Tesi di LAUREA SPECIALISTICA Titolo A Concrete Experience in Operative Risk Management and Capital Allocation Data 2011-12-20 Autore/i Lo Re, Valentina Relatore Barucci, E. Full text non disponibile Abstract During 2011, with a greater emphasis in the last few months, Italian banks often made it to the front pages and headlines of the media, as they faced important losses or reduced stability. Banks are in fact subject to nancial and economic pressure coming from dierent sides, they are affected by a multitude of internal and external factors, which are often entangled and diffcult to cope with. All these factors can be condensed in a key word, that emerges in the combative world of nance: risk . It is the amount of risk and the ability to handle it which at the end of the day determines the performance of a nancial institution, down to its stability and ultimately its survival. Consequently, being able to handle risks is the cornerstone on which banks fund their strategy and their power, and the weapons that institutions need to put in place are the analysis, modelling and prediction of risks. The paramount importance of the analysis of risk for the stability of our financial system constitutes the strong motivation behind this work, which is the result of a collaboration between the author and one of the major Italian financial groups. The main outcome of the internship and of the research which was carried out has been an extension in the modelling of the risk,which has been integrated into the existing risk model of the nancial group.The thesis is structured in two main parts: the rst one, drawn from the literature of the sector, and the second one, more innovative and experimental. In Chapter one, we start by providing a denition and a taxonomy of the various types of risk, followed by a broad overview which outlines why Operational Risk is of capital importance. We discuss the state of the art, showing how Operational Risk has been conceived and formalized in the last two decades and we point out the challenges and limitations in the way it is analyzed and modelled until now. We also outline the entities controlling the Operational Risk. In the last section of this chapter , we illustrate in details the core of our method, paying particular attention to the technique chosen by the Group for the study of Operational Risk. In Chapter two, we focus our attention on the description of the model. A rigorous mathematical approach for each part of calculation is presented, starting from the estimation of single losses until the Capital at Risk value. At rst, a clear distinction between two main aspects of our problem is explained. Drawing our issue, two are the questions which occured to us: how many yearly losses will occur and which will be the amount of each single losses. The answer to these two question are the prerequisite for the definition of the loss frequency distribution and of the loss severity distribution. At the beginning, all the analisys will concern each risk class; only in the last Section we treat the problem of the aggregation and we reach to the real Capital at Risk calculation. With the Chapter three the main subject of this thesis is presented: the Allocation of the capital to the subsidiaries of the Group. The fruit of our labours are mainly here and it concerns how to model loss data for the calculation of the stand-alone capital. In fact, we ve inserted in our model distributions that were not considered before, we ve tested them and, finally, we ve proposed a possible future method, taking into account all our analysis. Those were made thanks to the support of S-plus, a data analysis software system. In the last Chapter we report the main outputs of the calculation engine. They lead us to believe that the introduction of further loss distributions should be one of the steps to develop and to improve the allocation mechanism.