A component Markov regime‐switching autoregressive conditional range model

In this article, we develop one- and two-component Markov regime-switching conditional volatility models based on the intraday range evaluate their performance in forecasting daily of S&P 500 Index. We compare with that several well-established return- range-based models, namely EWMA, GARCH, FIGARCH GARCH model, hybrid EWMA CARR model. in-sample goodness fit out-of-sample forecast using a comprehensive set statistical economic loss functions. To assess use mean error metrics, directional predictive ability tests, evaluation regressions, pairwise joint tests; to appraise value at risk coverage tests management show proposed switching produce more accurate forecasts, contain information about true volatility, exhibit similar or better when used for estimation risk. Our results are robust choice proxy, sample size, period, alternative distributions.