Excess cash flows and diversification discount

Financial Management (Financial Management Association), Summer, 2004 by John A. Doukas, Ozgur B. Kan

IV. Additional Tests: Excess Cash Flows and Excess Value of Targets

It is possible that the discount documented thus far may be because bidders acquire targets trading at a discount. (8) Therefore, we analyze the pre- and post-acquisition excess cash flows and valuation of bidders while controlling for the excess cash flow and value of targets in year -1. We concentrate only on whole acquisitions in which bidders acquire identifiable target firms, and for which Compustat has information on these targets. We eliminate all firm-year acquisitions in which the bidders acquire units, assets, plants or partial targets because of difficulties in identifying the performance of these targets.

An advantage of the whole-acquisitions subsample is that it allows us to control for possible size effects, because we expect these targets to be comparable in size to bidders. Our selection procedure produced a subsample of 43 identifiable bidder-target pairs over our sample period. Appendix A reports the bidders and targets included in this subsample by year.

Table V shows that the 43 identifiable bidder-target pairs and the remaining 699 firm-year observations of our complete sample have different size characteristics. As expected, the size of fully identifiable targets is much greater, with a mean [median] value of $1.288 billion [$175 million], than that of the remaining acquisitions in which the bidders acquire units, assets, plants, or partial targets, with a mean [median] value of $270 million [$50 million]. The table also shows that the relative size of targets as a percentage of the bidders' total assets is considerably higher in the fully identifiable targets, with a mean [median] of 34.27% [15.09%], than in the rest of acquisitions with a mean [median] of 20.22% [8.91%]. These sample characteristics indicate that the size of targets in our subsample of 43 identifiable bidder-target pairs is more comparable to the assets of bidders than to the assets of the remaining 699 asset or partial acquisitions.

Therefore, we re-estimate the regressions of Table IV. We introduce an indicator variable for the subsample of 43 identifiable bidder-target pairs, its interactive terms with the change in excess cash flows ([DELTA]ECF), and the relative size of targets as additional explanatory variables. The coefficients of the indicator variable and its interactive terms suggest that our previous results remain robust. (9)

A. Bidders' and Targets' Pre-Acquisition Excess Performance

Table VI reports the pre-acquisition excess cash flows and value measures for both bidders and targets in our subsample of identifiable bidder-target pairs. In a careful inspection of the subsample of 43 pair observations we detected two outliers, both associated with related acquisitions (Bidder: Summa Industries (Core Business SIC 30) vs Target: Calnetica Corp (SIC 30) and Bidder:Bell Industries Inc. (Core Business SIC 50) vs Target: Milgray Electronics Inc (SIC 50)) with projected changes in excess value greater than 35%. All other target-bidder pairs had a projected change in excess value within the -15% to 15% range. Therefore, we decided to drop these two outliers from the original subsample. When we replicate the tests with the outliers, the results are slightly different. (10) Consistent with the evidence reported in Table II for the original sample, Panel A shows that bidders have positive and significant excess operating cash flows. Targets also appear to generate mean [median] cash flows greater than those of their industry peers. This evidence suggests that neither bidders nor targets have inferior excess cash flows prior to the acquisition.