What Is the Effect of Oil and Gas Markets (Spot/Futures) on Herding in BRICS? Recent Evidence (2007–2022)

Hang Zhang, Evangelos GiouvrisJournal of Risk and Financial Management 2023 · DOI

TL;DR

This study investigates how oil and gas markets influence investor herding behavior in BRICS stock markets from 2007-2022. It finds that energy markets have a minimal impact on herding, with investors only herding selectively during crises, and that low energy market volatility can trigger herding, while high speculative activity intensity drives it.

What Problem Does This Paper Address?

The paper addresses the gap in understanding how external market influences, specifically energy markets (oil and gas, both spot and futures), affect investor herding behavior in emerging economies. While herding—the tendency for investors to follow the crowd rather than making independent decisions—is a well-studied phenomenon, its drivers in the context of BRICS countries and its connection to global energy market dynamics, especially during periods of crisis, have not been thoroughly explored. The authors aim to provide a more nuanced view beyond the common belief that energy markets significantly impact herding, particularly during volatile times.

Methods

The study employs a quantitative approach, utilizing a modified version of the Cross-Sectional Absolute Deviation (CSAD) model to measure herding behavior. This model, originally proposed by Chang et al. (2000), is adapted to incorporate various factors:

  • Basic Herding Model: Regressing CSAD on the absolute market return and squared market return to detect herding.
  • Energy Market Influence: Adding returns from oil and natural gas (spot and futures) to the CSAD model.
  • Crisis Event Analysis: Modifying the model to analyze herding during specific crisis periods (Global Financial Crisis, European Debt Crisis, COVID-19, Russia-Ukraine War).
  • Granger Causality Tests: Used to determine the causal relationship between energy market performance and herding.
  • Market States Analysis: Examining herding during "upward" (positive returns) and "downward" (negative returns) periods in energy markets.
  • Volatility Analysis: Employing GARCH, EWMA, RS, and GK models to calculate volatility and then using a quartile-based approach to assess the impact of different volatility levels on herding.
  • Speculative Activity Analysis: Introducing a speculation ratio (trading volume to open interest) to analyze its effect on herding, including its intensity and volatility.

Datasets

The study uses daily data for the period from May 17, 2007, to May 13, 2022.

  • Stock Market Indices:
    • Brazilian IBOVESPA
    • Russian MOEX Index
    • Chinese CSI 300
    • Indian S&P BSE SENSEX
    • South African FTSE/JSE Africa All Share Index
  • Energy Market Data:
    • Henry Hub Natural Gas Index (spot)
    • Brent Oil Index (spot)
    • Generic 1st "CO" Futures (oil futures)
    • Generic 1st "NG" Futures (natural gas futures)

Data for stock indices were collected from Bloomberg, and energy market data were sourced from the U.S. Energy Information Administration and Bloomberg.

Key Insights and Ideas

  • Energy Markets' Limited Direct Impact: Contrary to popular belief, oil and gas markets have a surprisingly minimal direct impact on herding behavior in BRICS stock markets.
  • Selective Herding During Crises: Investors don't herd universally during crises; they tend to herd selectively at specific points in time or in response to particular energy market movements.
  • Low Volatility as a Herding Trigger: Low volatility in energy markets (both spot and futures) is a significant driver of herding across all BRICS countries, a finding consistent with previous research in other markets.
  • Intensity, Not Just Speculation: It's not the mere presence of speculative activity but its intensity or volatility that drives herding. As speculative activity becomes more intense, herding increases.
  • Brazil's Unique Position: Brazil stands out as an outlier, showing immunity to speculative activities in energy markets, even when intensity increases.
  • Oil's Greater Influence than Gas: While both are considered, oil markets generally show a stronger connection to herding than gas markets.

Key Results

  • Overall Herding Effect: Energy markets (oil and gas, spot/futures) have a minimal impact on herding in BRICS stock markets. Causality tests show oil futures Granger-causing CSAD only during the COVID-19 crisis for some countries, while gas has no effect.
  • Market States: Bullish or bearish energy market states do not significantly affect herding.
  • Volatility Impact: Low energy market volatility (first and second quartiles) consistently triggers herding across all BRICS countries. Higher volatility levels do not lead to herding. South Africa shows the strongest herding response to low volatility.
  • Speculative Activity Impact: Speculative activities during non-crisis periods have minimal impact. However, increased intensity/volatility in speculative activities leads to herding in most BRICS countries (except Brazil), with India being most affected.
  • Crisis Period Analysis: Herding during crises is limited and country/period-specific. Russia shows no herding across any crisis period. Brazil and Russia exhibit anti-herding during some crisis periods, suggesting their role as potential "counterbalancing forces" in portfolios.

Limitations and Future Work

  • Data Availability: The study was limited by data availability, preventing the inclusion of data prior to 2007 and the use of a wider range of oil and gas indices, especially for futures.
  • Geographic Scope: The study focused on BRICS. Future research could extend to frontier countries with varying dependencies on oil/gas to test the universality of the findings.
  • Model Extensions: The authors suggest extending the sample period and including more indices. Further research could explore other energy commodities and investigate the impact of geopolitical events more deeply.
  • Brazil's Anomaly: The unique behavior of Brazil, being immune to speculative activities, warrants further investigation.
  • Universality of Findings: The study's findings on herding and volatility could be tested across a broader range of global markets to assess their generalizability.

Summary updated

Abstract

We investigate the effect of gas/oil markets (spot/futures) on herding in stock markets in BRICS over 15 years (2007–2022). We consider the effect(s) of crises (Global financial, European debt, COVID-19, and Russia–Ukraine war), bull/bearish energy markets, volatility, and speculation. The effect of gas and oil markets on herding in stock markets is minimal, and investors herd selectively during crises. Even during the ongoing Russia–Ukraine war, the effect of energy markets on herding in BRICS is minimal. Causality tests show that oil (spot/futures) Granger causes CSAD during COVID-19 only. Gas (spot/futures) has no effect. We also find that energy (spot/futures) market states (bearish/bullish) have no effect on herding in stock markets. Low volatility in energy markets can trigger herding (consistent with previous research in US, China) in all BRICS. Speculative activities during (non)crises appear to have minimal impact on herding. However, as the degree of intensity (volatility) in speculative activities increases in oil/gas, it causes herding in all countries (India is affected mostly), except Brazil. It is not the speculation activity per se in (non)crises that causes herding, but the intensity/volatility in speculation activity. Overall, oil/gas markets (especially gas markets) appear to have a smaller impact on herding than expected, contrary to public belief; however, as the intensity/volatility in speculative activities increases, then herding also increases, which is expected given the uncertainty that speculation causes.
BibTeX
@article{zhang2023what,
  author = {Hang Zhang and Evangelos Giouvris},
  title = {{What Is the Effect of Oil and Gas Markets (Spot/Futures) on Herding in BRICS? Recent Evidence (2007–2022)}},
  journal = {Journal of Risk and Financial Management},
  year = {2023},
  doi = {10.3390/jrfm16110466},
}
Figure