The quantile dependence of commodity futures markets on news sentiment

Abstract

Focusing on energy commodities, industrial metals, and gold, this paper examines the degree to which commodity futures returns depend on news sentiment under various market conditions, and the structure of that dependence. We observe an asymmetric market reaction to positive and negative news sentiment, which changes in periods of financial turmoil. The quantile regression analysis shows that news sentiment's influence on the futures returns follows an upward trend at higher percentiles. This structure flattens for positive news during the global financial crisis, while the slope for the negative component steepens in backwardation periods.

Intermediary capital risk and commodity futures volatility

“…In line with Badshah (2013) and Koenker and Hallock (2001), who argue that the averaging effects of the OLS regression could lead to inaccurate estimation of the nonlinear relationships, we use a quantile regression method to explore the heterogeneous response of commodity futures volatilities to similar intermediary shocks across a volatility distribution. Following Baur (2013) and Omura and Todorova (2019), the evidence shows an increasing trend of the dependence between commodity futures volatilities and positive (negative) intermediary capital shocks toward upper volatility percentiles, and this structure is more significant for the negative intermediary capital shocks. However, empirical results also demonstrate that financialization in commodity futures markets flattens the dynamic trace of the quantile regression coefficients over the post‐2008 period.…”

“…In addition, according to He and Krishnamurthy (2012) and He et al (2017), the capital constraints of intermediaries are more likely to impact asset prices under unfavorable economic conditions since intermediaries must bear disproportionately large risk, and as a result, asset prices must adjust to make the greater risk exposure optimal. According to several empirical studies, the return distribution of commodity futures markets is affected by the heterogeneity of investors' responses (Omura & Todorova, 2019; Zhu et al, 2016). Considering the heterogeneous effects of intermediaries and the nonlinear relationships in commodity futures markets, a heterogeneity‐consistent regression method should be used to explore the various responses of commodity futures volatilities to similar intermediary shocks across a distribution as an extension of the OLS regression model.…”

“…Following Baur (2013) and Omura and Todorova (2019), we use the quantile regression method (a heterogeneity‐consistent method) that Koenker and Bassett (1978) introduced to examine the degree and structure of the dependence of commodity futures volatilities on intermediary capital risk factors throughout the conditional distribution of market volatility. This method is more efficient than OLS regression since focusing only on average effects could lead to inaccurate estimation of relevant coefficients or omission of important relationships (Badshah, 2013; Koenker & Hallock, 2001).…”

“…Based on Baur (2013), we transform the basic quantile model into the following two specifications: and Whether there are changes in the degree to which commodity future volatilities are associated with intermediary capital risk factors across the conditional distribution and the structure of that dependence can be determined by analyzing the coefficients of ICR in Equation (18). We follow Baur (2013) and Omura and Todorova (2019) in setting the average degree of the dependence to remain unchanged if the estimates of the coefficients across quantiles are zero. In Equation (19), the degree and structure of the asymmetric effect of positive and negative intermediary capital risk factors are captured by the coefficients of and , respectively.…”

The dynamics of cross‐boundary fire—Financial contagion between the oil and stock markets

“…The most popular way to identify the existence of financial contagion is to provide evidence of a significant increase in cross‐market linkages after a shock to one or a group of markets (Forbes & Rigobon, 2002). Several methods have been proposed to measure the cross‐market linkages such as the conditional correlations (Forbes & Rigobon, 2002), vector autoregression approach (Chichernea et al, 2019), multivariate generalized autoregressive conditional heteroscedasticity (GARCH) model (Isleimeyyeh, 2020), and quantile regression approach (Čech & Baruník, 2019; Omura & Todorova, 2019). Unfortunately, most of these methods are based on linear assumptions and ignore the nonlinear dependence between financial markets (Bampinas & Panagiotidis, 2017).…”