Research seminar by Yarema Okhrin (Universität Augsburg): «Cased-based reasoning for volatility modelling»

On Thursday, March 15 at 4.40 pm International College of Economics and Finance held Research seminar.
Speaker: Yarema Okhrin (Universität Augsburg)
Theme: «Cased-based reasoning for volatility modelling»
Venue: Pokrovski Bulvar, 11, Room Zh-822

Abstact: Explaining and forecasting volatility is a key issue in modern risk management. Whereas simple autoregressive models are generally suitable for modeling, the core question is to provide some economic reasoning for the chosen time series model. This paper proposes an economic motivation, which is closely related to a decision making process. We argue that dynamics of daily volatility is a phenomenon, which can be accessed via not rule based but case based reasoning using the arguments of the case based decision theory. The case based theory presumes analogous thinking of human beings in cases where current situations are evaluated by considering their similarity to previously experienced situations. Furthermore, volatility is a measure of uncertainty which reflects disagreement of market participants concerning future cash flows. Since an individual assessment of uncertainty is rather case based, the similarity based reasoning provides an intuitive explanation for the volatility dynamics. The econometric implementation of the case based reasoning is conducted via the empirical similarity concept, which is related to the nonparametric kernel estimation. The developed theory is applied to modeling and forecasting volatility of major financial indices.