Extension of loans is being considered as one of the most important and useful functions of the universal commercial banks in the economy. Russian banking has been gradually restoring after the financial crises, when the default on the government debt has showed that only those banks that are more concerned with the non-financial sector can survive external shocks. However, the fact that banks become more and more willing to extend loans to the corporate sector, due to the absence of other profitable financial instruments, exposes them more to risks of particular borrowers. Unpredictable macroeconomic conditions and inefficient contract enforcement mechanism are overlapped with risks of borrowers’ performance and value of their guaranties. To ensure themselves from risks on the micro level banks need to undertake costly and time-consuming process of assessment of their potential clients. In the situation when interacting parties possess different information or some data is not verifiable, credit relations demand not only a sound banking system and high management qualification, but also stability of financial records of the firms. Economic theory addresses different questions on this matter, and even we abstract from the commitment and moral hazard problem on the banks side, there exist a plenty of informational issues concerning borrowers. Even general financial situation is improving; small and medium size enterprises in Russia are constrained in resources. Most of them cannot undertake cumbersome and expensive procedure of issuing public debt. Newly emerged enterprises do not have credit history and performance records, and even those that have been working quite efficiently during a certain period cannot provide reliable financial documentation due to the presence of underground activities. High credit risks and costs due to the difficulties in assessing and choosing potential borrowers in the situation of informational imperfections are reflected in the interest rate on loans. As a result, enterprises with stable but not very high profitability, that is usual for non-financial sector, are constraint in their access to credit resources.While the situation in Russian banking is generally improving, risk of bad debts may affect the future of their financial stability. So both sides require recommendations in developing their strategies and possibility to rely on adapted theories of credit relations. In this respect, we illustrate bank-borrower relationships and bank decisions on the credit market in the context of the two game-theoretic models. We consider factors that influence the interest rate on loans and especially particular conditions of the Russian credit market i.e. costs and uncertainty related to the pledged assets, wide tax evasion and non-payments in the economy along with an absence of the standardise financial records of the borrowers. In the first model, banks, as Cournot oligopolists, extend loans to finance production of the firms on the security of collateral. Firms have two types of technology available, and a choice between them depends on the level of prudence in the society and cost of loans, through such specific factor as firm’s collateral valuation. Collateral requirements, set by the bank, are quite high and cover the principle with all interest and costs. But if the firms are able to provide falsified documentation and use illegally pledged assets, they will put a low value on their collateral in a loan agreement with a bank. Thus, particular conditions may lead to the choice of a risky investment, while improvement in the legal enforcement and relatively low interest rates will make a stable one more profitable.The second model is based on the theory of signalling and provides a more detailed analysis of the interest rate decisions of a representative bank that provides loans to the non-financial sector. In the present state of the Russian economy, enterprises are engaged in the tax evasion and maintain arrears no matter their actual profitability and investment possibilities. Due to financial uncertainty banks are not able to distinguish potentially creditworthy borrowers, and therefore insure themselves with high interest rates on loans. Enterprises can try to build more transparent relationships with banks and provide them with information indicating their performance and ability to repay loans.A bank’s decision at a fist stage of the game depends on whether it can distinguish between the two types of borrowers: with high and low investment risks. In order to obtain cheaper loans, the high quality (i.e. low risk) firms need to distinguish themselves from a homogeneous mass. We obtain a possibility for the latter firms to provide a credible signal to the bank through the improvement in their financial performance. The model’s results suggest that high quality firms can adjust their behaviour by refraining from chronicle non-payments in return for better credit conditions. This will serve as an indication of their efficient performance and ability to repay loans. The bank then conducts a particular credit policy separating high-risk borrowers from the low-risk ones and adjusts its interest rate on loans accordingly. L-type firms will continue to borrow at a high rate, as due to the limited liability and high upper values of their return distribution they benefit more from investing than from forgoing the project.