Genetic algorithms in forecasting commercial banks deposit

Deposits constitute the most important source of funds for commercial banks and are part of the total capital funds required for economic development of a country. The paper presents the application of genetic algorithms (GAs) in forecasting commercial bank deposits. The proposed forecasting algorithm captures patterns relation between the inputs: gross domestic product, money supply (held by the public), interest rate, number of branches of commercial banks, and loans, and the output: bank deposits, while keeping the absolute average forecasting error as minimal as possible. The results of computer simulations show the average error for a ten year forecast at around ±5%, which is within a satisfactory range. The results demonstrate that GAs can accurately forecast in the model of econometrics and finance, and that they do not restrict either the form or the regularity of the objective function. With GA forecasting, every component is taken into account to ensure the optimum forecasting value