A model for stock market returns: non-Gaussian fluctuations and financial factors

摘要

While there are various theories to account for the large variations in stock prices, some observed statistical aspects require further analysis. A model is proposed for aggregate stock prices, based on observed data, rather than any efficient market hypothesis, and considering jumps in statistical parameters between phases of generally increasing, or generally decreasing, aggregate stock prices. The model relates a critical parameter for short-term behaviour directly to financial factors, especially interest rates, to explain large short-term variations which follow a non-Gaussian distribution. Economic fundamentals may affect changes over longer periods.