The role of news-based uncertainty indices in predicting oil markets: a hybrid nonparametric quantile causality method

Abstract

A recent literature emphasizes the role of news-based economic policy uncertainty (EPU) and equity market uncertainty (EMU) as drivers of oil-price movements. Against this backdrop, this paper uses a k-th order nonparametric quantile causality test, to analyze whether EPU and EMU predicts stock returns and volatility. Based on daily data covering the period of 2 nd January, 1986 to 8 th December, 2014, we find that, for oil returns, EPU and EMU have strong predictive power over the entire distribution barring regions around the median, but for volatility, the predictability virtually covers the entire distribution, with some exceptions in the tails. In other words, predictability based on measures of uncertainty is asymmetric over the distribution of oil returns and its volatility.JEL Codes: C22; C32; C53; Q41 Keywords: Uncertainty; Oil markets; Volatility; Quantile causality  We would like to thank two anonymous referees for many helpful comments. However, any remaining errors are solely ours.

On the time‐varying links between oil and gold: New insights from the rolling and recursive rolling approaches

“…As we discussed in the next section, all of these asymptotic testing approaches might have series size and power distortions in finite samples and not robust to integration–cointegration properties of the data (Balcilar et al, ; Balcilar & Ozdemir, ; Dolado & Lütkepohl, ; Park & Phillips, ; Sims, Stock, & Watson, ; Toda & Phillips, ; Toda & Phillips, ; Toda & Yamamoto, ; and Yamada & Toda, ). The causality‐in‐quantiles tests are robust to misspecification and structural breaks (Balcilar, Bekiros, & Gupta, ; Balcilar, Gupta, & Pierdzioch, ). Although, both the nonparametric causality‐in‐quantiles tests and rolling and recursive rolling tests used in this study are robust against structural breaks, each one has certain advantages.…”

The role of time‐varying rare disaster risks in predicting bond returns and volatility

“…In this section, we present the methodology for the detection of nonlinear causality via a hybrid approach as developed by Balcilar et al (2017), which in turn is based on the frameworks of Nishiyama, Hitomi, Kawasaki, and Jeong (2011) and Jeong, Härdle, and Song (2012).…”

“…To achieve our objective, we conduct the predictability analysis based on the k ‐th order nonparametric causality‐in‐quantiles test recently developed by Balcilar, Bekiros, and Gupta (). As indicated by Balcilar et al.…”

“…As indicated by Balcilar et al. (), the causality‐in‐quantile approach has the following novelties. First, it is robust to mis‐specification errors as it detects the underlying dependence structure between the examined time series.…”

“…To achieve our objective, we conduct the predictability analysis based on the k-th order nonparametric causality-in-quantiles test recently developed by Balcilar, Bekiros, and Gupta (2017). As indicated by Balcilar et al (2017), the causality-in-quantile approach has the following novelties. First, it is robust to mis-specification errors as it detects the underlying dependence structure between the examined time series.…”