The proposed project studies the dynamics of extreme price movements, or price jumps, using high-frequency data of financial assets. In the first stage of the project, fellow address the issues of price jump indicators in the multivariate context as opposed the univariate price jump indicators dominating the literature. Then, he will compare his indicators with those in the literature using Monte Carlo analysis as well as real data. In the second stage, he focuses on the price jump dynamics methodology, which is rarely touched in the literature. First, he develops the two-stage method, where he first identifies price jumps using the previous mentioned indicators, and, in the second, step he estimate the market dynamics in different market regimes defined through the presence of jumps. He further extends the model and joins both stages into one. This new model based on the maximum likelihood will estimate price jumps and their dynamics in one step and thus provide more robust results. He further applies the methodology on the simulated data and the empirical high-frequency data for various stocks or stock market indices. Thus, he obtains the quantitative description of phenomena like information spread across financial markets—price jumps serves as a proxy for moment when information hits the market—or even market panic. The results can be directly applied to study financial integration of different markets, their stability in the financial crisis, contagion of market panic and to propose more efficient policies and construct more robust portfolios. Finally, fellow proposed a theoretical model based on the continuous-time DSGE methodology, which set the theoretical grounds for the observed phenomena. When fitted on the data, the model will significantly extend the current understanding of the financial markets and provides new theoretical methodology.