The Wolves Smell Blood: Brunnermeier and Pedersen on Predatory Trading

Almost every model of optimal execution shares a comfortable assumption. The market is a passive thing you push against. It has an impact function, it has a decay rate, and it responds to your trading mechanically, like a spring. It does not have opinions about you. It does not notice you. It certainly does not try to hurt you.

Markus Brunnermeier and Lasse Heje Pedersen's paper is about what happens when the market does notice, and the other participants are not springs but wolves.

The problem: your distress is somebody else's opportunity

Suppose a large fund is in trouble and must liquidate a big position. Perhaps it has hit a risk limit, blown through a margin call, or its investors are pulling money. Whatever the reason, it has to sell, and reasonably soon, and this is known or guessable by others.

The textbook view of what should happen is reassuring. Other traders, seeing a temporarily depressed price, should step in and buy. They are providing liquidity, they earn a fair return for it, the distressed fund gets a decent exit, and the market has done its job.

Brunnermeier and Pedersen show that under quite plausible conditions, the rational thing for those other traders to do is the exact opposite.

The key idea via analogy: front-running the fire sale

Imagine you know a neighbour is about to be forced to sell their car this weekend, at any price, because they need cash by Monday. You are also in the market for cars.

The naive strategy is to wait until Sunday and buy their car cheap. But there is a better one, if you happen to already own a similar car.

Sell your own car first, on Saturday, into the same market. That pushes the local price of used cars down. Now your neighbour, forced to sell on Sunday into a market you have just depressed, gets an even worse price. And then, on Monday, once the forced selling is finished and the panic has passed, you buy back in at the bottom.

You have made money twice over: once by selling before the crash you helped cause, and once by buying the bottom you helped create. And your neighbour, the distressed seller, got a far worse price than they would have if you had simply stood aside and bought from them.

That is predatory trading, and the paper's central insight is that it is not a market failure or a rogue act. It is what a rational, profit-maximising trader should do when they know somebody else is a forced seller. The strategy is: sell into the distress, then buy back the overshoot.

The consequences the paper draws out are what make it important.

Liquidity vanishes exactly when it is needed most. In normal times, a large seller finds willing buyers. In a crisis, when everyone knows a fund is being liquidated, the buyers turn into sellers. The market becomes illiquid precisely at the moment liquidity matters, and it does so not by accident but by the logic of the situation.

The price overshoots and then reverts. Because predators are pushing the price down beyond what the fundamentals justify, and then buying it back, you see the characteristic pattern of a violent drop followed by a sharp recovery once the forced selling is done. The overshoot is the predators' profit.

A trader can profit by triggering a crisis, not merely by exploiting one. This is the darkest implication. If pushing the price down far enough will cause a leveraged fund to hit its margin call and become a forced seller, then it can be profitable to do the pushing. You are not waiting for blood in the water. You are cutting the swimmer.

Contagion is a natural consequence. A distressed fund does not just hold one asset. When it is forced to liquidate, it sells everything it owns, which drags down assets that have nothing fundamentally in common with each other except that this fund held them both. Predators anticipating this will attack those other assets too. Distress spreads across markets and across traders through no channel other than the shared balance sheet of the victim.

Why it mattered

• It put a strategic adversary into the execution problem. This is the paper's deepest contribution. If your counterparties can infer that you are a forced seller, then the whole optimal execution literature is solving the wrong problem, because it assumes a mechanical market. Any large liquidation must now worry not just about its impact but about information leakage, because leaking your intentions turns the market against you.
• It explains why real crises look the way they do. Sudden, violent, self-reinforcing collapses followed by sharp reversals, with contagion across unrelated assets, are hard to explain with mechanical impact models. Predatory trading explains all three with one mechanism.
• It gives a rigorous account of why traders hide. Dark pools, iceberg orders, randomised execution schedules, careful broker selection: all of these are defences against exactly the threat this paper describes. The paper is the theoretical justification for the entire secrecy apparatus of institutional trading.
• It reframes liquidity as strategic rather than structural. Liquidity is not a property of an asset. It is a property of what the other participants believe about you.

The honest limitations

• The predators need to know you are distressed. The whole mechanism turns on other traders knowing, or confidently inferring, that a forced sale is coming. In practice this information is imperfect. A fund that successfully conceals its distress is much harder to prey on, which is exactly why concealment is so highly prized.
• Predation is risky and can backfire. If a predator sells aggressively expecting a distressed seller to follow, and the distressed fund is rescued, refinanced, or turns out not to have been distressed at all, the predator is short into a rising market. The paper's equilibrium is cleaner than the real gamble.
• The line between predation and liquidity provision is blurry. A trader who declines to buy from a distressed seller and instead waits for a lower price is behaving prudently, not criminally. A trader who actively sells to induce the collapse is doing something else. In practice, these shade into one another, and regulators have found the distinction extremely difficult to draw.
• It is a stylised model with few players. Real markets have many participants with heterogeneous information and beliefs. The stark predator-and-prey structure is a caricature, deliberately so.
• Competition among predators limits the profits. If many traders spot the same distressed seller, they compete to sell first, which erodes the profit and can restore something closer to the competitive outcome. The paper is aware of this and it tempers the conclusions.

The one-line takeaway

Brunnermeier and Pedersen showed that when the market knows you are a forced seller, rational traders will not buy from you, they will sell ahead of you and buy back the overshoot, which means liquidity evaporates precisely when it is needed most, distress spreads contagiously across unrelated assets, and the entire optimal execution literature's assumption of a passive, mechanical market quietly collapses.