Horizontal Mergers, Investments, and Industry Evolution

During the past two decades, U.S. industries have been dramatically shaped by horizontal mergers. The deal value of an average horizontal merger has risen from $48.4 million in 1995 to $504 million in 2016 (See Figure 1). Over the years, academic studies have addressed the issue of whether horizontal mergers reduce industry competition. It is widely acknowledged that the answer relies on the importance of the operating efficiency effect, under which the merger creates synergies which increase the non-merging rivals’ competitive pressure, relative to the market power effect, which allows the rivals to collude. However, it remains unclear how the non-merging rivals react to the merger-induced competitive pressure.

Figure 1 Trend of horizontal mergers

This figure displays the historical trend of horizontal mergers between U.S. firms in the United States from 1984 to 2016. The left axis shows the frequency of horizontal mergers in each year, and the right axis shows the median deal size in real US dollar value relative to that in 1984.

In my recent paper, I examine whether an average horizontal merger increases or decreases non-merging rivals’ competitive pressure. I use a sample of 16,873 horizontal mergers between U.S. firms from 1984 to 2016. The sample entails both public and private deals from 548 industries. 

I find that non-merging rivals have significantly negative market reactions to the announcements of a horizontal merger in the industry (See Figure 2). In the three years post-merger, the rivals have significantly lower profitability and valuation than in the pre-merger years. Based on the stylized assumption that horizontal mergers overall increase the non-merging rivals’ competitive pressure, this paper investigates how the rivals adjust investment policies in response.

Figure 2 Cumulative abnormal returns (CAR) around merger announcements

This figure shows the cumulative abnormal returns (CAR) in a 30-day period around the date of horizontal merger announcements of the merging firms and the non-merging competitors, defined as CRSP firms sharing the same 6-digit NAICS code with the acquirer (target). The gray dashed line draws the acquirer’s CAR movement. The gray solid line draw’s the target’s CAR movement. The dark solid line draws the deal-level equal-weighted average of rivals’ CAR movement. The dark dashed line draws the deal-level value-weighted average of rivals’ CAR movement. The daily abnormal returns are winsorized at the 1% and the 99% percentile. Below the graph are summary statistics of 7-day announcement returns (CAR [-3, +3]) of the merging firms and their rivals.

The rivals’ performance deterioration indicates that horizontal mergers are overall dominated by the operating efficiency effect. That is, the newly merged firm enjoys lower costs of production than the firms operating separately. In response, the non-merging rivals have motives to operate more efficiently as well as to match the marginal costs of the merged firm. Otherwise, the merged firm can lower product prices to a level below the rivals’ marginal costs and easily take over the entire market. Following this intuition, I hypothesize that in response to the merger-induced competitive pressure, the non-merging rivals would adjust their production functions to be more cost-efficient. To do so, firms have incentives to allocate investments to segments with the greatest input-output elasticity, allowing the firms to achieve increasing returns-to-scale.

Using the financial information from Compustat-Segment Database, I document how firms reallocate investments across segments following a horizontal merger. The rivals reduce investments in the segments with low growth opportunities (measured with Tobin’s Q) and raise research and development (R&D), property, plant, and equipment (PP&E), and labor investments in the segments with more growth opportunities. The lower-Q segments appear to be in the industries sharing the same NAICS-6 industry code with the merger, and the higher-Q segments are likely to be in the industries under the same 1 digit-NAICS code, but not in the exact same 6 digit-NAICS industry with the merger (i.e., the related industries). Also, the segment-level tests show that after horizontal mergers, the non-merging rivals’ segment R&D expenses become sensitive not only to the segment’s own Tobin’s Q, but also to that of other segments within the firm. The results reveal the rivals’ improved efficiency in internal capital markets.

Aggregating the investment adjustments at the firm level, I document that in the three years following a horizontal merger, the non-merging rivals increase R&D by 2.4% relative to the median. In the meantime, the rivals reduce investments in PP&E by 9.5% and in labor by 7.4% (See Figure 3). Note that there are significant variations in terms of the rivals’ R&D expenditures. For example, the standard deviation of firms’ R&D ratios is almost twice as large as the median R&D. For the firms with greater R&D increases, they have more reductions in the cost of goods sold, Selling, General & Administrative expenses (SG&A) ratios, and larger increases in asset turnover ratios. Thus, following horizontal mergers, the non-merging rivals appear to focus more on R&D investment to improve their operating efficiency, which is consistent with the operating efficiency effect prediction that horizontal mergers may impose greater competitive pressure on rivals.

Figure 3 Rivals’ investments in the three years around a merger

The four figures below respectively plot the coefficient estimates of rivals’ investments in PP&E (top left), employment (top right), and R&D in the three years surrounding a horizontal merger. An event is defined as instances where a horizontal merger is completed. For each firm exposed to the horizontal merger, investment (including net PP&E growth, employment growth, and R&D/Assets) is regressed on a set of seven event-year dummy variables labeled t-3, t-2,…, t+2, t+3, where the convention is that t takes on the value of one when the horizontal merger is completed. Deal × firm fixed effects, year fixed effects, as well as the control variables are the same as the model in Table 3. Standard errors of coefficients are clustered at the industry level. The solid line plots the coefficient estimates of the year dummies, and the dashed line shows the confidence interval at the 10% level.

Arguably, horizontal mergers and the incumbent firms’ investments are subject to endogeneity concerns. Thus, empirical results in this paper should be interpreted with caution in terms of causality. That said, I use my best efforts to create a control group of firms exposed to withdrawn horizontal mergers. If rivals’ investment adjustments do not result from horizontal mergers, both the treated group and the control group should exhibit similar investment patterns around mergers. However, I find that the observed investment changes only occur to the firms exposed to completed horizontal mergers, implying that the effect of horizontal mergers on rivals’ investment adjustments is unlikely to be spurious.

In addition to production costs, I investigate the alternative mechanisms of costs through which rivals attempt to improve operational efficiency. Namely, I assign the sample firms into two groups based on the pre-merger levels of financial constraints, and I find that the unconstrained rivals tend to increase investments in PP&E, labor, and R&D while the constrained rivals tend to reduce all these investments, suggesting that financing costs may be crucial for the rivals’ consideration of improving efficiency. Furthermore, I assign the sample firms into two groups with their pre-merger levels of innovativeness considering that innovation requires a large sum of sunk costs as well as uncertainty. I find that the more innovative rivals tend to actively engage in PP&E, labor, and R&D investment following a horizontal merger, while less innovative rivals significantly decrease all three types of investments. 

Conclusion

My paper illustrates how firms modify investment strategies following horizontal mergers in their industry. I argue that the investment allocations are consistent with the firms’ cost minimization incentives. That being said, in industry organization, heterogeneities exist not only across firms within the same industry, but also across different industries. For instance, the incentives to invest in R&D depends on firms’ expected loss of profits in the face of competition. In this paper, I find that neck-and-neck rivals significantly increase investments in PP&E, labor, and R&D in the three years following a horizontal merger, while the laggard rivals tend to reduce all the investments. This result implies that horizontal mergers may introduce the most expected loss of market share for neck-and-neck competitors, and these firms would have the greatest incentives to actively keep up with the merging firms.

Since the net effect of horizontal mergers on industry competition depends on the relative strength of the operating efficiency effect to the market power effect, the outcome can be determined by the ex ante level of industry competitiveness. In less competitive industries—e.g., oligopolies—a horizontal merger may facilitate collusion among the remaining incumbents rather than encourage them to engage in efficiency-improving competition. I assign the sample deals into deciles by the pre-merger industry-level profit margins. I find that in the industry decile with the highest profit margins, both the merging firms and the leader rivals have increased profit margins and return on assets following horizontal mergers. Also, the rivals tend to reduce investments in PP&E, labor, and R&D, implying that in the least competitive industries, the horizontal mergers’ market power effect dominates, so the rivals would have fewer incentives to invest.

Dongxu Li is an Assistant Professor in Finance at WangYanan Institute for Studies in Economics & School of Economics, Xiamen University.

This post has been adapted from his paper, “Horizontal Mergers, Investments and Industry Evolution” available at SSRN.

Leave a Reply

Leave a Reply

Your email address will not be published. Required fields are marked *