Commodity and counterparty risk

# Commodity and counterparty risk

In 2015, Delta Air Lines held a portfolio of jet-fuel hedges and refinery interests that, on paper, was supposed to protect the airline from spiking energy costs. Then crude collapsed from $100 to under $40 a barrel. Delta's hedges went the wrong way, forcing more than $2 billion in cumulative losses and margin calls over two years. Southwest, long the industry's hedging poster child, had already begun unwinding aggressive positions. The lesson wasn't that hedging is dangerous, it's that the line between hedging and speculation is thinner than most boards understand, and it is the CFO's job to draw it, police it, and defend it.

This lesson is about the two exposures that most often turn a treasury function into an unintentional trading desk: the price of what you buy, and the solvency of who sits on the other side of your contract. Both are manageable. Both become career-ending when governance lags the position.

Framing commodity exposure: from physical reality to hedgeable risk

Before you hedge anything, you have to know what you are actually exposed to, and most companies overstate their true exposure because they confuse *gross* input spend with *net economic* exposure.

Start with the distinction between transactional, translational, and structural commodity risk. Transactional is the near-term purchase you've committed to. Structural is the multi-year exposure baked into your business model, an airline is structurally long fuel demand; a food processor is structurally short its key crop. The critical CFO judgment is: which portion of exposure passes through to customers, and over what lag?

Here is the analytical move that separates sophisticated shops from naive ones. Your hedgeable exposure is not your commodity spend. It is:

> Net exposure = Physical volume × (price sensitivity of costs − price pass-through to revenue) × unhedged duration

A quick-service restaurant chain that buys beef but reprices its menu every quarter has *far* less economic exposure than its beef invoice suggests, because pricing power absorbs much of the shock. Hedging the full notional would over-insure and create a new speculative position on the wrong side. Conversely, a fixed-price government contractor with a locked-in bid and a volatile steel input has near-total exposure and a short pass-through window, it must hedge or it must bid differently.

Choosing the instrument to the exposure

Once you've sized net exposure, match the instrument to the *shape* of the risk, not to what your bank is selling:

• Forwards and futures for linear, high-conviction exposure where you're willing to give up upside for certainty. Cheap, liquid, but they create margin-call cash drag, the very thing that nearly broke Delta's treasury.
• Options (caps, floors, collars) when you want asymmetric protection, insurance against a bad outcome while retaining favorable moves. You pay premium, but you avoid the mark-to-market cash bleed on the downside.
• Swaps for converting floating exposure to fixed over a defined tenor, common in energy and metals procurement.
• Basis and proxy hedges when a perfect instrument doesn't exist. A jet-fuel buyer often hedges with heating oil or crude because there's no deep jet-fuel derivatives market. This introduces basis risk, the hedge and the exposure move imperfectly together, and it is the single most underestimated risk in commodity programs.

The practical rule: the closer the hedge tracks the physical exposure, the less you can be surprised. Every basis mismatch, every tenor gap, every proxy correlation that "usually holds" is a place where your P&L will diverge from your intention at the worst possible moment.

Building the hedging mandate: the governance that prevents speculation

Here is the uncomfortable truth: the difference between a hedge and a speculative bet is *not* the instrument. A futures contract can be either. The difference is whether an offsetting physical exposure exists and whether the position sits inside an approved mandate. Governance, not instrument choice, is what keeps you honest.

The CFO's core deliverable is a hedging policy that a board can understand and an auditor can test. It has five components:

1. Objective statement. Define *why* you hedge in one sentence the board endorses: "To reduce the volatility of gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → from input-cost movements within a defined budget-protection band." This single line does enormous work, it explicitly rules out "to profit from commodity views." When a trader later argues for a directional position, the policy answers before the meeting starts.

2. Coverage ratios and layering. Specify the *percentage* of forecast exposure that may be hedged, tiered by time horizon, for example, 70-90% of the next 6 months, 40-60% of months 7-18, and 0-30% beyond. This layered, ratcheting approach avoids the trap of hedging 100% at a single price (which is itself a bet) and forces you to average into positions. It also caps how far out you can go, containing the tenor risk that makes long-dated positions so dangerous.

3. Approved instruments and prohibited structures. List what treasury may use and explicitly ban exotic structures, leveraged swaps, knock-in barriers, and anything with embedded optionality that increases loss beyond the notional. Many corporate blowups (recall the 2020 losses at a major Chinese state oil trader, or numerous mid-cap manufacturers) came from structures that looked like cheap hedges but embedded a sold option that magnified losses.

4. Limits. Three limits matter: notional limits (how much exposure can be hedged), stop-loss / value-at-risk limits (how much mark-to-market loss triggers escalation), and counterparty limits (covered below). Each limit needs a named owner and an escalation path.

5. Segregation of duties. The person who executes trades cannot be the person who confirms, values, and reports them. This is the control that catches rogue behavior, the Barings and Société Générale failures were failures of segregation, not of markets.

The risk committee in practice

A policy is inert without a body that meets, reviews, and can say no. The risk committee, typically the CFO, treasurer, head of procurement, a business-unit leader, and a risk officer, should convene at least monthly and review:

• Current coverage ratios versus mandate
• Mark-to-market and cash-flow-at-risk on open positions
• Basis performance (are proxy hedges still tracking?)
• Counterparty exposure and collateral movements
• Any limit breaches and remediation

The committee's real function is cultural. When procurement wants to "lean into" a falling market or a business head wants to lift hedges because "prices are obviously going higher," the committee is where an *ex ante* discipline confronts an *ex post* temptation. The CFO's job is to make sure the mandate wins that argument every time, because the moment you allow one discretionary view, you have re-founded a trading desk under your P&L.

Hedge accounting reinforces this discipline. To qualify for hedge accounting treatment (which lets you defer gains/losses rather than run them through P&L each period), you must document the hedge relationship, the risk being hedged, and prove effectiveness. The *discipline of qualifying* is itself a governance filter: a position you can't document as hedging a real exposure probably isn't one.

Counterparty risk: who fails when you need the hedge most

A hedge is only worth the solvency of the party on the other side. This is the exposure that vanishes from the model right up until it detonates, as it did in 2008, when firms holding perfectly reasonable hedges discovered their counterparties (or their counterparties' counterparties) couldn't perform. The bitter irony: counterparty risk peaks in exactly the stressed markets where your hedge is most in-the-money and most valuable.

The CFO manages counterparty exposure along four levers:

1. Measure the right exposure. Your counterparty exposure is not the notional, it's the replacement cost if they default: the current mark-to-market value in your favor, plus a potential future exposure add-on for how much that value could grow before you could unwind. A swap that's currently at zero can still carry meaningful future exposure. Track both current and potential exposure per counterparty.

2. Diversify and set counterparty limits. No single bank or dealer should hold more than a defined share of your hedge book. This feels obvious until a treasurer defaults to the one relationship bank offering the tightest spread. Cheaper pricing from a concentrated counterparty is a discount you pay for with tail risk.

3. Use collateral and netting agreements. The ISDA Master Agreement with a Credit Support Annex (CSA) is the standard architecture. Netting collapses your many positions with one counterparty into a single net exposure. The CSA requires the out-of-the-money party to post collateral as marks move, dramatically reducing replacement risk. The trade-off, and it is a real one for the CFO, is that collateral posting introduces liquidity risk: when your own positions move against you, you must fund margin calls in cash, fast. This is precisely the mechanism that strained Delta and that turned Germany's Uniper and other utilities into liquidity crises during the 2022 energy spike, when soaring prices forced billions in margin postings even on economically sound hedges.

4. Watch the funding-liquidity linkage. The modern insight is that market risk, counterparty risk, and liquidity risk are not separate, they are the same risk viewed at different moments. A large in-the-money hedge is a counterparty exposure to *them* and a collateral-funding exposure to *you* if the market reverses. The CFO must stress-test the hedge book for the cash-flow consequences of large adverse moves: How much liquidity must I have on standby to keep good hedges alive through a price swing? Answering that question in advance, via committed credit lines sized to a stress scenario, is what separates firms that survive volatility from those that are forced to liquidate hedges at the worst moment.

Putting it together on Monday morning

The integrated workflow looks like this: (1) size *net* economic exposure, not gross spend; (2) decide coverage ratios by horizon under the board-approved mandate; (3) select instruments that minimize basis and unwanted optionality; (4) execute across diversified, ISDA/CSA-documented counterparties within limits; (5) monitor mark-to-market, basis performance, counterparty exposure, and collateral liquidity in a monthly risk committee; and (6) pre-fund a liquidity buffer sized to a stress scenario so a winning hedge never becomes a losing cash crisis. Each step is a control point where governance either holds or fails.

Key Takeaways

• Hedge net economic exposure, not the invoice. Subtract price pass-through and account for pricing lags before sizing any position; over-hedging is itself a speculative bet on the other side.
• Governance, not instruments, defines the hedge/speculation boundary. A board-endorsed objective statement, layered coverage ratios, banned exotic structures, hard limits, and segregation of duties are what keep treasury from becoming a trading desk.
• Make the risk committee win every discretionary argument. The moment you allow one "view" to override the mandate, you have re-founded a speculation operation, so build the committee to enforce *ex ante* discipline against *ex post* temptation.
• Model counterparty exposure as replacement cost plus future exposure, and diversify with ISDA/CSA netting and collateral. Never let spread savings concentrate your tail risk in one dealer.
• Pre-fund the liquidity that keeps good hedges alive. Stress-test collateral calls for adverse price moves and size committed credit lines accordingly, the failure mode isn't a bad hedge, it's being forced to liquidate a good one to meet a margin call.