That is, while there is a positive relation between the sequence of equity issue and the amount raised, there is no significant difference in offer size relative to firm size between the initial two equity issues and subsequent equity issues.
The implicit assumption behind the methodology of averaging returns is that all equity issue announcements are independent observations, and that for a firm that conducts multiple issues, investors do not react any differently to the announcements of the first few offerings than to those announced later in the sequence.
Second, we investigate whether investor reactions to the current equity issue are influenced by the firm's operating and stock performance following the previous issue.
There is no evidence that the market's reaction to equity issue announcements is influenced by firms' past performance or by expectations of improvements in their future performance.
We obtain information on type of equity issue, number of shares issued, issue price, and filing and issue dates from the Securities Data Corporation's (SDC) Global Financing Database.
The main method of analysis we use is to treat the cumulated abnormal returns for both domestic and foreign equity issue announcements as a dependent variable in an OLS regression, to be explained by a set of dummy variables and a measure of the size of the equity issue.
Figure 2 shows the cumulated average abnormal return for the sample of foreign equity issue announcements.
The price-reactions to domestic and foreign equity issue announcements discussed here are averages for companies and issues that differ in several respects.
In model (1) we include only telsize and find that an equity issue of equal size of the pre-announcement market value (relsize = 1), will decrease the market value by 15,5%.
In model (9) we include all three interactive dummies; the effect of international listing and - equity issue now increases to +11,4% and becomes highly significant.
In the test, the first day on which the announcement of the equity issue appeared in the Wall Street Journal is numbered event day t = 0.
i^ is the price of a share of firm i's common stock the month before the equity issue announcement.
To measure the abnormal earnings forecast revision at the announcement of an equity issue, we use a simple expectations model for expected forecast revision whose assumptions reflect the following salient time-series features of the earnings forecast data: (i) forecast revisions are subject to a bias that varies across firms and is not related in any systematic manner to the month during which a fiscal year or quarter ends, and (ii) the mean monthly percent of analysts revising their earnings forecasts is approximately 20% and again does not seem to be influenced by months during which a fiscal year or quarter ends.
In the next section, we study analysts' forecast revisions of earnings around equity issue announcements to disentangle these plausible explanations for the significant positive relation between TDAR and institutional ownership.
A similar result was documented earlier by both Brous |2^ and Jain |12^, who interpret it to indicate that an equity issue announcement does convey information regarding the firms' earnings.