Quelques applications du filtre de Kalman en finance: estimation et prévision de la volatilité stochastique et du rapport cours-bénéfices
François-Éric Racicot () and
Raymond Théoret ()
Additional contact information Raymond Théoret: Département de stratégie des affaires, Université du Québec (Montréal)
Abstract:
The popularity of Kalman filter is increasing in financial studies, notably to estimate diffusion processes. In this article, we show how we can use it to forecast the volatility of returns and the price-earnings ratio of the S&P500. The Kalman filter is consequently very versatile when variables, as volatility or forecasted price-earnings ratio, are unobserved. But the forecaster must use his judgment when he uses the Kalman filter. An error of specification in the model may give way to very biased forecasts.