Skip to content
GetProfitable
Search
Dictionary

TWAP algorithm

Slicing an order evenly over a time window so the average fill approximates the time-weighted average price. Simple, predictable, and easy to detect.

Buying 60,000 shares over three hours means roughly 333 shares a minute, ideally with randomised sizes and intervals so the pattern is not obvious. Its virtue is that it makes no assumptions about the volume profile.

Its weakness is the same thing: it ignores volume, so it trades the same amount during a thin lunch hour as during the active open, which means paying more impact per share at exactly the wrong time. vwap-algorithm addresses this.

Predictability is a real cost. A perfectly regular schedule can be inferred by other participants, who then trade ahead of the remaining slices. Randomising size and timing, and varying venues, is not paranoia, it is part of the design.

Related: vwap-algorithm, execution-algorithm, pov-algorithm, market-impact

See it drawn

Original diagrams for the ideas on this page. Illustrative, not real market data.

Intraday price swinging around VWAPA price line for one trading day weaving above and below a smoother VWAP line, with a band drawn one standard deviation either side of it.INTRADAY PRICE AND VWAPprice9:3012:4516:00+1 SD bandVWAP−1 SD bandIllustrative session. VWAP starts fresh at the open and firms up as the day fills in.
VWAP and its standard-deviation bands. VWAP is the day's average price weighted by how much volume traded at each price, so it shows where the bulk of the day's business was done. The bands sit one standard deviation either side, and price here swings between them all session.

Educational only, not advice. Spotted an error? Post in Site Feedback.