Why Passive Orders Get Picked Off: Adverse Selection Explained

Imagine you run a small tea stall. You post a price: 20 rupees a cup, all day, no haggling. Most passers-by pay it and walk off happy. Then one afternoon a man in a hurry buys ten cups, and you notice the tea leaves wholesaler down the street has just raised prices. You were the last person selling cheap. You did not lose anything on paper yet, but you will feel it tomorrow when you restock.

That is the everyday version of a market idea called adverse selection — and it happens to anyone who lets an order sit and wait to be filled.

First, what a limit order actually is

Most markets offer two simple ways to trade. A market order says “get me in or out now, whatever the price.” It is fast and certain, but you pay for the certainty.

A limit order says “I will buy, but only at this price or better.” It does not execute straight away. Instead it sits in a queue — the limit order book, which is simply the running list of everyone’s resting buy and sell instructions — waiting for someone to come along and trade against it.

Posting a limit order is sometimes called being passive or providing liquidity. The appeal is obvious: instead of crossing the gap between the best buy and best sell price (the bid-ask spread), you might earn that gap yourself. You are, in effect, the tea stall that always has stock ready.

The catch is what you learn when your order finally gets filled.

The fill you wanted versus the fill you got

Here is the uncomfortable part. Your limit order does not fill because the world politely decided your price was good. It fills because someone wanted to trade with you — and that someone may know something you do not, or may simply have gotten there first.

Think about a buy limit order resting below the current price. It fills when sellers push the price down to your level. Sometimes that is harmless noise. But sometimes the price keeps falling after you are filled, and your freshly bought position is immediately worth less than you paid. The order you were quietly hoping would fill did fill — precisely because it was about to become a bad trade.

Researchers describe this as being “picked off,” and studies of liquid futures markets have found that a large share of limit order fills turn out to be adverse in this way. Many people who simulate trading strategies that ignore this effect end up with results that look far better than what real trading delivers.

Why this is a structural feature, not bad luck

It helps to know who is on the other side. Broadly, order flow comes in two flavours.

Informed flow comes from participants acting on a view or an edge — perhaps they have faster data, better models or simply saw the news first. Trading against them tends to cost money.

Uninformed flow is everyday, more or less random activity. Trading against it tends to earn the spread.

Market makers do not treat those two the same. They cannot always tell them apart in the moment, so they price for the risk that any given counterparty might be informed. When that risk looks high, they widen their quotes. When it looks low, they tighten. Over time, the cost of protecting against sharp flow gets shared by everyone using the market.

The key asymmetry is this: an order sitting in the book is available to anyone who wants it. The person choosing to trade against it gets to decide. That is a real advantage, and it explains why resting orders tend to be hit at the moments that suit the taker.

Not every market hides information the same way

How much adverse selection you face depends heavily on market design, and this is where things get genuinely interesting — and genuinely uncertain.

In traditional markets, order flow is often separated by route. Retail orders travel through brokers and ordinary connections; professional firms pay for faster links and proximity to the exchange’s matching engine. That gap means a slow retail order and a fast professional order rarely race each other head-on.

On public blockchains, that separation largely disappears. A casual swap from a phone wallet, an institution’s programmatic trade, and an arbitrageur’s automated move can all appear in the same public queue, looking much the same. Because intent is visible before it executes, liquidity providers onchain often assume the worst — that every trade might be informed — and price accordingly. The result is wider spreads for everybody, including people with no special information at all.

Different venues have tried different fixes. Some batch trades together and settle them at one price, so being first no longer pays. Some give market makers a brief window to cancel stale quotes. Some let fast traders self-identify through a priority lane and pay for it. Some route orders privately so nobody can see them coming. Some adjust fees automatically when volatility rises. None of these is free, and each shifts who bears the cost. In some jurisdictions these design choices are also shaped by regulation, so what is available where you live can differ meaningfully.

What this means if you place limit orders

None of this argues against limit orders. They are a legitimate, useful tool, and for patient investors they can reduce costs. But a few honest expectations help.

A resting order that fills instantly and comfortably may not be a gift. It may mean the market was already moving your way, and someone was happy to take the other side before you repriced. That is not a scandal. It is the price of being passive.

It also means that backtests and strategy simulations deserve suspicion when they assume fills and price moves are unrelated. If a model hands you a fill every time the price touches your level and never asks whether the price kept going, it is quietly handing you money that reality will not.

And it means “free” liquidity is never quite free. Someone is absorbing the risk that you might be the informed one, and they charge for it — through wider spreads, higher fees, or slower execution.

The honest summary

Adverse selection is not a flaw in the system or a sign you did something wrong. It is the natural consequence of letting other people choose when to trade with you. The queue pays you for waiting; the price of that payment is the occasional fill you would rather not have had.

Understanding it will not tell you what to buy, and nothing here is a recommendation or a promise about returns. But it will stop you being surprised — and in markets, not being surprised is worth a great deal.