Социальные медиа и макропруденциальная политика: влияние новостного сентимента на потребительское кредитование в России

Журнал Новой экономической ассоциации. 2027. № 1.

Lapteva E., Pashkov E., Кузьминых Е. А.

This paper explores how the information environment influences consumer credit dynamics within the framework of macroprudential policy. The analysis is based on a dataset of messages from Russian economic Telegram channels. News sentiment and topical relevance are assessed using a multimodel approach that combines several large language models, thereby mitigating sensitivity to any single model’s architecture. The resulting aggregate measures are used to construct a financial stability sentiment index, which is then incorporated into an econometric analysis of consumer lending growth using bank-level panel data and macroeconomic indicators over the period 2017-2025. The results indicate a negative association between sentiment and consumer credit growth, as well as a structural shift in this relationship after 2022, when more favorable assessments of financial stability begin to move in tandem with credit expansion. The strength of these effects varies across banks depending on their characteristics and becomes more pronounced under tighter monetary conditions, confirming the importance of the information channel in the transmission of macroprudential measures. A series of robustness checks consistently supports the validity and stability of the main findings.

Keywords: потребительское кредитование, financial stability, финансовая стабильность, макропруденциальная политика, consumer lending, news sentiment, Macroprudential policy and regulation, медиа сентимент, large language models (LLMs), большие языковые модели (LLM)