When Oil Volatility Infects Everything Else
Commodities: Markets, Performance, and Strategies
Editors: H. Kent Baker, Gregory Filbeck, Jeffrey H. Harris
ISBN: 9780190656010
After 2008, everyone wanted to know if a shock in oil would spread to corn, copper, or gold. Frankie Chau and Rataporn Deesomsak build a tool to measure exactly that: the Commodity Volatility Spillover Index (CVSI).
Why this matters now
Commodity futures went from niche hedging to mainstream portfolio holdings. Investment fund activity hit $330 billion by 2012, up ninefold from the early 2000s. That raised a fair question: are commodity markets more connected, and more fragile, than before?
Some blame financialization for distorted prices. Others, like Stoll and Whaley, say the evidence is thin. Either way, regulators and portfolio managers need to know how volatility jumps from one market to another.
Nine markets, three sectors
The study covers weekly data from April 1990 to December 2016 across nine futures:
- Agriculture: corn, soybeans, coffee
- Energy: Brent crude, WTI, natural gas
- Metals: copper, gold, silver
Energy is the most volatile sector. Gold and silver were stable before 2004, then surged as investors treated them as crisis hedges. Correlations are positive across the board, with the tightest links between corn-soybeans (0.61), gold-silver (0.74), and Brent-WTI (0.89).
How the spillover index works
Chau and Deesomsak use the Diebold-Yilmaz (2012) framework. Think of it as a weather map for volatility shocks. A VAR model with GARCH volatility estimates decomposes how much of each market’s variance comes from its own shocks versus spillovers from others.
You get three useful numbers:
- Total spillover - how connected the system is overall
- Directional TO - how much volatility a market sends out
- Directional FROM - how much it receives
What the full sample shows
Over 1990-2016, total spillover averaged 24.8%. That means about three-quarters of each market’s volatility is idiosyncratic. The rest bleeds across borders.
Biggest transmitters: Brent crude (45.7% TO), WTI (43.3%), silver (35.8%), gold (33.4%)
Biggest receivers: Same energy and metal markets also absorb the most spillovers.
Most isolated: Coffee barely sends or receives volatility. Natural gas is 97.8% driven by its own shocks.
Net transmitters: Silver (+4.5%) and gold (+3.4%) push volatility out. Agricultural markets are net receivers.
Crises change everything
The static average hides the real story. A rolling 200-week window shows spillovers jumping above 50% during the global financial crisis and Eurozone debt crisis. Volatility interconnectedness spikes in recessions.
Directional plots reveal crude oil as the dominant transmitter after 2000. Gold was a major spillover source around the dot-com bust, then flipped to receiver mode by 2006. Agricultural markets mostly absorb stress rather than create it.
Do not blame just one market
The authors warn against scapegoating gold or oil for regulation. Focusing on one contract misses the systemic nature of the commodity network. Higher margins on gold futures (CME raised them 22% in August 2011) might distract from broader interconnectedness problems.
Robustness checks pass
Results hold when you change the VAR order, forecast horizon, or rolling window length. The pattern is stable.
Practical takeaway
If you hold a diversified commodity portfolio, energy and precious metals volatility will hit your agriculture positions during crises, even if corn fundamentals look fine. The CVSI is essentially an early warning gauge for systemic commodity stress.
For regulators, the message is: watch the network, not just individual contracts.
Navigation
← Previous: Benchmarking Commodity Performance (Chapter 17)
Next: Commodities in Emerging Markets (Chapter 19) →