The Hidden Cost of Your Own Trade, Explained Simply

Imagine you walk into a busy vegetable market and ask for two hundred kilograms of tomatoes. The first few kilos cost the posted price. But word spreads fast, and by the time you reach kilo one hundred, the price has crept up. You did that. Your own appetite moved the market.

Financial markets work the same way, just faster and with fewer tomatoes. When a large investor buys or sells a big chunk of something, their own order shifts the price. That shift is called market impact, and estimating it — before and after the trade — is one of the quiet crafts of professional trading.

Why a big order costs more than its sticker price

Start with a simple definition. Market impact is the difference between the price that would have existed if your order had never been sent, and the price you actually ended up paying. It's a cost you create simply by showing up.

This cost splits into two pieces. Temporary market impact is the immediate jolt your buying or selling gives the price — like the tomato seller raising his price the moment he sees you coming. In a deep, liquid market, that jolt fades as other participants step in and absorb it. Permanent market impact is the part that sticks around, the lasting change your trade leaves in the price.

Think of it like tossing a stone into a pond. There's the splash — temporary — and there's the ripple that keeps travelling outward long after the stone sinks. Traders care about both.

The measurement toolkit: transaction cost analysis

Professionals measure these costs through something called Transaction Cost Analysis, or TCA. At its heart, TCA compares the price you actually got against a benchmark price — usually the price at the moment the trading decision was made. The gap between the two is called implementation shortfall: the cost of turning an idea into reality.

That gap has several ingredients. Some are fixed, like commissions. Others are random, like the spread — the small difference between the price someone will buy at and the price someone will sell at. Market impact is one of these random, shifting costs, and it depends heavily on what the market looks like when you press the button.

Two traders can place the same order on the same day and pay very different prices, simply because one caught a calm, liquid hour and the other waded into a thin, jumpy one.

Going slow costs less — until it doesn't

Here's the natural fix: if trading fast moves the price against you, trade slowly. Slice the big order into smaller pieces and spread them over time. This is called a trading schedule, and it's where the well-known Almgren-Chriss model earns its keep.

Published by Robert Almgren and Neil Chriss in 2000, the model frames execution as a balancing act. Trade too quickly, and you pay a heavy impact cost. Trade too slowly, and you're exposed to timing risk — the danger that the price drifts away from you while you're still holding shares you meant to sell. Nobody enjoys watching a price run off without them.

If you only cared about impact, the maths says run a steady, evenly paced schedule — known as a TWAP, or time-weighted average price — and stretch it out as long as possible. But a very long TWAP means a very long period of not knowing what the market will do. So the model introduces a risk-aversion knob. Turn it up, and the schedule front-loads: most of the trading happens early, then tapers off. Turn it down, and you drift back toward the even pace.

Each setting on that knob gives one expected cost and one expected variance. Sweep across all settings and you trace what's called the efficient cost-risk frontier — the set of schedules where you can't reduce expected cost without accepting more uncertainty. It's a menu, not a magic answer.

The square-root surprise

Now for the part that genuinely surprises newcomers. When researchers measure how prices actually move in response to large orders, they keep finding something neat: impact grows with the square root of the total volume traded.

Double the size of your order, and impact rises by roughly forty percent — not double. Ten times the volume, and impact rises about threefold. This is called the square-root law, and it's shown up across equities, futures, options and crypto. It also appears largely independent of how long you took to execute, provided you're not taking too big a share of the daily volume.

That last detail matters. It suggests the cost mostly depends on how much you trade, not how cleverly you slice it. You can read that as liberating or slightly humbling, depending on your mood.

Recent research on the Tokyo Stock Exchange, using complete trading records from 2012 to 2018, found this square-root pattern holds even at the level of individual child orders — provided the market gets time to digest them. Researchers also built "synthetic" orders by shuffling which trader sent which trade. The impact looked the same. That points to something mechanical about impact — a matter of supply, demand and available liquidity — rather than a signal that traders know something special.

Some studies go further and suggest a double square root is at work: each individual order moves prices with a square root of its own size, and its effect then fades away over time in inverse-square-root fashion.

Why this matters beyond the trading desk

There's a bigger implication. If impact were a brief hiccup, it would be a minor operational nuisance. But evidence suggests impact decays slowly — over days or weeks — and only partially. Some research argues that buying one dollar of the US equity market can raise total market capitalisation by several dollars over months. One person's trading cost, in other words, quietly becomes part of the price everyone else sees.

Traders increasingly estimate impact before the trade as well as after, using historical data, volume profiles and liquidity assessments. Execution algorithms, which follow instructions to buy or sell a set amount of currency or stock, help institutional users manage these costs. But a report from the Bank for International Settlements notes they transfer execution risk from dealers to end users and raise the bar for anyone trying to navigate the market — they need data, skills and tools, and so do the central banks watching them.

There's also a caution flag. The same BIS report notes that these algorithms could create self-reinforcing loops that sharpen price swings, though early observations during the pandemic suggested the risk may be less severe than feared. The honest position: that question isn't fully settled.

What to carry with you

Market impact is real, measurable in principle, and stubbornly hard to pin down in practice. Models like Almgren-Chriss give a sensible structure. The square-root law gives an empirical anchor. But every model rests on assumptions — linear impact, well-behaved prices, stable liquidity — that live markets cheerfully violate.

Impact coefficients have to be estimated from data, and if you get them wrong, every number downstream is wrong too. Market conditions differ by venue and by jurisdiction, and rules governing execution and disclosure vary widely. This is education, not a recipe. There's no single right answer sitting in a formula, waiting to be discovered.

What there is: a clearer understanding of why rushing costs money, why patience carries its own price, and why the market always notices when you show up hungry for tomatoes.