Together Cash Flows and Consumption Account for Differences in Asset Returns
The level of dividend exposure to consumption influences the amount of asset risk compensation.
ANN ARBOR, Mich. — Asset pricing hinges, in part, on the idea that differences in exposure to systematic risk should justify differences in risk compensation. Economic theory suggests the source of this systematic risk should be variation in aggregate consumption. However, researchers have not been able to uncover strong empirical links between aggregate consumption and differences in risk compensation across assets.
In a new study, Robert Dittmar of the Stephen M. Ross School of Business at the University of Michigan and colleagues explore how the relationship between dividends paid on a portfolio and aggregate consumer spending helps to explain the differences in expected returns on assets.
Asset returns can be broken into components related to cash-flow growth and to valuation ratios. The authors isolate the cash-flow growth portion of the return and argue that assets with greater cash-flow exposure to consumption risks (i.e., larger cash-flow beta) should offer higher risk compensation. Conversely, lower cash-flow risk exposure should result in lower compensation.
"Our overall message is clear," said Dittmar, assistant professor of finance. "Estimates-of-risk measures based solely upon the relation between cash flows and consumption explain a considerable amount of the cross-sectional variation in measured risk premia."
In contrast, risk measures based on the relation of total return and consumption explain little of the difference in average returns across assets.
In their study, Dittmar and colleagues Ravi Bansal of Duke University and Christian Lundblad of Indiana University examine 30 portfolios sorted into sets of 10 on the basis of three characteristics: market capitalization, book-to-market ratios and past returns (momentum). The dividends paid on these portfolios are the measure of cash flow. A portfolio's cash-flow risk is measured by the degree to which its dividends move in tandem with aggregate consumption.
The researchers say their cash-flow risk measures display striking patterns within the portfolio-characteristic groupings. In particular, the large-firm, low-book-to-market and worst-performing ("loser") portfolios display risk measures that are lower than those of the small-firm, high-book-to-market and best-performing ("winner") portfolios, respectively. In contrast, most total-return risk measures exhibit the reverse relation, or no relation at all to these characteristics.
"This illustrates an important point," Dittmar said. "Portfolios with high cash-flow risk measures are portfolios with high average returns, while portfolios with low risk measures are portfolios with low average returns. In other words, portfolios with higher cash-flow covariance with consumption have larger risk premia."
Further confirming their findings, the winner and high-book-to-market portfolio dividend growth rates demonstrate close pro-cyclical movements with the smoothed consumption growth rate, whereas the loser portfolio dividend growth rate demonstrates strong counter-cyclical movements.
"A striking feature of our results is that the constant exposure of the cash flow paid on momentum portfolios to aggregate consumption appears to be closely connected to the average returns earned on these portfolios," Dittmar said. "In the past, momentum has proven a particularly challenging dimension for asset-pricing models to explain."
Overall, the researchers say, differences in cash-flow betas account for more than 60 percent of the cross-sectional variation in risk premia for the 30 portfolios, compared to 6.5 percent for the Capital Asset Pricing Model, the main risk-based model used in practice.
"The market price for risk in cash flows is highly significant," Dittmar said. "Therefore, cash-flow risk is important for interpreting differences in risk compensation across assets."
For more information, contact:
Bernie DeGroat, (734) 936-1015 or 647-1847, firstname.lastname@example.org