Event Study Methodology for Capturing Abnormal Trading

Besides return event studies, there are also event studies investigating whether the trading volumes of assets display statistically significant anomalies. These volume event studies apply the general principles of the event study methodology to time series of trading volumes. One major application of volume event studies is the investigation of insider trading preceding M&A announcements. In the US, volume event studies are increasingly used as evidence in security fraw litigation cases.

Measures of Abnormal Trading

The main difference of abnormal volume event study from abnormal return event study is that instead of returns, the log-transformed relative volume per firm is used (Campell and Wasley, 1996), namely

\begin{equation}V_{it} = \log{\left(\frac{n_{it} + .000255}{S_{it}} \cdot 100\right)},\label{eq:log}\end{equation}
where $n_{it}$ is the number of shares traded for firm $i$ on day $t$ and $S_{it}$ is the outstanding share of firm $i$ on the trading day. Ajinkya and lain (1989) and Cready and Ramanan (1991) recommend strongly to use the log-transformed value instead of the non-log-transformed formula:
\begin{equation}V_{it} = \frac{n_{it}}{S_{it}} \cdot 100.\end{equation}
The constant $.000255$ in Equation $\ref{eq:log}$ is added for avoiding the log-transformation on zero values.
Campbell, CynthiaJ. and Wasley, CharlesE. 'Measuring abnormal daily trading volume for samples of NYSE/ASE and NASDAQ securities using parametric and nonparametric test statistics'. Review of Quantitative Finance and Accounting, 6.3 (1996): 309-326