Finance & Accounting Jan 1, 2011
Dividends Driving Volatility
Erratic stock prices due to concerns over short-term dividend payments
Thirty years ago, Yale economist Robert Shiller presented a puzzle in the form of a line chart. He drew one line representing the actual price of stocks over decades versus the prices one would expect based on dividend payments that corporations deliver to their shareholders. The line representing actual prices is a spiky one with erratic peaks and valleys, while the line of expected prices gently rolls. Since the smooth undulations of expected dividends are thought, in part, to drive stock prices, then stock prices should be smooth. Thus, the puzzle is, why is the stock market so volatile?
Stock market experts have attempted to unravel this so-called excess volatility puzzle for decades. Now, Jules van Binsbergen, an assistant professor of finance at the Kellogg School of Management, Michael Brandt, a professor of finance at Duke University, and Ralph Koijen an assistant professor of finance at the University of Chicago, have discovered that concerns about dividends in the next few years can account for most of the excessive volatility in stock markets, in addition to concerns about the dividend payments over longer terms, which has been the focus of previous studies.
“This is the first time we can measure people’s concerns regarding the first few years of dividends directly,” Koijen says.
To measure investors’ short-term concerns, van Binsbergen and his colleagues studied two alternatives for purchasing stocks: Investors can purchase stocks and hold them, pocketing any dividend payments along the way, or they may choose to buy a future or option contract on that same asset. In the latter case, the purchase price is based on the expected price of the stock or index at some time in the future, say one year. At the end of the year, the investor receives that asset, no matter its price. If the price at the end of the year is higher than the price paid, the investor pockets the money. But while stocks purchased normally pay dividends over that year, futures investors do not receive back payments for the dividends. Thus, the difference between the price of the futures contract and the current index price reflects the value investors place on the dividends for the period of the futures contract. Using actual data from the S&P 500 Index, which catalogs the prices of 500 large stocks, the team saw what investors are willing to pay today to receive two years of dividends.
Investor purchases indicated an underlying concern about short-term dividend payments. Dividends are cut in the worst of times—a recession or a depression—which also happens to be when investors need their cash the most, says van Binsbergen. So although the risk of a dividend cut may be low, the fact that it coincides with an economic meltdown leads investors to demand higher returns, he says.
Another puzzle in the stock market—the equity premium puzzle—is that investors request an excessively high compensation to hold stocks, which are considered a risky form of investment. Equity premiums, also called risk premiums, operate in the following way: risk averse investors will refuse to buy a stock for $40 today, even when it is expected to yield $40 within the next year, but they may purchase it for $35. For this $5 gain, the risk or equity premium, they will risk investing in the stock market. More specifically, the risk or equity premium is the returns on stock minus the risk-free rate of keeping money in the bank.
According to standard economic models, people are implausibly risk averse, requiring very high risk premiums. Here is a real world example: If you invested in the stock market fifty years ago, your money would have grown by 10 or 11 percent whereas if your money was in the bank, it would have just grown by 5 percent. Even this incredibly high risk premium of 5 percent fails to convince many people to purchase stock.
“Most of the finance literature suggests that investors require these high premiums because they’re afraid of what might happen to dividends in five to thirty years,” van Binsbergen says. “It says investors don’t care about dividend payments over just the next three years—but we find that they do.”
“By parsing out what matters to investors in the short- and long-term, we can learn how certain events impact market expectations of future dividend growth for the years to come.” — Koijen
When van Binsbergen and his colleagues dissected futures data in the S&P 500, they were able to discern what risks people would take within the first three years of purchasing stock. It turns out they require a large incentive for taking on even a year of risk. “Investors worry a lot about those first few dividends,” van Binsbergen notes, “Why they are so risk averse is a different question, we just know they seem worried about the short-term, more than we thought they were.”
As a result of the high-risk premiums investors desire, those who buy and sell stocks price them far lower than what expected corporate dividends would suggest.
“Some days you may feel fearful,” van Binsbergen says, “but maybe tomorrow you’ll take a gamble.” And here is where Shiller’s volatility puzzle with the concordant lines comes into play: changing attitudes towards short-term risk and long-term risk make stock prices far more volatile than expected corporate earnings and dividends would suggest. In part, it is why “dividend payments are smooth, but stock prices fluctuate a lot over time,” van Binsbergen explains.
“By parsing out what matters to investors in the short- and long-term, we can learn how certain events impact market expectations of future dividend growth for the years to come,” Koijen remarks.
“Now more than ever, in the context of the current financial crisis,” van Binsbergen says, “we need to understand the role of short-term and long-term risk for financial markets.”
Related reading on Kellogg Insight
van Binsbergen, Jules H., Michael W. Brandt, and Ralph S.J. Koijen. 2012. On the Timing and Pricing of Dividends. American Economic Review, June, 102(4): 1596-1618.