Enron and the Accounting Trick That Faked $74 Billion in Earnings

In December 2001, the seventh-largest publicly traded company in the United States — a company with reported revenues of $101 billion in the prior year, with offices in forty countries, with twenty-five thousand employees, with a corporate headquarters tower in downtown Houston so distinctive it had become a city landmark — filed for bankruptcy. The filing was, at the time, the largest in American corporate history. The company's stock, which had traded above $90 per share fourteen months earlier, was worth twenty-six cents.

Within six weeks of the filing, the company's auditor — Arthur Andersen, one of the five largest accounting firms in the world — would itself be indicted, convicted of obstruction of justice, and effectively dissolved. Within a year, Enron's chairman would be indicted on six counts of fraud and conspiracy. Within three years, the company's CEO would be sentenced to twenty-four years in federal prison.

The collapse of Enron is taught, when it is taught at all, as a story about corporate fraud and personal greed. Both elements were present and important. But the deeper story — the one most relevant to operators trying to understand how an apparently legitimate company can sustain a fraud at this scale for years — is about a single accounting choice the company made in the early 1990s that, by 1999, was the only thing keeping the company alive.

Mark-to-market: the rule that changed everything

For most of the twentieth century, American accounting standards required companies to recognise the revenue from a long-term contract as it was actually earned over the contract's life. A natural-gas company that signed a twenty-year supply agreement worth, say, six billion dollars in total revenue would record approximately three hundred million per year in revenue across the twenty years of the contract. The total revenue was the same. The recognition was spread across the period in which the work was done.

In 1991, Enron — under the direction of Jeffrey Skilling, who had recently joined as the head of the company's trading operation — petitioned the Securities and Exchange Commission for permission to use a different accounting method for its long-term energy contracts. The method Skilling proposed was called mark-to-market accounting. Under mark-to-market, the company would estimate the total future profit of a long-term contract at the moment the contract was signed, and book that entire profit as current-year earnings.

This was a radical departure from energy-industry accounting practice, and the SEC was sceptical. Enron made the case, persuasively, that its trading operations functioned more like a financial-services business than a traditional energy company, and that mark-to-market — the standard for trading firms — should apply. After more than a year of negotiation, the SEC granted approval in 1992.

The decision is, in retrospect, one of the most consequential regulatory rulings in American business history. It changed the structure of Enron's incentives in a way the SEC almost certainly did not appreciate at the time.

What mark-to-market actually did to the company

Under mark-to-market, the moment Enron signed a contract, it could record a number on its income statement that represented the contract's projected lifetime profit. That number was based on Enron's own internal estimate of future commodity prices, future interest rates, future volume, and future operating costs over the contract's term.

Two effects compounded. First, the company's reported earnings became almost entirely unmoored from the company's actual cash position. A contract that Enron signed today might not produce a single dollar of cash for years, but Enron could book hundreds of millions of dollars of "earnings" from the contract this quarter. Second, the size of those earnings was determined by Enron's own forecasts of variables (future gas prices, future weather, future regulatory environments) over time horizons of ten to thirty years. The forecasts were not, in any meaningful sense, audited. Arthur Andersen, the auditor, was technically required to verify them, but the verification was largely deferential to Enron's own projections.

The company quickly discovered that aggressive forecasting could produce essentially any earnings number it wanted. A long-term contract that more conservative assumptions would have valued at three hundred million dollars could, with slightly different assumptions about future gas prices and contract performance, be valued at nine hundred million. The conservative number was probably closer to truth. The aggressive number was the one Enron's traders and executives were compensated on.

By the late 1990s, the gap between Enron's reported earnings and Enron's actual cash flow had become enormous. The company was reporting record profits while its operating cash flow was deeply negative. The mismatch was, in retrospect, the most visible warning sign anyone looking at the company carefully should have caught. Almost no one looked.

The special purpose entities that hid the losses

The second mechanism that allowed Enron's fraud to scale was a structure called special purpose entities (SPEs) — separate legal partnerships that Enron created and controlled but that, under the accounting rules of the period, did not appear on Enron's balance sheet.

The SPEs were used, increasingly aggressively from about 1997 onwards, to absorb losses that Enron did not want to report. When a contract that had been booked at a large profit under mark-to-market began to underperform — when actual gas prices diverged from Enron's forecasts, or when a counterparty went bankrupt, or when a regulatory ruling went against Enron — the company would, instead of recognising the loss, transfer the underperforming contract or asset to one of its special purpose entities.

The transfer technically removed the asset from Enron's balance sheet. The SPE then "owned" the underperforming asset and absorbed the loss. The SPE's losses, because the entity was technically separate from Enron, did not flow through to Enron's reported earnings. Enron continued to report the original mark-to-market profit on the contract; the loss disappeared into a structure the financial statements did not show.

By 2000, Enron had created hundreds of these entities. Some were named for Star Wars characters (the most famous, a fund called LJM — for Lea, Jeffrey, and Matthew, the names of Enron CFO Andrew Fastow's wife and two sons — was managed by Fastow himself, in what was an obvious conflict of interest that Enron's board explicitly approved). Others were named for the deals they were created to absorb. Collectively, the SPE structure was holding billions of dollars of losses that did not appear on Enron's published financial statements.

The collapse, and what triggered it

Through 2000 and into 2001, Enron's stock continued to climb. The company was named the most innovative company in America by Fortune magazine for the sixth consecutive year. The company's executive team was profiled, repeatedly, as the smartest in the energy industry.

What broke the structure was a relatively obscure analyst named Jonathan Weil, who in September 2000 published a piece in the Wall Street Journal questioning the gap between Enron's reported earnings and its operating cash flow. The article was technical and did not get widespread attention at the time. But it triggered a slow shift in attitudes among other analysts and short-sellers, who began to look more carefully at the disclosures the company was making about its derivative positions and SPE structure.

The decisive event came in August 2001, when Enron's CEO, Jeffrey Skilling, abruptly resigned with no explanation. Skilling had built mark-to-market accounting at the company. He had championed the SPE structure. He had been the public face of the company's growth narrative for a decade. His sudden departure, with no compelling personal reason offered, was the signal short-sellers had been waiting for.

Within two months, the company was forced to disclose a $1.2 billion write-down related to two of its SPEs. The disclosure triggered cascading questions about the rest of the SPE structure. Within six weeks of that disclosure, the company's credit rating was cut to junk, its trading partners stopped extending credit, its commercial paper market evaporated, and it became unable to refinance its short-term debt. Enron filed for bankruptcy on December 2, 2001.

The cumulative restatement of Enron's financials, after the bankruptcy, removed approximately $74 billion of fictitious earnings the company had reported between 1997 and 2001.

What the case actually demonstrates

The Enron story is told, in business school casebooks, as a study in corporate ethics and the risks of executive compensation tied too tightly to short-term stock performance. Both elements are real. But the deeper structural lesson is about what happens when accounting rules are written for one kind of business and applied to a fundamentally different kind.

Mark-to-market was a perfectly reasonable accounting method for a financial trading firm with liquid, observable market prices for the assets it held. Goldman Sachs uses mark-to-market for its bond inventory because there are public markets in which those bonds trade every day. The valuation is, while imperfect, anchored to observable reality.

The SEC, in 1992, allowed Enron to apply mark-to-market to long-term energy contracts where there were no liquid markets, where the underlying commodities had to be projected decades into the future, and where the company itself was generating the projections. The same accounting rule, applied to fundamentally different inputs, produced fundamentally different consequences. It produced, eventually, the largest accounting fraud in American history up to that point.

The pattern is worth recognising in any context where new financial instruments, new accounting methods, or new structuring techniques are being applied to industries they were not designed for. The history of financial regulation is, in large part, a history of these mismatches — rules that worked in one context, applied to another, until the second context produced a catastrophic failure that forced new rules. Enron was that failure for mark-to-market in non-financial industries. It is still, more than two decades later, the cleanest case study of the pattern available.