decades, more companies—including some of the largest by market capitalization, such as Tesla, Amazon, and Alphabet—have chosen to pay few or no dividends to shareholders. In separate research in progress, the economists are exploring the implications of incorporating other ways of measuring value returned to shareholders, such as stock buybacks.
Jeff Horwich is the senior economics writer for the Minneapolis Fed. He has been an economic journalist with public radio, commissioned examiner for the Consumer Financial Protection Bureau, and director of policy and communications for the Minneapolis Public Housing Authority. He received his master’s degree in applied economics from the University of Minnesota.
