Margin of Pain

Amaranth / Jun 02, 2026

Amaranth and the Trade That Got Too Big

A natural gas calendar spread is one of the cleanest trades in commodities. By September 2006, Amaranth's version of it was a large share of an exchange contract's open interest. That is not a position. That is a financing problem wearing a position's clothes.

AI-generated editorial still life of a single steel rod bent into a long arc that is visibly sagging and deforming under its own weight
AI-generated editorial illustration. It represents a curve bending under the size of its own exposure, not source evidence from Amaranth Advisors, NYMEX, CFTC, FERC, or any energy market.

On paper, a natural gas calendar spread is about as straightforward as commodity trading gets.

The mechanics are relatively simple: you buy a winter contract, short a summer contract, and bank on the seasonal spread widening as winter heating demand drives prices up while summer storage injections weigh them down. It is a historically reliable, almost mechanical trade—until someone tries to execute it at a scale the market cannot actually absorb.

Brian Hunter did not invent this spread, but he traded it at a volume that redefined what the energy market could handle.

By April 2006, Amaranth Advisors' natural gas book was sitting on nearly $2 billion in year-to-date profits. Five months later, that same book had wiped out roughly $6.6 billion—two-thirds of the fund's entire capital—in a matter of weeks. While post-mortems often focus on whether Hunter's fundamental thesis was flawed, the reality of the collapse is more straightforward: the trading thesis itself was perfectly reasonable, but the sheer volume of the position was not.

The Curve That Had Always Worked

Hunter had arrived at Amaranth in mid-2004 after running a gas desk at Deutsche Bank. A Canadian with a physics degree, he had a reputation for taking a highly structured, analytical approach to energy markets.

In 2005, that approach paid off spectacularly. When Hurricanes Katrina and Rita wrecked Gulf Coast infrastructure and left storage levels depleted, Amaranth's long-winter/short-summer position rode the ensuing spike in winter contracts. Hunter's personal payout for the year was rumored to be upwards of $75 million, and his profit share was bumped to 15 percent. Naturally, he was given a longer leash and more capital to deploy.

But the trouble with a highly visible, highly profitable trade is that everyone else has screens too. By late 2005, banks, proprietary desks, and rival hedge funds were crowded into the exact same winter-versus-summer trade, flattening the seasonal curve before the winter season had even started. While Amaranth's historical data models were looking backward at a reliable seasonal pattern, the actual market was already changing under the weight of crowded capital.

When a Position Becomes the Market

By early 2006, Amaranth didn't just have a position in natural gas; they essentially were the market.

Hunter was holding upwards of 100,000 NYMEX futures contracts. According to academic reviews of the collapse, Amaranth controlled anywhere from 30 to over 80 percent of the open interest in specific delivery months. When a single player controls that much of a market, the standard rules of trading break down.

First, you lose the ability to price the asset independently. Any attempt to trim or adjust your position moves the price against you, meaning your own liquidity needs—rather than market fundamentals—dictate the price.

Second, the primary constraint shifts from risk management to basic financing. When clearing brokers and counterparties see that level of concentration, they start demanding higher margin payments. Suddenly, your real risk limit isn't the Value-at-Risk (VaR) model on your dashboard—it's your immediate capacity to post cash.

Finally, there is no viable exit. A book of that size cannot be unwound quietly or quickly; any attempt to liquidate forces the curve to collapse.

The Curve Collapses

As it turned out, the winter of 2006–2007 was exceptionally mild, and gas storage remained highly resilient. The March-April spread—a core bet for Amaranth—collapsed from nearly $2.50 per MMBtu in late August to about $0.58 by the end of September.

In calendar spreads, a compression of that scale is catastrophic when leveraged. On September 14 alone, the fund lost $560 million. Once the margin calls started, the end came quickly. By late September, Amaranth was forced to sell its entire energy portfolio to JP Morgan and Citadel at a massive discount, and the fund announced its liquidation shortly after.

It is a familiar story in finance, but it raises the obvious question of why the firm's internal alarms never went off.

The Risk Report That Said Yes

The failure of risk management at Amaranth wasn't due to a lack of data; the desk had plenty of reports. The real breakdown was cultural. Because Hunter was generating the lion's share of the firm's profits, his trading limits were repeatedly adjusted upward to accommodate his growing positions, rather than forcing him to scale back.

On paper, risk limits are absolute boundaries; in practice, they are often treated as negotiable thresholds when debated face-to-face with a star performer. Academic post-mortems later showed that even a standard, statistically normal drawdown for Hunter's strategy could easily exceed 20 percent—a figure that should have prompted serious discussions about sizing long before the final collapse. But as long as the P&L was positive, those warning signs were easy to ignore.

The Counterweight

To understand why Amaranth failed, you only have to look at John Arnold, who was running Centaurus Energy at the exact same time. Like Hunter, Arnold was young, highly analytical, and focused almost entirely on natural gas.

But while Amaranth was collapsing, Centaurus reportedly made around $1 billion in 2006, posting returns of over 300 percent. The fundamental difference wasn't their market outlook—both often held similar views on the winter gas curve. The difference was execution and size. Arnold kept his positions small enough that he could maneuver and exit when the market turned, whereas Hunter had built a position so massive that he was essentially trapped. It's a reminder that having the "right" macro view doesn't matter if your position is too large to survive a temporary adverse move.

The Verdict

The regulatory fallout took nearly a decade to resolve. In 2007, the CFTC accused Amaranth and Hunter of trying to manipulate NYMEX gas prices during expiry days by aggressively selling contracts in the final minutes of trading to benefit larger positions elsewhere. After years of litigation—including a FERC fine that was eventually thrown out on jurisdictional grounds—Hunter settled with the CFTC in 2014, paying a $750,000 penalty and accepting limited trading restrictions without admitting wrongdoing.

But for anyone managing money, the legal outcome is secondary. Long before the lawyers and regulators got involved, the market had already delivered its verdict. The fatal blow wasn't a regulatory filing; it was the simple mechanics of a margin call on a position that had grown too big to exit.

Ultimately, the Amaranth story isn't a cautionary tale about natural gas trading or seasonal spreads. It is a lesson in how a highly profitable strategy can slowly warp an entire firm's risk tolerance. When a single trader's book grows to represent the majority of a fund's assets and identity, the fund stops managing a portfolio and starts managing a single, highly leveraged bet.

Disclosure: Margin of Pain publishes research and commentary about traders, markets, and risk. This article is not investment advice or a recommendation to buy, sell, short, or hold any security, derivative, futures contract, currency, commodity, or asset.

Source trail