Reconsidering Returns

Research Retrieved: January 2019
Retrieve the paper from the SSRN

Issue

Asset-pricing models usually assume that all investors understand and observe performance as returns, made up of the capital gain, or change in price of the asset, as well as cash flows like dividends or coupons. The paper at hand provides evidence that this assumption does not hold. Many brokerage statements, financial newspapers and financial websites report price changes as returns, resulting in a market-wide nudge that causes investors’ perceptions of performance to be biased. Many market indices that are used today, such as the S&P500, were constructed in the early 20th century, when academic articles generally assessed prices and dividends separately, with the indices designed to capture only the price component. This feature is still retained today. The market seems to be confused about the implications of this, which affects, for example, flows to mutual funds, the tone of journalists’ reports on market performance and the predictability of short-term market price movements.

Key Findings

  • Data on returns of major stock indices is difficult to obtain and rarely displayed by default: Returns of major indices are not displayed according to their almost universally agreed upon definition. That is, displays of performance include only price changes, while not taking into account cash flows such as dividends or coupons. This is the case for a majority of brokerage statements, as well as for financial newspapers, such as the Wall Street Journal and for financial websites, including for example Bloomberg or Morningstar. Instead, investors that are looking for actual returns measures can only make approximations based on manual calculations.
  • Confusion resulting from this misreporting of performance can be observed in the use of language in newspaper articles: Because of the misreporting of financial returns, stock indices underperform on ex-dividend days by construction. Newspaper articles are overly pessimistic on these days, which shows that price indices are mistaken for returns.
  • Confusion resulting from this misreporting of performance affects investor behavior: The display of performance in terms of price changes, rather than returns with dividend reinvestment, may help in explaining behavioral biases in investment decisions. For example, if investors view the performance of their investments relative to a price index, rather than a return index, they will not receive feedback that failing to reinvest dividends could lead them to underperform a return benchmark. This could help to explain investors’ tendencies to not reinvest dividends and view dividends as a free source of income, separate from the price level of the security. Similarly, the equity premium puzzle, the fact that equities outperform bonds by a high margin, has been the source of much debate in finance. While by no means a full solution, if investors are viewing the performance of equities in terms of price indices, they will perceive less outperformance of stocks relative to bonds, making them more willing to allocate capital to bonds.
  • Confusion resulting from this misreporting of performance affects asset prices: Investors mistake price changes for returns, just like newspaper journalists. The authors empirically show that stocks move with price changes of indices and not, or at least to a lesser extent, with market-wide dividend yields. As a result, on days on which a dividend is paid out, information on prior returns is not fully incorporated into prices, meaning that there will be excess returns in the future. In line with this, the authors show that a high dividend yield on the market today predicts higher market returns the next trading day. In contrast, in Germany, where market performance is displayed as returns, the prior day’s dividend yield does not predict market performance. This is consistent with the notion that displays of performance drive asset prices.
  • Performance displays affect flows to mutual funds: Flows to mutual funds are related to relatively uninformative metrics of performance that are prominently displayed on financial websites. For example, Yahoo! Finance allows different options for benchmark comparisons of funds, namely the S&P500, the Dow Jones and the NASDAQ. All of them exclude dividends. As a metric of performance of the fund itself, Yahoo! Finance displays changes in Net Asset Value. This fails to correct for any distributions, including dividends, realized capital gains distributions, and returns of capital. Still, the authors empirically find that mutual fund flows are higher for funds that outperform the S&P 500, based on their percentage change in Net Asset Value, rather than the funds’ return.

Relevance for Practice

Displays of asset performance represent an understanding of returns from the early 20th century. The authors show that investors’ confusion about this misreporting of returns affects their perceptions of asset performance, market prices, flows to mutual funds, the tone of newspaper articles on market performance and the predictability of short-term market price movements. Therefore, it is important to educate investors about the meaning of this misreporting, while it is also a worthy policy goal to shift the default display of performance to return indices.