UC Berkeley

First chapter of this thesis finds a new consumption growth predictor linked to macroeconomic fundamentals: the technology gap, the dierence between potential and actual productivity of capital. I construct a representative firm business cycle model, in which the technology gap generates specic patterns of short- and long-run consumption growth, and consumption growth volatility. Intuitively, a high technology gap acts as an economic shock that increases consumption in the long term due to a higher future productivity

level. I use quality-adjusted price indices of durable investment goods to create a proxy for the technology gap. Consistent with the model, I find empirical evidence that a high technology gap predicts: (i) strong consumption growth at longer horizons, (ii) high consumption growth volatility, and (iii) high risk-free rate.

Second chapter demonstrates the relationship between research and development expenditure, and firm productivity. I construct a model which implies that firm-level R&D optimal policy should be dependent on ex-ante productivity. Firms ex-ante further from the frontier optimally invest more in R&D. Ex-post productivity depends on the amount of R&D investment and the match between new technology and existing production factors. Firms investing more in R&D are ex-post on average closer to the frontier, controlling

for theoretically motivated endogeneity. I present empirical evidence supporting the model. Using data envelopment, I construct a measure of firm-level distance from industry-wide productivity frontier. On average, a 1% larger distance from the frontier causes a 0.5% increase in R&D intensity next quarter. R&D activity in turn predicts high stock return volatility.

Third chapter tests the existing durable consumption-based asset pricing model of Yogo (2006). Consumption risk is measured by the covariance between asset returns and future durable consumption growth, rather than contemporaneous growth, as in the original model. I present empirical evidence that excess returns on Fama-French portfolios are correlated more with future than contemporaneous durable consumption growth. I transform the original Euler equations of the model to use information about

the future consumption growth. As its correlation with returns is higher, the estimate of risk aversion from the model decreases substantially compared with Yogo (2006). I also find that the altered consumption risk measure increases the explanatory power of the

model. I approximate the original model and show that it can be estimated in the simple OLS framework. Cross-sectional R square is highest when the consumption growth is sampled over six to eight quarters ahead. This result is robust to dierent sets of test assets.