One of the main issues of the basic VWAP algorithm is its exposure to deviations in trading volume profiles from historical averages. Percentage of volume, or POV algorithms, as the name suggests, try to participate in the market with a specific percentage of the market volume. In other words, they are following the actual market volume so that deviations from historical averages don't matter. Once again, the first step is to determine the total order size in the second step, the POV percentage is set. That means the trader decides on the portion of the market volume they wish to capture. The algorithm then constantly monitors the market volume by analyzing reported trades using this set percentage and observed volume. The algorithm calculates the order size for each interval and executes accordingly, and then it's rinse and repeats. The processes continue adjusting to the ebb and flow of the market's trading volumes until the whole order is completed or a specific end time. For example, the end of the day is reached, whichever happens first. Let's have a look at a concrete example to reduce complexity. We are not looking at a whole trading day, but just a two hour window that we have split into 15 minute time intervals. Let's assume we're looking to buy 10,000 XYZ shares with a 5% participation rate based on the observed market volume, target execution volumes, as well as the aggregate position, the POV algorithm looks as follows. Assuming the POV is able to execute in line with its target values, the order completes at 10:30 AM as the 10,000 units have been purchased. Note that when using the POV algorithm, there is no certainty about completion of the order total in a specific time period, as this will depend on the actual total market volume. Another important consideration in using POV algos is that if several large orders are worked by similar POV algorithms, at the same time, this competition for liquidity could lead to significant market impact.