Development of adaptive regulatory thresholds for volatility in the Russian stock market

Abstract

The current mechanism for limiting volatility on the Russian stock market — a discrete auction triggered when the MOEX index falls by more than 15% within ten minutes under Bank of Russia Regulation No. 437-P — was calibrated to prevent catastrophic single-day crashes resembling the 1987 Black Monday and does not account for the crisis scenarios typical of the domestic market. The research problem lies in the absence of an empirically grounded methodology for setting regulatory thresholds differentiated by the intensity of market stress. The study aims to develop and verify a multi-level system of regulatory volatility thresholds for the MOEX index, calibrated on intraday drawdown from the opening price. The sample covers 2006–2025 (5,004 trading days). Methods include conditional volatility modelling via EGARCH(1,1) with a Student’s t-distribution, market regime identification through a Markov regime-switching model, threshold optimisation based on a loss function with operational constraints on the annual number of triggers, and retrospective verification across four crisis episodes. The current threshold is found to detect no more than 2.3% of statistically extreme trading days, whereas the second level of the proposed three-tier system detects 18.3% of such days at a comparable trigger frequency and covers all four crisis episodes examined, including early identification of the February 2022 military shock. A regime-dependent adaptation mechanism is developed that automatically tightens thresholds once the market enters a stress regime. The findings are applicable to improving the regulatory practice of the Bank of Russia and the Moscow Exchange.