This paper presents an application of neural network and simulation modeling to analyze and predict the performance of 883 Russian Banks over the period 2000-2010. Correlation analysis was performed to obtain key financial indicators which reflect the leverage, liquidity, profitability and size of Banks. Neural network was trained over the entire dataset, and then simulation modeling was performed generating values which are distributed with Largest Extreme Value and Loglogistic distributions with estimated parameters providing robust results. Next, a combination of neural network and simulation modeling techniques was validated with the help of back-testing. Finally, we received nine bank clusters that describe the structural performance within the Russian Banking sector.