Can financial markets be decoded using hidden signals? This paper presents a self-calibrating model for financial signal processing, offering a new approach to understanding market dynamics and pricing securities. By considering the short rate process as a function of an unobserved Markov chain, the model enables explicit expressions for the prices of zero-coupon bonds and other securities. Discretizing the model allows the application of signal processing techniques from Hidden Markov Models. This innovative approach enables the estimation of not only the unobserved Markov chain but also the parameters of the model itself, making it self-calibrating. The model's architecture simplifies the complex calculations often involved in financial forecasting. The estimation procedure is rigorously tested on a selection of U.S. Treasury bills and bonds, demonstrating the model's practical applicability. The findings suggest that incorporating signal processing techniques can improve the accuracy of financial models. Future research could explore the model's performance with other asset classes and in different market conditions.
Published in the International Journal of Theoretical and Applied Finance, this paper aligns well with the journal's scope by presenting a novel model for financial analysis. By applying signal processing techniques to understand the short rate process, the research contributes to the theoretical understanding of financial markets. The self-calibrating feature of the model enhances its practical value, fitting the journal's applied focus.