Over the last few years counterparty risk management has become one of the most important factors influencing financial markets. The main modern models of partner banks assessment are based on rating. But ratings have essential faults. A new approach to counterparty risk assessment is construction by a bank of forecast spreads of credit default swaps (CDS) for its counterparties.In this article, two models are constructed: the model of assessment of CDS theoretical values for banks of the BRICS group, CDSs on the debt of which are not traded in the market, and the model for assessment of default probability of Russian banks based on the results of the first model. The results of the study contribute to improvement of the existing methods for counterparty risk assessment and can be used for further estimation of default probability of counterparties of Russian financial institutions and companies.