This methodology will misclassify some volume. Our goal is not to classify correctly each individual trade (a hopeless exercise in any case), but rather to develop an indicator of overall trade imbalance that is useful for creating a measure of toxicity. We use time bars to allow time for the market price to adjust to the trade direction information that we attempt to recover through bulk classification. Later in the article we present evidence that our bulk classification procedure leads to more useful results for the purposes of estimating flow toxicity than those based on the itemized classification of raw transaction data.16