Negative Midday Prices and a Flat Forward Hedge
A cold-storage operator consumes 120,000 MWh annually, with 55% of volume in summer afternoons. It currently buys 70% on a two-year fixed price at $58/MWh and 30% at day-ahead spot. The local grid has added 4 GW of solar in two years; midday prices now frequently settle between -$12 and $8/MWh, while evening prices average $74/MWh. The calendar-2026 forward is $52/MWh. Natural gas forwards are $3.10/MMBtu, and the on-peak heat rate implied by the forward market is 7.8 MMBtu/MWh. The CFO wants to lock the remaining spot exposure at the forward price. Operations expects to add 15,000 MWh of flexible load by 2027 and can shift 3 MW for four hours.
1. Using a variable O&M cost of $4/MWh, calculate the gas-fired marginal cost floor from the stated heat rate and gas forward. Then explain why a $52/MWh calendar forward can exceed that floor and why negative midday spot prices do not by themselves prove the fixed price is overpriced.
Read the worked answer
Fuel cost = 7.8 MMBtu/MWh × $3.10/MMBtu = $24.18/MWh. Add $4 variable O&M = $28.18/MWh marginal cost. The $52 forward leaves $23.82/MWh above that floor. That gap can reflect evening scarcity, capacity accreditation, carbon or environmental costs, shape and risk premium, and congestion. A flat forward is not a forecast of every hour; it is a tradable annual strip. The negative midday prices occur when solar output is high, but the operator's load is 55% summer afternoon and its spot exposure is not flat 24/7. If it locks flat baseload at $52, it may buy energy when its actual load is low and still buy evening energy at $74 if the fixed block does not match shape. The relevant comparison is the cost of the hedged shape, not the average spot price. Negative prices are evidence of shape value, not proof that a flat fixed price is too high.
2. Quantify the spot exposure and the value of shifting 3 MW for four hours on 90 days from $74 evening to $0 midday. Should the CFO lock 100% of the 36,000 MWh spot tranche at $52? Propose a hedge structure.
Read the worked answer
Spot exposure = 30% × 120,000 = 36,000 MWh. Locking all at $52 costs $1.872 million for that tranche. Load shifting: 3 MW × 4 h × 90 days = 1,080 MWh moved from a $74 evening price to a $0 midday price, saving about 1,080 × $74 = $79,920 before rebound or operational cost. Locking 100% of the spot tranche is too blunt: 55% of load is in summer afternoons, and a flat block may not follow the operator's shape. A better structure is a fixed block sized to base load, for example 20,000–24,000 MWh; a shaped or peak cap for summer afternoons; and an index tranche for flexible load. If the CFO's main concern is budget certainty, set a cap near $60/MWh rather than paying $52 for all hours. The 2027 addition of 15,000 MWh and 3 MW of shiftable load argues for keeping some volume index-based and testing a demand-response or battery case.
3. What non-price factors could reverse the recommendation?
Read the worked answer
Supplier credit, collateral, volume tolerance, basis or congestion, gas correlation, and operational ability to shift. A $52 fixed price from a BBB- supplier requiring a parent guarantee or mark-to-market collateral may be worse than a $54 price from an A-rated supplier. If the 3 MW shift cannot be dispatched on 90 days, savings fall. If summer afternoon congestion charges are high, the fixed nodal product may not match the load. If gas forwards are volatile, the implied heat rate can move. Also check whether the fixed price includes RECs or carbon and whether negative-price hours are curtailable. The decision should compare all-in hedged cost and liquidity risk, not just $/MWh.