Introduction. Currently, the study of the role of financial intermediation as an important auxiliary mechanism of economic growth has received considerable attention in the theoretical and empirical literature. The problems of economic and mathematical modeling of causal relationships between the rates of economic growth and the dynamics of the financial system development attract the attention of a large number of both foreign and Russian specialists. Most authors believe that not only depth growth, but also a change in the financial sector structure (the ratios between its various segments) can have an impact on economic growth. A quantitative assessment of the impact of the financial market’s type of structure (bank-oriented or based on the securities market) on economic growth is of practical interest. The aim of the work is an econometric study of the influence of the ratio of bank credit volume and the issuance of securities on the rate of economic growth in Russia. The observation period is from Q1 2003 to Q4 2017. Theoretical analysis. In Russia, a bank loan has penetrated into the economy much deeper than the securities market, which lags far behind in terms of depth and efficiency from the world average. The indicators characterizing the structure of the financial market are the volume of bank lending (with the allocation of loans to individuals and organizations) and the total market capitalization of shares traded on the Moscow Stock Exchange. The paper uses an econometric methodology for studying the statistical relationship between non-stationary time series, including tests for the Ingle-Granger cointegration, the study of causality and the response to shocks based on the vector-based error correction model (VECM). Empirical analysis. A comparison of time series of quarterly values of financial structure indicators, as well as Russia’s GDP for 2003–2017 is carried out. The article presents a statistical comparison of the level of development of the Russian stock market relative to the markets of developed and developing countries. The modern econometric Gretl package was used for calculations and modeling. Results. The cointegration of non-stationary time series has been established: gross domestic product, total capitalization of the Moscow Exchange, and bank lending to individuals and legal entities. The Ingle-Granger test found a cointegrating relation that confirms the long-term equilibrium relationship between variables and the authenticity of their correlation. It is shown that economic growth largely depends on the development of a bank credit and to a lesser extent on the growth of the market capitalization of shares. It is shown that in terms of contribution to economic growth, loans to organizations are more than twice as large as loans to individuals. Decomposition of the variance of forecast errors in the medium term revealed the influence of loans to organizations on the dispersion of economic growth and bank credit.