In the past we have discussed that some investors demand dividends. (Here is a nice post by Larry Swedroe on the topic and there are more holistic measures, such as shareholder yield, which are better predictors of future returns). A few posts we have on the topic highlight that CEOs cater to dividend demand,mutual funds “juice” the dividend yield, and examine the returns to dividend payers in payment months. These papers support the behavioral finance view of the financial marketplace.
Behavioral finance experts, Samuel M. Hartzmark and David H. Solomon, have a new working paper titled, “The Dividend Disconnect,” which examines investor behavior around dividend paying firms. The paper highlights the fact that humans use mental accounting and this affects decisions that investors make regarding positions in their portfolios.(1)
This paper has two primary predictions, which we will describe below.(2)
The Dividend Disconnect Summary
The first prediction in the paper is that, “Capital Gains and Dividends Viewed as Distinct Desirable Attributes.”
But what does that mean?
The authors highlight that when assessing stock positions, an investor has two options for how to assess the performance — (1) simple price appreciation/depreciation, or (2) total return. Note that price appreciation/depreciation is simply the price appreciation/depreciation on the position, while total return includes both the price appreciation/depreciation plus the dividend return.
Directly from the paper:
For many positions, either price changes or returns including dividends will yield the same category of gain or loss. However, some positions are at a gain when dividends are included, but at a loss without their inclusion. Do investors treat such positions as being at a gain or at a loss when evaluating whether to sell the position? This is equivalent to asking whether investors adjust for the mechanical decrease in shares price that results from dividend payments.
To test this, the authors regress a sell signal against two variables:
The results are found in Table 2. As shown in Columns 1 and 2, the authors find that individuals are around 7.5% more likely to sell a gain as opposed to a loss, also known as the disposition effect. Examining the “Gain only with Dividends” variable, the authors find that individuals sell winners based only on total return of the position (around 1% more likely to sell unambiguous gains compared to unambiguous losses in Column 2); however, this rate of selling winners on a total return basis (~1%) is much lower than the simple price appreciation assessment (~7.5%). The general results are similar for institutions and mutual funds who exhibit a positive disposition effect–stocks are more likely to be sold based on a simple price assessment compared to total return.(4)
