The first chapter of this dissertation highlights the role of technology in asset pricing by demonstrating market return predictability based on aggregate technology shocks from both theoretical and empirical perspectives. I solve simple general equilibrium models, in which technology shocks drive conditional mean and volatility of future economic growth. The expected market returns and premiums therefore vary across time. This implication is strongly supported by my empirical study, in which I find that the technology shocks, i.e. unexpected growth of patents and R&D expenditures, have strong and distinctive explanatory power for market returns and premiums in both short- and long-term predictive regressions.
The second paper, joint with Yu-Chin Hsu, proposes a stepwise superior predictive ability (SSPA) test that is more powerful in identifying predictive models or profitable investment targets in a large-scale multiple testing framework. We provide formal proofs and employ simulations to confirm that the proposed SSPA test is more powerful than White's reality check (2000), Hansen's superior predictive ability test (2005), and Romano and Wolf's stepwise multiple testing procedure (2005). We apply the SSPA test to examine the performance of mutual funds and hedge funds in a multiple testing framework.
In the third chapter, I consider the role of technology in the cross-section of stock returns. In an intertemporal asset pricing model, technology shock changes future investment opportunity for investors, which causes technology risk in stock returns. This study empirically tests this intuition using patent and R&D data. We construct a tracking portfolio to measure the change in expectation of future technologies and develop a technology factor. We test asset pricing models including this technology factor and find that its price is of statistical and economic significance.