Can Investment Shocks Explain the Cross-section of Stock Returns?∗

In this paper we assess the role of capital-embodied technology shocks in explaining properties of the cross section of stock returns. Existing theories seem to disagree on the sign and magnitude of the price the market demands for bearing the risk of such shocks: while a negative price is necessary to justify the value premium or the cross-sectional predictability of returns by firm characteristics, a positive price is required to justify the existence of momentum profits. Our direct empirical analysis, conducted over the 1930–2010 period, finds weak support for the existence of a significant price of risk for investment-specific shocks and indicates, instead, that inferences based on commonly used proxies of investment-specific shocks are sensitive to the time period considered (pre vs. post 1963). Taking the magnitude of the value premium or the momentum profits as given we further infer that the empirically observed magnitude of these shocks appears to be too small to explain such phenomena. Finally, exploiting the fact that investment shocks manifest themselves through the creation of new capital stock, we propose novel cross-sectional tests that rely on firms’ capital intensities to indirectly assess the impact of investment shocks on stock returns without measuring these shocks directly. The empirical evidence we gather lends mild support to the hypothesis that investment-specific shocks can explain a significant part of the cross sectional variations in stock returns. JEL Classification Codes : E22; G12; O30