In this section, we describe a general methodology for constructing forecasting models for market impact. These types of models are very useful in predicting the resulting trading costs of specific trading strategies and in devising optimal trading approaches.
Explicit transaction costs are relatively straightforward to estimate and forecast. Therefore, our focus in this section is to develop a methodology for the implicit transaction costs, and more specifically, market impact costs. The methodology is a linear factor–based approach where market impact is the dependent variable. We distinguish between trade-based and asset-based independent variables or forecasting factors.
Some examples of trade-based factors include:
Probably the most important market impact forecasting variables are based on absolute or relative trade size. Absolute trade size is often measured in terms of the number of shares traded, or the dollar value of the trade. Relative trade size, on the other hand, can be calculated as number of shares traded divided by average daily volume, or number of shares traded divided by the total number of shares outstanding. Note that the former can be seen ...