Mark-to-Market: Booking Tomorrow's Profits Today
The single accounting decision that let Enron report a deal's entire estimated future profit the moment it was signed — not as it was actually earned.
In 1992, Enron won SEC approval to use mark-to-market accounting for its long-term energy trading contracts. Under ordinary accrual accounting, a company recognizes revenue and profit from a long-term contract gradually, as it's actually delivered over the contract's life. Under mark-to-market accounting, a company instead estimates the entire contract's future value today and books that estimated profit immediately, in the current quarter — then adjusts the estimate in later periods as actual conditions unfold.
For a genuinely liquid, easily-priced asset — a publicly traded stock, say — mark-to-market is a reasonable, standard practice, because the "estimate" is just today's observable market price. The book's reporting documents how differently this worked for Enron's actual contracts: many were for exotic, illiquid, long-duration deals (some running twenty years) with no real observable market price at all, meaning the "estimate" underlying the immediately-booked profit was, in practice, a number Enron's own models produced.
| Accrual accounting | Mark-to-market accounting | |
|---|---|---|
| When profit is recognized | Gradually, as the contract is actually delivered over its life | Immediately, based on the estimated value of the entire future contract |
| What the reported number depends on | Cash and deliveries that have actually already happened | A model's estimate of future prices, volumes, and costs — not yet realized |
| Effect of a bad estimate | Limited — you haven't booked the profit yet, so there's little to unwind | The already-reported profit can require a large, embarrassing restatement later |
Once a division's reported profit depended on the assumptions inside its own valuation model, the people building those models had a direct, personal incentive — compensation, bonuses, and career standing all tied to reported earnings — to lean toward the more optimistic end of any reasonable estimate. The book documents that Enron's own auditors and even some of its own executives later acknowledged that mark-to-market values on the company's more exotic, long-term deals were, in practice, extremely difficult for anyone outside the models to independently verify.
The incentive problem compounds because the same people building the models were often the ones whose compensation depended on the models' output — a structural conflict distinct from any individual's honesty. Even a careful, well-intentioned analyst asked to estimate the value of a twenty-year contract has real latitude in choosing assumptions about future prices, volumes, and discount rates, and when every reasonable assumption within that latitude points toward a bigger bonus, the aggregate effect across many deals and many quarters tends to drift optimistic even without any single person consciously deciding to inflate a number.
Consider a hypothetical twenty-year contract to supply energy at future prices. Under mark-to-market rules, Enron could estimate the total profit across all twenty years using an internal model's assumptions about future energy prices, and book a large chunk of that estimated profit in the current quarter — years, in some cases, before a single dollar of it had actually been collected. If the twenty-year forecast later proved too optimistic, the shortfall wouldn't show up until much later, often after the executives who booked the original estimate had already been paid a bonus tied to it.
Accounting rules generally require that a value used for financial reporting be reasonably estimable and verifiable — standards that are straightforward to apply to an asset with a daily quoted market price, and considerably harder to apply to a bespoke, multi-decade energy contract with no comparable transactions to check against. The book's reporting documents that Enron's auditors repeatedly signed off on mark-to-market values for contracts of exactly this second kind, a judgment call that, in hindsight, extended the technique well past the kind of asset it was originally designed to value.
This is the specific reason the mark-to-market chapter of Enron's story became the most-cited accounting lesson from the collapse: it isn't a story about a technique that was inherently flawed, but about a technique's reliability depending entirely on the verifiability of the estimate underneath it — a distinction any investor evaluating a company's use of estimate-based accounting can still apply today.
- Mark-to-market accounting is a legitimate, standard practice for genuinely liquid assets — the issue documented in the book is specifically its use on illiquid, model-dependent, multi-decade contracts with no real market price to check the estimate against.
- The core problem wasn't that the technique was invented improperly — it's that it let reported profit diverge, sometimes for years, from actual cash collected, with compensation tied to the reported number rather than the eventual cash.
- The incentive problem didn't require individual dishonesty to compound — reasonable assumption latitude, combined with compensation tied to the output, tends to drift optimistic in aggregate across many deals.
- The reliability of any estimate-based accounting method depends on how verifiable the underlying estimate actually is — the same question worth asking of any company using similar techniques today.
- This chapter's mechanism — profit recognized on a model's estimate rather than on delivered results — recurs, in a different form, in nearly every later chapter of this course.