Noncausal and noninvertible models in financial econometrics: theory and applications

Financial markets often behave in ways that standard models find challenging to explain. Asset prices can fluctuate dramatically during crises, exhibit episodes of extreme volatility, and sometimes even form "bubbles" before collapsing. My research investigates how incorporating investors’ expectations about the future can enhance our understanding and prediction of these abrupt market movements. By extending traditional forecasting methods—which typically examine only past data—to also include forthcoming information, I present new models that reflect the intricate, real-world patterns observed in financial returns. These models more effectively manage explosive price growth and sharp volatility spikes, which frequently occur during events like the 2008 financial crisis or more recent disruptions. Through both theoretical approaches and empirical testing on stock and commodity markets, I demonstrate that employing forward-looking tools can significantly enhance our capacity to assess risk, anticipate market fluctuations, and develop more resilient financial strategies.