Volatility Modeling and Spillover: The Turkish and Russian Stock Markets

This study investigates the internal and external (spillover) characteristics of the volatility of the Turkish and Russian stock market indices. To this end, generalized autoregressive conditional heteroskedasticity models that are classified as short memory (GARCH, EGARCH, GJR-GARCH, APARCH) and long memory (FIGARCH, FIEGARCH, FIAPARCH, HYGARCH) considering adaptive structure (Fourier series), and the rolling Hong causality methods are used. The analysis spans the years 2003-2020, revealing that the asymmetric power autoregressive conditional heteroskedasticity model is the most appropriate method in terms of both stock indices and leverage and long memory effects are evident in the volatility series. Bidirectional volatility spillovers between Turkish and Russian stock market indices are also evident in all time horizons. Investors can use volatility results for stock valuation, risk management, portfolio diversification, and hedging, and policymakers can consider the volatility results to evaluate the fragility of financial markets.