Determinants of financial stability in Kazakhstan's banking sector: a hierarchical regression approach

Abstract

Financial stability in the banking sector is a critical prerequisite for macroeconomic stability, particularly in the context of rising credit risks. The study aims to quantitatively assess the impact of liquid assets (FSLT), interest income structure (FSEIM), and return on assets (FSERA) on the level of non-performing loans (FSANL). The scientific novelty of the research lies in the application of hierarchical regression analysis to evaluate the stability of Kazakhstan’s banks, which revealed the hidden significance of the income structure that was not evident through simple correlation analysis.

The research methodology is based on 65 observations. Normality testing using the Shapiro–Wilk test confirmed a statistically significant deviation of all variables from the normal distribution (p < 0.05), justifying the use of hierarchical modeling (models M₀–M₃).

Key results demonstrated that the final model M₃ explains 62.7% of the variation in the FSANL indicator. Contrary to the results of pairwise correlations, where the strongest relationship was observed for liquidity (r = -0.468), the dominant predictor of stability in the complex multifactorial model was interest income structure (β = -0.781), followed by liquidity (β = -0.761) and profitability (β = -0.560). The negative sign of the coefficients confirms that strengthening the internal financial base of banks leads to a reduction in credit risks. Foreign exchange position (FSSNO) and extended liquidity (FSLS) indicators were found to be statistically redundant for this model.

The practical significance of the work lies in the possibility for regulators to use the M₃ model to monitor systemic risks and improve supervisory practices. The value of the study consists in the empirical evidence that for the banking system of Kazakhstan, the stability of interest income flow is a more critical factor for resilience than the volume of liquid assets in the short term.

Keywords: second-tier banks, financial stability, regression modeling, banking sector of Kazakhstan, banking supervision, liquidity, income structure, non-performing loans (NPL).