A Deterministic Pricing Model Approach to Merger Analysis

Abstract

This paper extends the merger pricing model associated with Larson‐Gonedes to the general question: how well does the premium developed from the pricing model forecast the securities market reaction of the actual merger? Based on a sample of 91 common stock mergers, shareholders in participating firms incur wealth losses about half the time but the magnitude of the gains outweighs the losses such that statistically significant gains are reported for both buyers and sellers. Removal of market wide price movements further increases the gains to shareholders. However, the premium consistently overstates the gain obtained by acquired firms and bears no systematic relationship to the gains registered by shareholders of acquiring firms. Financial analyses of mergers have focused almost exclusively on mergers as “events” with resultant measurements in abnormal returns surrounding the merger announcement/consummation to shareholders, and occasionally bondholders, in both buying and selling firms. Recent reviews of these studies by Halpern (1983), Jensen and Ruback (1983), and especially Roll (1986) stress the tentativeness of the findings and the ambiguity of their interpretation. The common feature of all this analysis has been on the ex post valuation of the merger event by the securities market from an informational content perspective. Alternatively, these studies have evaluated indirectly whether the price premium paid in an acquisition exceeds, equals, or is less than the market's valuation of the net present value of the merger, and how the spoils/losses are distributed between acquirers and acquirees. But never is the bid premium itself determined and then compared to the market's reaction upon public announcement. As Roll argues, the merger process involves three steps: “First, the bidding firm identifies a potential target firm; second, a ‘valuation’ of the equity of the target is undertaken…; third, the ‘value’ is compared to current market price… If value exceeds price, a bid is made…” Roil (1986, p. 198). This paper links the price premium offered in mergers to the market's reaction to the news of the merger, or alternatively, it compares Roll's steps two and three. The merger pricing model used is the exchange ratio determination model developed by Larson and Gonedes (1969) and applied to mergers by Conn and Nielsen (1977). The pricing model, commonly cited in finance texts (eg. Copeland and Weston (1988, pp. 757–763), has the advantage of being deterministic and thus provides a direct measure of the bid premium subject to a pareto optimal wealth constraint for shareholders in both buying and selling firms. The principal question this paper asks is: Does the price premium provide a consistent, unbiased forecast of the market's reaction? This is an important question from both the bidding firms' and target firms' perspectives for several reasons. First, the terms of the negotiated merger may signal important information to the securities market regarding the degree of agency costs in the merging firms. For example, an excessively high negotiated price for the target may indicate either the bidder has inept management or management insulated from shareholder interests. Thus, the terms of a merger may reflect not only the participants' expectations regarding the merger itself, but also be influenced by existing — although previously unknown — agency costs. The signalling information contained in merger announcement may obviously mask the expectational information, creating ambiguity in interpretation of market reaction. Second, distribution of the market reaction for buyers and sellers is important not only to participating firms' shareholders, but also to the effectiveness of the market for corporate control. A perfectly competitive merger market assures that merger premiums equal the expected value of the increased market values of merging firms. Thus, divergences between premiums and subsequent market reactions may have important implications for assessing the degree of competitiveness in the merger market, and hence, the effectiveness of mergers as a disciplinary force in the market for corporate control. Finally, the adequacy of ex ante merger pricing models remains an unexplored issue. Using an improved methodology, the Larson and Gonedes (LG) model is expanded to adjust for market wide movements in PE ratios; thus, merger specific influences on wealth positions are more clearly focused upon in contrast to the earlier work by Conn and Nielsen (1977). The earlier finding by Conn and Nielsen that approximately one half of mergers sampled in the 1960s failed to meet the pareto wealth constraint for participating firms is therefore re‐examined with an improved methodology and more recent sample of mergers occurring through 1979. The paper is organised as follows. Section I reviews and critiques the Larson‐Gonedes merger pricing model. Section II describes the empirical methodology and sample. Section III presents the empirical results and Section IV concludes with a summary.