Kraken Pro adds orders that trade over a chosen time window

One order, executed in smaller pieces

Kraken Pro launched native TWAP orders on 7 September 2026 across web, Desktop, mobile and its APIs. TWAP stands for time-weighted average price. Users select a total quantity and a window from one minute to thirty days. The system sends smaller orders during that period, subject to execution controls and the chosen price boundary. Kraken explicitly says the feature guarantees neither the best price nor completion of the entire order.

What the order controls actually do

Kraken’s Bot Orders guide describes market and limit modes. Market mode uses available prices when each slice is placed. Limit mode sets a maximum buying price or minimum selling price; a slice can be skipped when the market is outside that boundary. The guide says slice sizes are randomized, so an equal-slice illustration is not a reproduction of Kraken’s algorithm.

Running orders appear in the Bots tab with quantities, progress and average fill price. Canceling stops the remaining work; completed trades remain in the account. The guide also notes a gradual rollout. These are useful operational details alongside the dated launch, rather than a promise that every account has identical access.

A worked example of timing, not a price forecast

Suppose a fictional buyer wants 120 units over six hours. An intentionally simplified teaching model divides that into six equal groups of twenty units. The first group is attempted near the beginning and the remaining groups are distributed through the window. The article’s interactive illustration uses this kind of equal division solely to make quantity and time visible. Actual Kraken slicing should not be inferred from it.

Imagine three completed groups of twenty units at invented prices of $10, $11 and $9. Their costs are $200, $220 and $180. The buyer has spent $600 for sixty units, producing an average of $10 per unit before fees. That average describes completed trades only. It does not yet describe the price of all 120 intended units because half the quantity remains unfilled.

Why a price limit can leave an unfinished order

Continue the fictional example with a maximum acceptable buying price of $10.50. A proposed purchase at $11 conflicts with that boundary. If it is omitted in our simplified model, the completed groups at $10 and $9 cost $380 for forty units, an average of $9.50. The average looks better, but the buyer owns fewer units. Comparing only the average would hide the trade-off.

Now imagine the market stays above the limit for the rest of the intended window. The original goal of purchasing 120 units remains incomplete. This is not automatically a malfunction: the price boundary and the quantity target can become incompatible. The example explains why an execution report needs both the filled amount and the average price. One attractive number cannot stand in for the other.

How to judge the completed run

A clear review starts with the instruction the buyer actually gave: direction, total amount, duration and any price boundary. It then compares those inputs with what happened. In the fictional case, forty units bought within the boundary is a different outcome from 120 units bought at a higher average. Neither outcome can be called better without knowing which constraint mattered to the buyer.

Time also introduces uncertainty. A longer schedule leaves more of the intended trade exposed to later market conditions. A shorter schedule concentrates execution into fewer moments. The purpose of the tool is to carry out a chosen schedule, not to establish that one schedule will outperform another. Understanding that distinction makes TWAP easier to use and makes the final report easier to read: it is an account of executed orders, not a verdict on whether the original investment idea was correct.

The same reasoning applies to a seller. Selling fewer units at a preferred price may leave an unwanted holding, while selling the full amount may require accepting different prices. Completion and price discipline remain separate measurements in either direction.