• maximum entropy method
• fire-sales
• financial markets
• complex systems
• community detection

The aim of this thesis is to provide a different approach for analyzing and better understanding the financial markets. At variance with the large part of previous approach, we focus directly on the trading activity of market participants, i.e. the purchase and sell of an arbitrary quantity of financial assets that are present in the considered market. In this approach we focus on some general properties of the system and from these we extract informations on the behaviour of the system's elements. Our attitude is different from the previous frameworks, where the study of single constituents had a central role. For this purpose we have used a special database, maintained by Euroclear Finland, that reports the activity, in terms of purchased and sold shares, of all the financial institutions present in the Finnish market. To describe dynamically this system and concentrate the large amount of information in an analytical tractable structure, we have used a class of mathematical objects: the complex networks.