Following the requirements of the Basel Committee on Banking Supervision, banks need to comply with a capital ratio according to risks. Optimization techniques refine the risk model.
For many years now, multinational banks have had to comply with particular rules, one of which is to keep a ratio higher than 8% between capital and current risks. It means that banks always have to have money to support potential financial needs in case of an incident, crisis … This percentage includes operational risks (system failure, fraud…) and leads banks to develop a tool to estimate risks (loans, market…). To comply with Basel II regulations, they designed models to estimate risks with more accuracy.
As a part of recasting its computation engine for risk assessment, a major French bank asked us to develop an optimization algorithm for a risk model with Matlab environment (Matlab, Statistics Toolbox, Optimization Toolbox, Matlab Compiler, Matlab builder JA).. Once this had been done, the code developed was compiled into Java so as to be run by operators from a computation server.
More information : www.banque-credit.org/pages/comite-de-bale.html