Self-fulfilling prophecies and stock price volatility.

Summary This thesis focuses on the measurement and determinants of stock price volatility. The first part begins by assessing, using time series econometric analysis, the magnitude and persistence of cyclical stock price fluctuations. We then present the main financial anomalies (notably excessive price volatility and predictability of returns) in a unified and synthetic framework. The analysis leads us to relativize the scope of the informational efficiency hypothesis of the markets, and to emphasize the need to explain the large variations in the discount rate that cause the price movements. The second part of the thesis proposes a theoretical model, based on the notions of equilibrium indeterminacy and self-fulfilling prophecy, capable of accounting for the stylized facts thus identified. Using two simple asset pricing models, we show that a large number of financial anomalies, far from demonstrating the irrationality of markets, can be explained by the multiplicity of equilibria that can appear in rational expectations models.