Nonlinear intermediary pricing in the oil futures market

We study the state-dependent trading behavior of financial intermediaries in the oil futures market, using structural vector autoregressions with Markov switching in heteroskedasticity. We decompose changes in futures price volatility into changes in the slopes of traders' demand curves and in the variability of their demand shocks. We find that the downward-sloping demand curve of intermediaries steepens significantly during turbulent times. Moreover, the variance of intermediaries' own demand shocks doubles during these episodes. These findings suggest that the futures pricing of intermediaries is nonlinear and increases the hedging costs of producers and processors of oil when volatility is high.