Аннотация:Emerging markets often go through periods of financial turbulence and the estimation of market risk measures may be problematic. Online search queries and implied volatility may (or may not) improve the model estimates. In these situations a step-by-step analysis with R and Russian market data is provided. Four classes of models are considered (GARCH, HAR, ARFIMA, and realized-GARCH), and a detailed forecasting and backtesting investigation is performed.