Оценка вероятности отзыва лицензии банка с использованием методов регуляризации
In this study open data is used to estimate the probability of license withdrawal. The use of regularization allows taking into account the entire amount of data suggesting automatic selection of the optimal model. A number of predictive models are estimated. The methodology could be implemented by market actors in purpose to continuous monitoring of counterparties’financial stability.
This work contains the analysis of quotes dynamics for the biggest five Russian banks. The investigation is based on the trend of daily quotes. Pieces of news concerning these banks were researched in order to determine the reasons for the changes of trends. Main news groups were compared with the quote dynamics to investigate the power of their influence. It allowed to find fundamental factors influencing the quote dynamics changes and to estimate the power of the impact.