Commodity Options and Convenience Yield

Commodity options are options on a physical market’s futures or forwards, where storage costs, inventory scarcity, seasonality, and delivery rules shape the underlying curve. Convenience yield is the non-cash benefit of holding inventory; it helps explain backwardation and means an equity-style spot carry model is often the wrong starting point.

Cost of carry and convenience yield

For a storable commodity with spot S, risk-free rate r, proportional storage cost u, and convenience yield y,

F(0,T) = S0 * exp((r + u - y) * T)

High inventory scarcity raises y, lowering futures relative to spot and potentially producing backwardation. The formula is an equilibrium shorthand, not a mechanical arbitrage identity: physical storage capacity, financing constraints, location grades, and delivery optionality can prevent a trader from extracting the apparent spread.

Curve state	Typical carry relation	Economic interpretation
contango	F > S	financing and storage exceed convenience yield
backwardation	F < S	inventory convenience is valuable
seasonal curve	varies by delivery month	predictable demand/supply pattern
stressed nearby	sharp front backwardation	scarcity, logistics, or squeeze risk

The futures curve itself is the pricing input for options on futures. Infer an implied convenience yield only after specifying spot, storage, and financing conventions, and do not use it as a stable forecasting factor without inventory evidence. Futures curve roll yield explains the distinction between curve shape and realized roll return.

Black-76 pricing

Most exchange-traded commodity options are priced from the futures price using Black-76:

V_call = D(0,T) * [F * N(d1) - K * N(d2)]
d1 = [ln(F/K) + 0.5*sigma^2*T] / (sigma*sqrt(T))
d2 = d1 - sigma*sqrt(T)

The discount factor is separated because the futures price is a martingale under the appropriate futures measure. The option expiry can differ from the underlying futures’ last-trade or delivery date; use the precise contract calendar. American-style energy options, average-price options, and options with exercise into physical positions need different engines.

import numpy as np
from scipy.stats import norm

def black76_call(future, strike, vol, expiry, discount):
    st = vol * np.sqrt(expiry)
    d1 = (np.log(future / strike) + .5 * vol**2 * expiry) / st
    d2 = d1 - st
    return discount * (future * norm.cdf(d1) - strike * norm.cdf(d2))

def implied_convenience(spot, future, r, storage, T):
    return r + storage - np.log(future / spot) / T

Surface construction and spread risk

Commodity volatility is not one surface. A crude-oil option references a particular delivery month; its volatility depends on option expiry, futures delivery, strike, and often season. A calendar-spread option depends on two futures and their correlation. Interpolating only by expiry while silently rolling the underlying contract can create artificial volatility jumps.

Smile skew often reflects crash risk, producer hedging, and supply shocks. In natural gas and power, mean reversion and seasonal spikes can make lognormal Black extrapolation especially unreliable. Quote cleanup should enforce calendar and butterfly arbitrage checks in price space, retain bid-ask weights, and flag illiquid wings instead of forcing a smooth but non-tradable fit.

Hedging an option book

Black delta hedges the specific futures contract, not spot. As expiry approaches, liquidity may migrate to the next contract, creating roll basis. A long-dated option may require a strip of futures to hedge its curve exposure, particularly under a stochastic-convenience-yield model where curve factors move independently.

dV approximately delta_F*dF + vega*dvol
              + cross_gamma*dF_near*dF_far + theta*dt

For crack, crush, and calendar structures, hedge ratios should come from a joint model or historical beta but then be stress-tested against dislocations. Commodity spread trading covers the underlying economics; an option overlay adds nonlinear exposure to those spreads.

Delivery and operational risk

A futures option can exercise into a futures position with delivery obligations. Contract specifications may include delivery location, quality grade, exchange-for-physical rules, position limits, and final settlement methodology. Negative WTI prices in 2020 demonstrated that assumptions such as “futures are always positive” are not operationally safe. Black lognormal formulas and collateral systems must have defined fallbacks for extreme price regimes.

Realized volatility may jump with weather, outages, geopolitical events, or inventory reports. Stress both a parallel futures move and a curve reshaping, with volatility and correlation changes. Historical backtests that use continuous futures prices can hide the actual option’s delivery-month basis and should not be used as a direct pricing validation.

Key takeaways

• Convenience yield represents inventory value and helps explain futures-curve shape.
• Price listed options from their specified futures contract and settlement calendar.
• Commodity surfaces are delivery-month and season dependent, not just expiry dependent.
• Futures delta hedging leaves roll, curve, smile, and delivery risks.
• Extreme physical-market regimes require explicit model and operational fallbacks.