In this lecture, we are going to look at M&A from the perspective of financial economists rather than investment bankers. We have been heavily focused on getting the right price for the target firm. We learned how to insert our beliefs and assumptions about a target firm into our DCF valuation models. Additionally, we learned how to look at historical transactions to make sure that our offers or acceptance prices are reasonable or justifiable. However, as economists, what we care about now is whether mergers and acquisitions actually created value for shareholders.
How do determine whether a merger created value? The most ideal way is to study firm performance after a merger took place. Perhaps we can look at a firm’s profitability over time and see if profits grow faster, products increase quality, or new products are made. The problem with this method is that a lot of things can occur in the aftermath of a merger. For example, the economy could go into a recession, the firm can pursue additional acquisitions, the firm could be acquired, etc. Thus, it is difficult to accurately attribute changes in profitability to one particular transaction the longer we extend the window of our period of study.
To address this first problem, the simple solution is to restrict our study of a firm to a short window. But how short? It could be years after a merger is initiated before any synergies are actually realized or acted upon. Moreover, every firms timeline is different. So, we should conclude that looking at subsequent corporate decisions after a merger is not the most consistent way to analyze mergers. The method of analysis proposed here requires us to assume that markets are efficient and thus looking at the evolution of stock returns after a merger announcement can inform us about the value created for shareholders.
Hopefully, as finance majors, we should all believe to a certain extent that markets are good at aggregating information about a firm’s true or fundamental value. If you have new information about a firm and you acted upon it, either by buying the firm’s stock when it is undervalued or selling the firm’s stock when it is overvalued, the stock price should eventually reflect that information. This is called the efficient market hypothesis.
If we assume that markets are efficient, all information about the consequences of a merger should be captured eventually by the market price. So long as we restrict our window small enough to avoid any other major events that may confound the merger announcement, we can estimate the value added or destroyed by a merger -- or at least the market’s take on it. This is called the event study method.
Let’s formally go over the steps to conduct an event study:
- Remove changes in returns that is explainable by systematic risk: In this course, we will just use a 1 factor model or the CAPM to explain the tradeoff between risk and return. In this fictional world, all returns are explained by 1 source of risk: the market portfolio. Beta is therefore how correlated an asset return is to the overall market return. The hats on top of these variables indicate estimated parameters from running a regression. t indicates time, i indicates a firm. R_it is therefore firm i’s returns at time t. R_mt is the market return at time t. For this course, just take betahat_i as a given and assume alpha_hat is zero. Alpha is interpreted as returns not explained by the market. If CAPM is true, alpha should be not be statistically different from zero.
- In the second step, we calculate the “abnormal returns” of a firm’s stock as equal to the actual stock return minus the returns that are explained by the market risk. Abnormal returns therefore should capture new information specific to that particular firm.
- Since information can drift into the markets slowly, we add up the abnormal returns in a window around the merger announcement. Tau_1 is the starting date and tau_2 is the ending date. Before the announcement, the CAR should be around 0. After the merger announcement, the merger could have positive, negative, or 0 impact. If the merger created shareholder value, CAR should be positive after the announcement. If the merger destroyed shareholder value, CAR should be negative. If the merger didn’t do anything, CAR should be around zero.
We can do this event study for the universe of merger transactions and to average the CARs to see if they are positive, negative, or around zero.
Maytag acquires Magic Chef (1986)
In today’s lecture, I will go over 6 representative merger examples to basically summarize economists’ attempts at generalizing what are good or bad mergers.
The first example is a friendly product line acquisition. This is not a pure diversification strategy since the product line is relevant to the firm’s core competency. Maytag is a American company that mainly focuses on white goods or kitchen and laundry appliances. Since the 1960s, Maytag has predominantly focused on washers and dryers. However, during the 1970s and 1980s, American appliance manufacturers faced fierce competition from Japanese firms. Thus, this period was marked by rapid consolidation to cut costs. Eventually four heavyweights emerged: General Electric, Whirlpool, White Consolidated, and Maytag. These are few of the many M&A transactions that took place around this time period.
We will focus on Maytag’s acquisition of Magic Chef in 1986. Magic Chef actually started out making stoves but later acquired Admiral in 1979. Admiral started out making televisions in the 1950s but later expanded into the refrigerators. The obvious synergies from Maytag’s acquisition of Magic Chef is economies of scope and combined distribution network of similar complementary goods. Maytag was willing to pay a 23 percent premium for Magic Chef. The market responded positively to the announcement: Maytag shareholders earned 13 percent abnormal returns around the acquisition announcement.
Hoberg and Phillips (2008)
In terms of friendly acquisitions, Maytag’s acquisition of Magic Chef is a prototypical example of the kind of mergers that creates the most shareholder value, according to Hoberg and Philips paper. What Hoberg and Philips do is create a distance measure between each pair of firms by how closely related their products are based on textual analysis. Using this measure, they try to determine whether product similarity is an important determinant of post-merger success. To be more precise, Hoberg and Philips claim that the following characteristics of the target is critical for a successful merger:
- Target has products similar to yours. This is necessary to create new synergies
- Target has product dissimilar to nearest rival
- Target has technologies that add barriers to entry such as patents or brand loyalty
Hoberg and Philips find that merger transactions that follow this recipe experience higher sales growth, profitability, and likelihood of new product generation.
Google and Motorola 2011 Merger: Recent Example
A recent example of the Hoberg and Philips findings is the acquisition of Motorola by Google. In 2011, Google wanted to break into the mobile phone business. Their core competency at the time was software, but lacked the hardware expertise and technology to compete head on. Let me briefly discuss the role that each of these companies played with respect to Google:
- RIM or Research in Motion, which was renamed Blackberry Limited, is a Canadian company that created the Blackberry cell phone. Because they made cell phones using their own mobile OS, they were considered competitors. It was unlikely that Blackberry would be willing to sell its hardware IP only and abandon or spinoff its own in-house mobile OS.
- Apple is similar to RIM. They both have proprietary hardware and software technologies. Apple’s iOS is intimately integrated into its hardware, so forming any sort of smartphone partnership is extremely unlikely.
- HP interestingly also used to compete in the mobile OS market. HP used to own webOS, which is an operating system developed by Palm some time ago. WebOS was later purchased by LG that used the technology for smart TVs, and then again by Qualcomm. In the smartphone industry, HP is considered a competitor to Google because their primary product is software.
- Cisco system, who are best known for their routers, also entered the smart TV product space in 2011 through their set top boxes. Similar to Apple and Blackberry, Cisco owned both strong hardware and software patents and IP. Any partnerships with Google is unlikely. Thus, they are considered competitors.
- Nokia, similar to Blackberry, Apple, and Cisco had strong IP in both hardware software. Their partnership with Microsoft makes them a competitor to Google.
Thus, the only targets should consider based on the recipe proposed by Hoberg and Philips are those that had strong hardware IP and weaker software IP. Google thus had many partnerships with firms with strong hardware IP, such as LG, HTC, and Samsung. Although technically they are also Google’s own competitors, since Google wanted to build phones to compete directly against firms it had product partnerships. The seemingly best candidate is thus Motorola, which had strong hardware IP in both smartphones and set top boxes.
General Electric acquires Roper (1988)
But what happens in the case in which an acquirer makes a competitive bid for a target’s product line? Do we still realize the same abnormal returns? Unfortunately no for the case in which General Electric bought out Roper against Whirlpool. Roper, like Whirlpool, made household appliances such as washing machines. GE, for the most part, just wanted to prevent Whirlpool from gaining a significant market share in home appliances. However, GE paid a very high price. GE paid a 107% premium. GE shareholders experienced abnormal returns of negative 3 to 4 percent after the announcement of a competitive bid.