TWAP. Time-Weighted Average Price. An algorithm that chops a large order into smaller trades, spaced evenly across a time window.

Goal: get an average fill price close to what the asset was naturally trading at, without your own order pushing the market around.

The Problem TWAP Solves

You need to buy $5 million of a token. Place that whole order at once and you create an instant demand spike. Price surges before the order even finishes filling. You end up paying way above where the market was sitting when you started.

That's slippage. On orders this size, it runs into hundreds of thousands of dollars.

TWAP handles it by spreading the order out. Instead of one big splash, small consistent bites that don't spook the market or shove the price around.

How TWAP Works

Three settings: total size, time window, interval between trades.

Say you want to sell 1,000 ETH over one hour. TWAP sells roughly 16.67 ETH every minute for 60 minutes. Each individual trade barely registers on the chart. Average selling price lands close to what ETH was doing naturally during that hour.

Without TWAP, dumping 1,000 ETH in a single order crushes the price. You'd get far less per ETH because your own trade wrecked the market before it could finish filling.

Who Uses TWAP

Institutions, hedge funds, automated bots. Anyone whose positions are too large for normal market orders.

Retail traders rarely need it at their typical size. But understanding TWAP explains those stretches of steady, consistent buying or selling that run for hours without an obvious catalyst. Usually a TWAP order ticking along in the background. Serious money moving deliberately.

Binance, Bybit, and OKX all support TWAP orders. Minimum sizes around $1,000 are typical.

TWAP vs. VWAP

Gets confused constantly. Related, not the same.

• TWAP weights price purely by time. Every interval counts equally, regardless of how much volume traded during it. Simple, predictable.
• VWAP weights price by volume. Periods with heavy trading pull the average more than quiet ones. Tracks more closely to where real market activity happened.

Institutions benchmark against VWAP. Bought below it, you beat the market average. Bought above it, you overpaid relative to where volume was concentrated. TWAP is easier to execute. VWAP is more useful as a performance check.

Where TWAP Can Fail

Not a perfect tool. A few things work against it.

Predictability is a liability. Fixed intervals mean bots can detect the pattern and front-run each execution. Some implementations now randomize the timing specifically to counter this.

It doesn't adapt to the market. Price trending hard against you during the window? The algorithm keeps executing anyway. Buying into a falling market on a timer.

Built for normal conditions. A two-hour execution window in high volatility can span enormous price swings. TWAP assumes some stability during the window. It doesn't always get it.