Is implied volatility more informative for forecasting realized volatility: An international perspective

Abstract

Inspired by the commonly held view that international stock market volatility is equivalent to cross-market information flow, we propose various ways of constructing two types of information flow, based on realized volatility (RV) and implied volatility (IV), in multiple international markets. We focus on the RVs derived from the intraday prices of eight international stock markets and use a heterogeneous autoregressive framework to forecast the future volatility of each market for 1 day to 22 days ahead. Our Diebold-Mariano tests provide strong evidence that information flow with IV enhances the accuracy of forecasting international RVs over all of the prediction horizons. The results of a model confidence set test show that a market's own IV and the first principal component of the international IVs exhibit the strongest predictive ability. In addition, the use of information flows with IV can further increase economic returns. Our results are supported by the findings of a wide range of robustness checks.

Forecasting bitcoin volatility: Evidence from the options market

“…The results suggest that the information content of is generally smaller than that contained in historical prices in predicting 1‐day ahead volatility. This finding is in line with Pong et al (2004) in fiat currencies but in contrast to Martens and Zein (2004) in the S&P 500 stock index, implying that BTC is more currency‐like from a volatility forecasting perspective.…”

“…We use the options trades to calculate IV and analyze the forecasting power of this volatility measure in comparison to GARCH, ARMA, and HAR‐type models. Although IV generally underperforms parametric time‐series models (especially ARMA‐type models) in forecasting 1‐day ahead volatility, it is superior in long‐term prediction consistent with the findings in Martens and Zein (2004) and Pong et al (2004). We explain this conditional forecasting accuracy with the fact that time‐series models such as ARMA and GARCH are designed to forecast at short‐term horizons whereas IV captures long‐term expectations and is thus better suited to forecast long‐term volatility.…”

“…According to Martens and Zein (2004) and Pong et al (2004), volatility forecasts based on historical realized volatility estimated using intraday returns can provide superior accuracy compared to GARCH variances and IV. Thus, we incorporate in our analysis a HAR model with quarticity and jumps (HARQ‐F‐J) developed by Corsi (2009) and Bollerslev, Patton, and Quaedvlieg (2016) which is documented in Shen et al (2019) to provide superior forecasting power among many HAR‐type models.…”

“…In addition to HARQ‐F‐J, we also include in our analysis ARMA(1,1) and ARFIMA(1,d,1) models whose significant forecasting power has been documented in a number of studies (e.g., Dunis, Kellard, & Snaith, 2013; Kambouroudis et al, 2016; Martens & Zein, 2004; Pong et al, 2004, among others). ARMA(1,1) and ARFIMA(1,d,1) are specified in Equations (7) and (8) respectively as follows: where is the degree of fractional integration, is the lag operator, and are the polynomials that defines autoregressive and moving average components respectively, and is the expected value of variable .…”

“…Earlier studies that compare the forecasting power of option‐IV and GARCH‐type models generally conclude that the former outperforms the latter (e.g., Day & Lewis, 1992 in stock markets and Jorion, 1995 in foreign exchange markets). Instead of daily data used in GARCH‐type models, recent studies increasingly use models based on intraday data and realized volatility (e.g., ARMA or ARFIMA) and find that such models generate comparable or even better forecasts than IV (e.g., Kambouroudis, McMillan, & Tsakou, 2016; Martens & Zein, 2004; Pong, Shackleton, Taylor, & Xu, 2004). However, IV still provides some incremental information and thus improves the forecasting performance when it is combined with model‐based forecasts (Busch, Christensen, & Nielsen, 2011; Haugom, Langeland, Molnár, & Westgaard, 2014).…”