By Grant Mercer
There is a substantial risk of loss in trading commodity futures and options products. Losses in excess of your initial investment may occur. Past performance is not necessarily indicative of future results.
By late 1993, MG Refining & Marketing (MGRM) — the American petroleum subsidiary of Germany's Metallgesellschaft AG — had built one of the largest oil derivatives positions in the world: a long futures stack, equivalent to roughly 160 million barrels, hedging long-term fixed-price supply contracts sold to its customers. The strategy was economically coherent. The textbooks said it should work.
Then the margin calls arrived, and the textbooks went silent. In December 1993, the parent company's supervisory board liquidated the entire futures position, crystallizing reported losses of roughly $1.3 billion, and the 112-year-old conglomerate survived only through a bank rescue reportedly approaching $2 billion. This is the story of how a funding crisis — not a price forecast — destroyed the hedge, and why the episode remains the most debated disaster in derivatives history.
What happened: the squeeze
Through 1993, two market developments converged on MGRM's stack-and-roll design. The price of oil fell from the high teens toward the low teens per barrel. And the forward curve flipped from backwardation to contango. Both changes struck directly at the strategy's load-bearing assumptions.
The damage came through two channels, and understanding their interaction is the entire lesson of the episode.
The first channel was margin. MGRM's long futures stack was marked to market daily, and as oil prices fell, the position accumulated large paper losses — reportedly generating margin calls on the order of a billion dollars or more. These were not theoretical losses. They were cash demands, payable immediately, to the clearinghouse. Every dollar the oil price fell was a dollar per barrel that had to be wired, that month, regardless of the decade-long horizon of the customer contracts.
The second channel was the roll. With the curve now in contango, each monthly roll of the stack — selling the expiring contract low, buying the next month high — crystallized a loss. The positive roll yield the strategy had counted on became a negative roll yield: a steady bleed of cash, month after month, in addition to the margin calls.
Here was the cruel asymmetry at the center of the affair. The customer contracts — the long-term fixed-price delivery obligations the futures were hedging — were simultaneously becoming more valuable as oil prices fell. MGRM was locked into selling oil at above-market prices for years to come; in economic terms, the company held a large unrealized gain. But that gain existed only on paper. The customer contracts paid off over a decade; the margin calls were due this week. The hedge was, arguably, working: the combined position was economically coherent. But coherence is not cash, and the cash was flowing in only one direction.
This is the distinction the episode seared into the industry's consciousness: the difference between a P&L hedge and a cash-flow hedge. On a mark-to-market basis, MGRM's books may have been roughly balanced. On a cash basis, the company was hemorrhaging. And companies do not survive on mark-to-market valuations. They survive on cash.
The mechanism: the boardroom decision
In December 1993, Metallgesellschaft's supervisory board — alarmed by the mounting margin calls and reportedly unconvinced by the subsidiary's explanations — intervened decisively. It dismissed the executives responsible, liquidated the entire futures stack, and began renegotiating or unwinding the long-term customer contracts. The liquidation crystallized losses commonly reported at roughly $1.3 billion. The parent company, its creditworthiness shattered, required a rescue package led by Deutsche Bank reportedly approaching $2 billion. Metallgesellschaft never recovered its former shape; it was eventually dismantled, its remnants becoming an engineering company under a different name.
The liquidation is where the academic battle lines were drawn. In a famous 1995 analysis, economists Merton Miller and Christopher Culp argued that MGRM's stack-and-roll strategy was fundamentally sound: the hedge properly matched the economics of the long-term delivery obligations, and the margin calls were a foreseeable feature of the design, not a flaw. On this view, the disaster was not the hedge but the boardroom — a supervisory board that did not understand the strategy liquidated it at the worst possible moment, transforming temporary cash-flow pressure into permanent realized losses. Had the board held its nerve and funded the margin calls, the argument runs, the customer contracts' embedded gains would have vindicated the program.
The counterargument, advanced most prominently by Antonio Mello and John Parsons, held that the strategy was reckless regardless of its theoretical elegance. A hedge that requires its owner to post a billion dollars in margin during adverse moves — funded from where, exactly? — is not a hedge but a speculation on the owner's own liquidity. The strategy assumed access to funding that Metallgesellschaft did not, in the event, possess. On this view, the board's liquidation, however panicked, recognized a genuine insolvency risk: a company that cannot meet its margin calls does not get to wait for its decade-long thesis to play out.
The debate has never been fully resolved, which is precisely why the episode remains so instructive. Both sides agree on the central fact: the strategy's viability depended on funding, not just on price views. Where they differ is on whether that dependence was a manageable feature or a fatal flaw. For the student of risk, the takeaway does not require settling the debate — it requires understanding that the debate exists, because it reveals the dimension most hedgers neglect.
The warning signs the liquidation exposed
The funding question was never answered. The entire strategy rested on an implicit assumption: that when margin calls arrived in size, the parent company would fund them. That assumption was never tested until it failed. A hedge that requires its owner to post a billion dollars in margin during adverse moves is a hedge whose true counterparty is the owner's own treasury — and the treasury's capacity was never part of the design.
The board didn't understand the book. The supervisory board that liquidated the position reportedly did not fully grasp the strategy it was unwinding. An organization running one of the world's largest derivatives positions, overseen by directors who could not distinguish a funding squeeze from a failed strategy, was an accident waiting for a trigger. Whether the liquidation was wise or panicked, it was made under conditions of incomprehension — and governance that cannot evaluate the book under stress becomes the mechanism of the firm's destruction.
Liquidation converts pressure into permanence. Margin calls are, in principle, temporary: they are collateral against a position that may recover. Liquidation makes the loss final. The board's decision transformed what Culp and Miller would later call a funding problem into a realized $1.3 billion loss — and then the customer contracts, the supposed economic anchor of the whole program, were renegotiated at fire-sale terms, destroying the embedded gains that might have justified holding on. The sequence matters: the decision to fold is what made the losses real.
Creditworthiness is a finite resource. Once the margin calls became known within the banking system, Metallgesellschaft's credit standing collapsed, and the rescue package reportedly approaching $2 billion was triage, not a vote of confidence. A funding squeeze that damages creditworthiness feeds on itself: every dollar of perceived distress raises the cost of the funding needed to survive it.
The aftermath
The rescue of Metallgesellschaft was one of the largest corporate bailouts in German history to that point, and it permanently altered the landscape. German regulators and auditors took a new interest in derivatives exposures at industrial companies; the affair became a staple of every subsequent debate about corporate use of derivatives, cited by skeptics as proof that manufacturers and merchants should never run large trading books.
For the derivatives industry, the episode produced a lasting conceptual advance: the formal distinction between market risk and funding liquidity risk, and the recognition that hedging transforms the former into the latter rather than eliminating it. The phrase "stack and roll" entered the industry's vocabulary as a cautionary term — not because rolling is wrong, but because every roll is a financing decision disguised as a trading decision.
And the academic debate — Mello and Parsons versus Culp and Miller — became required reading in every serious derivatives curriculum, precisely because it has no clean resolution. It forces every student to confront the question the Metallgesellschaft board faced in December 1993: when the margin calls arrive and the thesis is long-dated, do you fund or do you fold? The honest answer, the episode suggests, must be determined before the calls arrive — because by the time they do, the decision is already made by the funding you have or don't.
What the liquidation teaches
A hedge can be economically sound and still bankrupt you. The Metallgesellschaft affair is the definitive case study of this truth. Markets settle in cash, and cash has a schedule. A position that is right about prices and wrong about cash flow is a wrong position.
Funding is part of the position. The largest cash demand a margined strategy can generate is not a footnote to the trade; it is the trade's binding constraint. Every leveraged or margined strategy should be stress-tested against its own funding before it is opened: what is the maximum margin call, and where does that cash come from? If the answer is "we'll figure it out," the strategy is incomplete.
Decide the unwind before the crisis. The fund-or-fold question cannot be answered well under the pressure of arriving margin calls. It must be settled in advance — in the risk committee's minutes, in the treasury's committed facilities, in the board's documented understanding of the strategy. By December 1993, the decision was already made by the funding Metallgesellschaft had or didn't.
Governance is a risk input. The board that liquidated MGRM's hedge may have been right or wrong — the academics still argue — but it was certainly deciding under conditions of incomprehension. An institution running large derivatives positions needs overseers who understand them well enough to tell a funding squeeze from a failed strategy. Otherwise, the governance meant to protect the firm becomes the mechanism of the firm's destruction.
There is a substantial risk of loss in trading commodity futures and options products. Losses in excess of your initial investment may occur. Past performance is not necessarily indicative of future results.