소비관련 거시변수를 통한 자산수익률의 예측

The basic pricing equation for consumption-based asset pricing models is derived from consumers’ first-order conditions on consumption and investment decisions, motivated by a desire to smooth consumption over both time and across states. That the price of an asset should equal the expected value of the asset’s future payoff, discounted by consumers’ marginal utility is widely regarded as the most fundamental principle of asset pricing. However, the performance of consumption-based asset pricing models have produced disappointing empirical results. For example, consumption- based asset pricing models have difficulty explaining several asset pricing phenomena, such as the high equity risk premium, the low real interest rate, the high equity volatility, and the cross-sectional variation in stock returns. In recent decades, a number of consumption-based measures have been proposed for forecasting asset returns, such as the consumption- aggregate wealth ratio (Lettau and Ludvgison, 2001; 2005), the surplus consumption ratio (Campbell and Cochrane, 1999) and the labor income to consumption ratio (Santos and Veronesi, 2006). Lettau and Ludvigson (2001) show that a variable that measures deviations of consumption from its stable relation with wealth has predictive power both in time series and cross- section of stock returns. Campbell and Cochrane (1999) use the surplus consumption ratio as a proxy for time-varying relative risk aversion and show that their model successfully matches the historical equity premium. Wachter (2006) uses a proxy of surplus consumption ratio to extend Campbell and Cochrane’s (1999) model to the bond market. The model of Santos and Veronesi (2003) implies that the ratio of labor income to consumption is inversely related to the conditional covariance between the asset return and the consumption growth. The ratio helps to explain time variation in expected returns and cross-section of stock returns. In this paper, we construct the recently proposed consumption-based measures using the Korean data and examine whether these measures predict future asset returns. Due to limited availability of macroeconomic data, the sample period for this study is much shorter than the sample periods used for the US. In constructing the consumption-based measures, we pay careful attention to the differences between the Korean and US data on household wealth and income in order to capture the informational contents of the proposed measuresas well as possible. We find that the consumption-based measures have predictive ability for the equity, bond, and housing risk premia in Korea. A regression of stock, bond, and housing price index returns on lagged consumption-aggregate wealth ratio (Lettau and Ludvgison, 2001; 2005) produces statistically significant coefficients. The surplus consumption ratio (Campbell and Cochrane, 1999; Wachter, 2006) predicts future stock and housing price index. The labor income to consumption ratio (Santos and Veronesi, 2006) predicts future stock, bond, and housing price index. Taken together, our findings suggest that these consumption-based measures capture the information relevant for time-varying risk premium in Korea.