Portfolio Insurance and the 1987 Crash

On October 19, 1987, the Dow Jones Industrial Average fell 22.6% in a single day, the largest one-day percentage drop in its history. A significant part of what turned an ordinary correction into a crash was a hedging technique that thousands of institutional funds were running simultaneously, called portfolio insurance, whose mechanical rule was to sell stock index futures automatically as prices fell, to synthetically replicate the payoff of a put option. The idea worked fine for any individual fund acting alone. It failed catastrophically once nearly everyone was running the same rule at the same time.

The mechanical feedback loop

Portfolio insurance, as practiced in the mid-1980s, used a dynamic hedging rule: as the market fell, the strategy called for selling a larger fraction of the portfolio's equity exposure via index futures, and as the market rose, buying it back — continuously adjusting to keep downside losses capped, without ever needing to buy an actual option. On a normal day with modest price moves and diverse participants, this works as intended: an individual fund sells a manageable amount into normal liquidity. On October 19, the market opened sharply lower on prior-week weakness, and portfolio insurers' models called for sizeable futures selling right at the open. That selling pushed futures prices down, which pushed the model's estimate of the "right" amount to sell even further down, triggering more selling, from the same funds and from others running similar strategies — a mechanical loop with no natural stopping point, made worse because the futures market couldn't absorb the size being demanded, causing futures to trade at a large discount to the actual index and confusing market makers about the fair price of stocks altogether.

Worked example

A pension fund running a portfolio insurance program holds $500 million in equities and is instructed by its model to hedge 20% of that exposure via S&P 500 futures once the market falls 5% from its recent high. On October 19, the market gaps down more than 5% at the open, and the fund's model, updating continuously through the day, calls for hedging closer to 60% of the portfolio's value as the decline deepens past its programmed thresholds — meaning the fund needs to sell roughly $300 million notional of futures, concentrated within hours, in a futures market that on a normal day might trade a small fraction of that in comparable size without moving price. The resulting futures selling pushes the futures index to trade at a discount of several percentage points below the still-lagging cash index, which index arbitrageurs then try to exploit by selling the underlying stocks and buying the cheap futures — adding still more selling pressure to the cash market and completing the loop back to the equities that started it.

What this means in practice

The 1987 crash led directly to circuit breakers — automatic trading halts triggered by large index moves, designed specifically to break exactly this kind of mechanical feedback loop by forcing a pause before selling programs could re-trigger each other — and it remains the standard case study for why a hedging strategy that is individually rational can be collectively destabilizing when correlated across enough capital.