Do Hedge Funds Use Stop-Loss? What Funds and Research Show

Some hedge funds use stop-loss rules and some do not, and the documented rules rarely look like a retail stop order. They work at the level of the trader or the whole fund: a drawdown limit on each portfolio manager's book, a cut in size after losses, or a clause that closes the fund when cumulative losses breach a threshold. Even at the trade level, large traders do not always leave a stop order with their broker: the Turtles, who traded large futures positions for Richard Dennis, were told to keep a stop price in mind and exit with a limit or market order when it was hit, so as not to reveal their positions.

This page sticks to what is published: how the large multistrategy funds are reported to apply stop-losses, what academic studies find, and how a trader can copy the parts that matter.

How big funds apply stop-losses

Millennium is one documented case. eFinancialCareers, summarizing a Wall Street Journal report in September 2024, wrote that "if you're down more than 7.5% at Millennium, you're almost certainly out", and that 15% to 20% of people leave the firm every year, typically after being stopped out for a loss over 7.5%. The same article says that Balyasny presents itself as more conversational about losses, and that Citadel does not operate a strict stop-loss strategy either.

The last two rows come from the rules that Richard Dennis's trainees followed, published free by one of them; they are detailed on our Turtle Traders page. The size rule is concrete: a Turtle with a $1,000,000 notional account who lost 10% traded as if the account were $800,000, and after another 10% loss, of $80,000, as if it were $640,000, until the yearly starting equity was regained. N is the Turtles' measure of a market's daily volatility, and each Unit was sized so that a 1 N move equaled 1% of the account, so a stop at 2 N risked 2%.

Do stop-losses help? What the research says

Kathryn Kaminski and Andrew Lo studied the question in "When Do Stop-Loss Rules Stop Losses?" (Journal of Financial Markets, 2014). Their working paper states that "under the Random Walk Hypothesis, simple 0/1 stop-loss rules always decrease a strategy's expected return, but in the presence of momentum, stop-loss rules can add value." Applied to U.S. equities, with long-term government bonds as the asset held while stopped out, certain rules added 50 to 100 basis points per month to buy and hold during stop-out periods, on monthly data from January 1950 to December 2004. The journal abstract adds that, at longer sampling frequencies, "certain stop-loss policies can increase expected return while substantially reducing volatility".

A 2022 study by Xin Hong, Ningjing Pang and Zhibin Wang (Pacific-Basin Finance Journal) looked at fund-level clauses that force a hedge fund to close when its cumulative losses breach a threshold. It finds them in 14% of hedge funds in China, finds that funds with these clauses generate higher future performance, and that funds become more risk-averse as they approach the threshold. The authors attribute the better performance to prompt terminations of poorly performing funds.

The fund's version: the uncle point and portfolio heat

Ed Seykota, in his essay on risk management, names the "UNCLE POINT": "the amount of draw down that provokes a loss of confidence in either the investors or the fund management". He warns that if either side becomes demoralized and withdraws, the fund dies, and that this point often gets little mention when the fund is sold. He calls the total risk of all open positions the "portfolio heat". A manager's stop-loss is the fund's way of fixing its uncle point in writing before the drawdown happens. More on his methods is on our Ed Seykota page.

The Turtles capped their heat in Units, each Unit sized to the same volatility risk: at most 4 in a single market, 6 in closely correlated markets, 10 in loosely correlated ones and 12 in one direction. Their rules explain why: the day after the October 1987 crash, when the Federal Reserve lowered rates overnight, Turtles loaded long in interest rate futures lost 20% to 40% of account equity in a single day in some cases. A stop cannot help when the market opens far beyond it; only the size limit capped that loss.

A worked example (hypothetical)

Take a hypothetical portfolio manager given $100 million of capital under a 7.5% drawdown limit, the threshold reported for Millennium. Adding the Turtle size cut, here at half the limit, gives three levels.

The figures are our own arithmetic on a hypothetical book, not data from any fund. The point is that the manager knows before the first trade the maximum loss the firm will accept, $7.5 million here, and every position is sized so that a normal losing streak stays well short of it. Our page on the risk of ruin shows how fast a string of losses approaches such a limit.

Applying it to a single trade

At the trade level, the tools are the same as for any trader: a stop-loss order set before entry, sized so that the loss is a fixed fraction of the account, and a trailing stop that follows a winning position.

The Turtles' rule ties both together: no trade risks more than 2% of equity, the stop sits 2 N from the entry, and when a Unit is added the stops of the earlier Units are raised by ½ N, so that they all sit 2 N from the latest entry. Other risk stories, good and bad, are in the famous traders section, including the worst hedge fund losses.