The $4.9 Billion Bet: Berkshire's Long-Dated Index Puts

The apparent contradiction

Warren Buffett described derivatives as "financial weapons of mass destruction" in Berkshire Hathaway's 2002 shareholder letter. It is one of the most quoted lines in finance.

Between 2004 and 2008, Berkshire sold equity index put options on major global indices and collected roughly $4.9 billion in premium, according to the company's 2008 annual report.

Both things are true, and treating them as a contradiction misses the point he was making. The 2002 warning was about counterparty risk, opacity and leverage in complex derivative books. The index puts were the opposite arrangement: simple structure, premium received up front, no collateral posting on ordinary market moves, and no obligation for well over a decade.

The structure

The contracts were long-dated European-style puts on indices including the S&P 500, written with expiries roughly fifteen to twenty years out.

Three features made the position unusual.

The premium arrived immediately and was investable. Berkshire received the money at inception and could deploy it for the entire life of the contract. That is the insurance float model applied to an options position: use of the cash for fifteen years, with the possible claim arriving at the end.

European exercise meant only the final day mattered. They could not be exercised early. Where the index sat in year six was irrelevant to the obligation; only the level at expiration counted.

The obligation was capped by the index reaching zero, which is a bound so theoretical it is close to meaningless — but it means the exposure was defined rather than open-ended.

The bet underneath was simple: over fifteen to twenty years, the market ends higher than it started. That is not a forecast about any particular period. It is a statement about the long-run direction of equity markets, expressed through an instrument the market was pricing on much shorter-horizon assumptions.

The part that gets left out

The story is usually told as a clean win. It was not clean.

Accounting rules required Berkshire to mark the contracts to market every quarter. During 2008 and 2009, as markets collapsed and volatility reached record levels, those marks moved dramatically against the position — reported mark-to-market losses reached around $10 billion.

Ten billion dollars of paper losses, on contracts with more than a decade left to run, reported publicly, quarter after quarter, against a position that could not be exercised early and had not lost anything real.

This is the most instructive part of the whole episode and it is the part the summaries skip. Berkshire held. The contracts began expiring in June 2018, with the last of them running to January 2026. The premium was kept.

What made holding possible

It is worth being precise, because "he held his nerve" is the wrong lesson to take.

The structure removed the pressure. European exercise meant nobody could call the position in during the worst of it. The marks were an accounting event, not a cash event.

The terms were negotiated. Berkshire arranged contracts that required little or no collateral posting on ordinary moves. A standard margin arrangement would have generated calls precisely when markets were at their worst.

The time horizon was genuinely long. Fifteen years leaves enormous room for recovery. A one-month contract has none.

The balance sheet could absorb it. Berkshire was never at risk of being forced to act.

Take away any one of those and the same trade becomes very dangerous. A retail seller writing short-dated puts on margin has none of them — no exercise protection, collateral requirements that tighten in stress, weeks rather than decades, and an account that can be liquidated.

What actually transfers

Three things, and they are structural rather than a matter of nerve.

The market chronically overprices long-horizon downside protection. Buyers of protection are frequently insurance-motivated rather than price-sensitive — who buys put options covers why pension funds and insurers pay for cover they expect to expire worthless. Consistent, price-insensitive demand pushes premiums above what the statistics alone would justify. That gap is the seller's structural edge, and it exists at retail scale too.

Being paid to accept an obligation is a real business. Berkshire was doing what it does in insurance, using a different instrument. The economics are underwriting economics: many small premiums, rare large claims, profitable if the ratio holds.

Marks are not losses. A short option position can look catastrophic and finish worthless. This is the single most useful takeaway for anyone running open positions through a bad stretch — what happens between opening and closing a trade covers the retail version of the same experience.

What does not transfer

The horizon. Fifteen years of recovery time is not available on a six-week contract. A retail seller's position is decided on one specific morning, and there is no version of "wait it out" that reaches past expiration.

Immunity from being forced out. This is the big one. Berkshire's structure and balance sheet meant nobody could make them close. A margin account in a drawdown offers no such protection, and the moment of maximum stress is exactly when the pressure to act arrives.

Scale. Terms negotiable at that size are simply not on offer to anyone else.

The certainty of the premium being kept. It worked. Fifteen years of equity market history could have gone otherwise, and the fact that it did not is one observation, not a law.

The honest lesson

Not "Buffett sells puts, so selling puts is validated". A great deal of the outcome came from structure and capacity that no individual has.

The transferable lesson is narrower and more useful: the edge in selling downside protection is structural, not predictive. You are not paid for knowing where the market is going. You are paid for accepting an obligation that others are systematically willing to overpay to avoid — and how much of that payment you keep depends almost entirely on whether your structure lets you hold the position to its conclusion.

Berkshire's structure did. Most do not, and that is the part to design around.

How StratosIQ approaches it

The scoring reflects the underwriter's ordering rather than the speculator's: safety carries the largest weight, not premium. The dominant input is the distance between the strike and a defensible floor built by ShieldIQ from how the underlying has actually behaved, reported as Fortified, Secure, Tight or Exposed. A contract failing that floor does not publish, however rich it looks.

Every contract is scored assuming it runs to expiration, which is the only assumption under which the "marks are not losses" lesson holds — a position you may be forced to close early is a different position entirely.

The full published record, contract by contract and including every loss, is on the performance page.

Worth reading next

The time Berkshire sold 50,000 puts on Coca-Cola covers the single-name version from 1993. Who buys put options covers the demand that funds the premium in both.