News, Noise, and Tests of Present Value Models

I use a present value framework to explore the e�ects of news (or noisy information) onstock prices and drive theoretical restrictions that link price volatility to noise and information. In particular, I show that market e�ciency implies that noise cannot explain more than half of price uctuations. I propose a novel methodology to decompose stock prices into a value component, related to information about future economic fundamentals, and a noise component. The key observation is that noise by construction cannot change future economic fundamentals, but affects stock prices. The advantage of my approach is that it does not require any particular assumptions on unobserved discount rates and econometricians' information set. Consistent with the predictions of the model, my estimates show that in the prewar period noise explains up to 28% of the S&P 500 index, and 36% in the postwar period. Finally, I �nd that the U.S. stock market was undervalued during the 1970s and overvalued during the 1990s, but there is no evidence that the market was overvalued before the crash of 1929.