The effects of mergers and acquisitions on the information production of financial markets

G Model QUAECO-982; No. of Pages 9

ARTICLE IN PRESS The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

The Quarterly Review of Economics and Finance journal homepage: www.elsevier.com/locate/qref

The effects of mergers and acquisitions on the information production of financial markets Marco Bade ∗,1 Berlin Institute of Technology, Chair of Finance and Investment, Sec. H 64, Straße des 17, Juni 135, 10623 Berlin, Germany

a r t i c l e

i n f o

Article history: Received 4 March 2016 Received in revised form 20 September 2016 Accepted 26 September 2016 Available online xxx JEL classification: G1 G11 G14 G18 G34

a b s t r a c t This paper shows that mergers and acquisitions (M&As) create opposing effects on the information production of financial markets. A merger between two related firms may generate technological synergy and profitability gains. This results in greater expected trading profits of speculators and incentivizes them to produce private information feeding back into investments. However, when merging firms announce the M&A deal, they typically disclose internal information. This levels the playing field among traders and eliminates speculators’ incentive to produce information. The resulting tradeoff determines the equilibrium information production of financial markets. © 2016 Board of Trustees of the University of Illinois. Published by Elsevier Inc. All rights reserved.

Keywords: M&A Information production Announcement Market efficiency Real efficiency

1. Introduction Over the past decades, more and more firms have decided to disclose synergy estimates within the announcement of merger and acquisition (M&A) deals. In fact, in the period from 1995 to 2008, the fraction of such firms has increased by 20 percentage points, from 7% to 27% (Dutordoir, Roosenboom, & Vasconcelos, 2014). When a major merger is announced, outsiders aim to interpret the disclosed information in order to estimate the stock’s future performance. One of the most cited examples is the merger between Hewlett-Packard (HP) and Compaq. On the day of the announcement, the stock price of HP fell by almost 19%. One possible explanation for this is that the announcement and disclosure of information related to the merger simply provided new information that led investors to sell the stock because of pessimistic expectations. However, there could be another explanation. That is, as investors and security holders obtain free access to such information, the announcement may have caused a reduction in informational asymmetries among traders. Speculators, brokers and market professionals lost their informational advantage over uninformed investors and sold the stock for lack of potential trading profits.

Moreover, in anticipation of a more level playing field, speculators had less incentive to produce costly, private information to trade on.1 To the best of my knowledge, this effect has not been studied in the context of M&As. Nonetheless, it is an important field of investigation because the information production of financial markets feeds back into firms and influences real decisions.2 As empirically shown by Kau, Linck, and Rubin (2008) and Luo (2005), managers of merging firms learn from stock price movements and may cancel M&A deals, when cancelling the deal is easy, when the market provides new and valuable information, or when the market predicts low returns from the deal. This implies that the market responds to M&As. However, up to date, there is no theoretical study analyzing the preliminary stage of the aforementioned feedback effects. That is, the reaction of the market to the deal and its announcement with respect to information production. This paper aims to fill this gap by presenting a model to study the effects of M&As and their announcement on the information production of financial markets. For this purpose, imagine a situation in which two related firms decide to merge in order to create technological synergy and improve efficiency. Typically, positive synergies emerge when the merging firms are related (Alhenawi & Krishnaswami,

∗ Fax: +49 3031421125. E-mail address: [email protected] 1 I thank the editor Narjess Boubakri and the anonymous referee as well as Prof. Dr. Hans Hirth for constructive comments that significantly improved the paper.

1 Gao and Liang (2013), and Han et al. (2014) show that direct disclosure of firm internal information about growth prospects reduces speculators’ incentive acquire private information. 2 There is empirical evidence on the information feedback effect from prices to real decisions (Chen, Goldstein, & Jiang, 2007; Foucault and Fresard, 2014; Ozoguz and Rebello, 2013).

http://dx.doi.org/10.1016/j.qref.2016.09.006 1062-9769/© 2016 Board of Trustees of the University of Illinois. Published by Elsevier Inc. All rights reserved.

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model QUAECO-982; No. of Pages 9

ARTICLE IN PRESS M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

2

2015) which means that, for example, the firms have overlapping businesses or a certain organizational and strategic fit. Hagedorn and Duysters (2002) show that mergers can improve the technological performance of high-tech companies after the merger, if they fit organizationally and strategically. The neoclassical theory suggests the decision to merge with or acquire another firm may be the result of an industry shock (Mitchell & Mulherin, 1996). Jovanovic and Rousseau (2002) state that mergers have to be seen as a response to the reallocation opportunities that enhance the profitability of the firms. The present model adopts the central prediction of the neoclassical theory by assuming that mergers lead to profitability gains. However, as this paper focuses on the information production of financial markets, I do not endogenize the merging firms’ incentives and decisions with respect to the merger or model the industry shock causing the merger. Within the later announcement of the deal, the firms disclose internal information to the market which they may have learned during the M&A process. As mentioned above, direct disclosure levels the playing field among traders and thus eliminates informational asymmetries. Hence, speculators, who have better information with respect to macroeconomic and market issues relevant for the deal as well as investments, may produce less precise information. Interestingly, this model predicts that there is no clear effect. Instead, the analysis highlights a tradeoff between two effects caused by the deal and its announcement. First, when the two firms merge, they generate synergies which enhance net expected cash flows from investment opportunities. This is in accordance with empirical results from the literature. Based on Data from mergers in the U.S. between 1979 and 1984, Healy, Krishna, and Ruback (1992) show that mergers enhance operating cash flow returns from more productive assets in place, especially in the case of highly related firms. Further, the authors point out that the merging firms maintain high levels of both investment and R&D rates.3 More productive investments enable speculators to generate greater expected trading profits because the traded shares are also more valuable. Consequently, there is an incentive to increase private information production. This effect is labeled the “merger effect” which is new to the literature. Second, the “announcement effect” results from the potential disclosure of internal firm information which preempts speculators’ informational advantage. This is consistent with empirical evidence (e.g., Brown & Hillegeist, 2007; Healy, Hutton, & Palepu, 1999; Heflin, Shaw, & Wild, 2005; Leuz & Verrecchia, 2000; Welker, 1995). As a result, there is a more level playing field among traders which reduces the speculator’s incentive to acquire private information (e.g., Gao & Liang, 2013; Han, Tang, & Yang, 2014). The tradeoff between these two effects determines equilibrium information production of speculators. One central assumption of this model is that financial markets are rational. That is, speculators are rational and trade on private information. Moreover, stock prices are the most informative source of information which is an immediate implication of the efficient market hypothesis. In this regard, the present paper is in accordance with Roll’s (1986) hypothesis of rationality. However, unlike to Roll (1986), in this paper, managers are also rational and make rational investment decisions based on information learned from stock prices. Given the hypothesis of rationality of financial markets and decision makers, the question then arises to what extend markets and investment decisions are efficient since, in the economic research, efficient markets are regularly viewed as one desirable goal (Fama & Miller, 1972; O’Hara, 1997). In order to help explain this question in the context of M&As and their announcements, the present paper studies efficiency measures, such as market efficiency and real efficiency. One prediction is that the tradeoff affecting the information production of financial markets translates into an ambiguous effect on market efficiency. With regard to real efficiency, the model identifies another and yet similar tradeoff between profitability gains and an effect on the information production that can be both positive and negative. These findings shed an entirely new light on efficiency effects caused by M&As and their announcement. The remainder of the paper is organized as follows: Section 2 presents the model setup. Section 3 derives equilibrium investments and information production. Furthermore, it analyzes the tradeoff determining equilibrium information production. Section 4 then discusses the effects of M&As on efficiency measures and highlights another tradeoff of the model. Section 5 relaxes the assumption that announcement is mandatory. Section 6 concludes the paper. All proofs are relegated to the Appendix.

2. The model In this Section, I present a model to analyze the effects of M&As on the information production of financial markets. There are four dates. Table 1 illustrates the timeline of events.

3 There is no consensus regarding the findings in the literature on the effects of M&A on technological synergies and R&D success. In fact, unlike Healy et al. (1992), Ravenscraft and Scherer (1987) do not find profitability gains of conglomerate mergers. However, the present model adopts the idea that M&As generate technological synergies since the focus is on related firms. Hence, this paper can sharpen the understanding of the effects of M&As on financial markets in the case of positive synergies.

Table 1 Timeline. Date t = 0

t=1

t=2

t=3

A and B commit to M&A agreement Speculator acquires private information

Closing or cancellation of the deal Firms learn private information Announcement Secondary market opens

Observation of stock price Investment decisions

Realization of cash flows

The economy is divided in two sectors, namely the public sector and the private sector. At date t = 0, the public sector consists of a single peer firm (A) which is publicly traded. Firm A represents an existing incumbent firm. The private sector consists of two nontraded firms (B and C) which are born at date t = 0 and which can be viewed as new market entrants. In the first stage, the three firms have no private information to enhance research and development. Additionally, they share the same stochastic source of uncertainty  which is either high (H) or low (L) with equal probability. For simplification, I assume H > 0 and L = 0. Each firm consists  of an  investment opportunity Gi . Hence, firm value of firm i ∈ A, B, C is given by: Gi = gi Ii −

1 2 I 2 i

(1)

Ii is the volume of investment. gi > 0 measures the profitability of investment. The net cash flow from the investment at date t = 3 determines firm i’s expected value at date t = 0. Furthermore, at date t = 0, thereis a risk-neutral speculator who acquires a private signal y ∈ H, L about the stochastic technology with









+1

Pr y = H| = H = Pr y = L| = L = 2 and  ∈ [0, 1].  measures the speculator’s quality of private information and will be determined endogenously. If  = 1, the speculator learns  perfectly. If  = 0, he learns nothing about . Expending resources on private information quality is costly for the speculator. The cost function is increasing and convex and given by: C () =

c 2  . 2

Now, consider the following scenario. The peer and one of the two private firms (firm B) agree to merge at date t = 0 in order to generate technological synergy because they assume to learn private information later. Think of a high-tech economy with a high level of technological uncertainty. For example, the three firms all produce electric motors or batteries. It is reasonable to presume that firms within such industry may not yet have private information when they make early M&A decisions to create competitive advantages. Consequently, learning during the M&A process may happen. Moreover, none of the firms may assume to learn perfect information about the future technological shock. Thus, in order to improve success in research and development and to gain competitive advantage, there is an incentive to generate learning synergies by merging with another firm. As a result, the two firms A and B combine knowledge and improve internal information production. However, they do not yet announce their plans at date t = 0 because the deal will be closed later at date t = 1. The probability of closing is exogenously given by . Thus, with probability  the deal will be closed and announced at date t = 1 and with probability 1 −  the deal will be cancelled.4

4 One can regard  as either the probability that non-influencable circumstances will lead to the conclusion of the agreement, or as the probability that the deal will be approved by the regulator (e.g., the Department of Justice and Federal Trade

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model

ARTICLE IN PRESS

QUAECO-982; No. of Pages 9

M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

also at date t = 1, firm i privately learns a signal zi ∈  Thereafter, 

, ø at no cost. The signal reveals  perfectly with probability fi ∈ (0, 1) and is completely uninformative with probability 1 − fi . This means, the exogenous parameter fi represents the quality of a firm’s internally available information. In case of closing, immediately after private learning, the merged firm (AB) announces the deal and discloses all information, that is, for example, information regarding advancements in research and development, or simply private information about the technological shock . Depending   on the new peer’s information zAB , the announcement v ∈ , ø reveals  perfectly with probability fAB ≡ fA + fB − fA fB or is uninformative for the market with probability 1 − fAB . In case of cancellation, the peer firm (A) does not disclose any information. This results in four different cases as illustrated in the following table5 : Then, the secondary market opens and firm shares are traded. In case of cancellation, only firm A is traded. Otherwise, if the deal is closed, shares of the newly merged firm AB are traded. The speculator and noise traders interact in a Kyle-type setting. Aggregate demand xn of noise traders is either 1 or −1 with equal probability. When the speculator decides to trade (cases 2–4), he camouflages his trade with the noise trading and buys (xs = 1) one unit of the stock if y = H and sells (xs = −1) one unit if y = L. In case 1, he does not trade as will be show in Section 3.2. It can easily be shown that he does not have an incentive to deviate from this strategy. In combination with the demand of noise traders, the aggregate order flow Xj = xs + xn takes three different values, i.e.,









Xj ∈ −2, 0, 2 where j ∈ A, AB . As in Kyle (1985), the competitive market maker does not learn any private information. Instead, he observes the total order flow that he cannot break down to its components. Hence, the price the market maker sets is assumed to equal the net expected  cash  flow from the traded firm, conditional on Xj , i.e., Pj (X) = E Gj |Xj . At stage t = 2, all firms observe the price Pj of the traded firm





and make investment decisions Ii zi , Pj conditional on their information sets. Since the investment opportunity of the public firm is traded, there is feedback in both directions, that is, from price to investment and from investment to price. The publicly traded firm learns from its own stock price while the private sector also learns from the peer’s price because all firms in the economy share the same source of uncertainty. Cash flows from investments are realized at date t = 3. This timeline captures the aforementioned economic setting in which firms learn both, private information during the M&A process and public information from price after the M&A deal is closed or cancelled.

3. Equilibrium This section solves the model recursively and presents the results with respect to the investment decisions at date t = 2, the secondary market trading game at stage t = 1, and the optimal information acquisition of the speculator at date t = 0.

Commission). As well described by Polasky and Mason (1998, pp. 1f.), increased market concentration resulting from horizontal mergers leads to increased prices and decreased social welfare. This can cause the regulator to cancel proposed mergers. In fact, the number of mergers investigated by the regulator increased dramatically. This makes clear that it is important to consider the possibility of the cancellation of M&A deals by exogenous reasons. 5 This order of events is important to depict the idea that firms learn private information during the M&A process, that is, after agreeing on the deal in order to enhance learning but before the announcement. Otherwise, there would be no private learning during the M&A process. In fact, the results do not change if the firms learn private information before the deal is closed or cancelled by exogenous circumstances as long as the merging firms are assumed to be committed to the agreement.

3

3.1. Equilibrium investment decisions At date t = 2, firm i makes the investment decision conditional on its information set that is composed of two ingredients. First, firm i learns a private signal zi at stage t = 1. Second, the firm observes the stock price PA in case of cancellation of the deal or PAB in case of closing. Given the information set, firm i’s expectation of its own investment opportunity is:









E Gi |zi , Pj = gi E |zi , Pj Ii −

1 2 I . 2 i

Because of noise, firm i is unable to fully deduce complete information about  from Pj . Thus, the firm can only partially presume information about the speculator’s signal from the peer’s stock. The first-order condition determines the optimal investment decision:



dE Gi |zi , Pj dIi







= gi E Gi |zi , Pj − Ii = 0.





Hence, firm i chooses to invest Ii∗ = gi E Gi |zi , Pj . Substituting ∗ Ii into Eq. (1) yields the ex post return:





Gi = gi2 E Gi |zi , Pj −

 1 2 E Gi |zi , Pj . 2

The tradability of the investment opportunity of peer firm (j) creates a feedback loop. In setting the price of the peer Gj , the market maker forecasts the value of  conditional on his information. In addition, he  forecasts  the peer’s investment decision at date t = 2, i.e., Ij∗ = gj E Gj |zj , Pj . Because the peer firm itself observes Pj and conditions the investment on it, the price both reflects and affects the peer firm’s expected value. This feedback loop results in a fixed point problem, which precludes closed-form solution (Gao & Liang, 2013, p. 1411). This paper adopts the approach of Bade and Hirth (2016) by assuming a binary trading structure. Thus, the investment opportunity of the public peer firm A or AB is tradable. 3.2. Equilibrium information production After private information acquisition by the speculator and after the deal is closed or cancelled, the speculator, uninformed investors (noise traders), and a market maker trade shares of the peer in a Kyle-setup. In cases 2–4, based on his information set after receiving a private signal, the speculator buys when he learns y = H or sells one unit of the traded firm’s stock when he learns y = L. Uninformed investors are hit by a liquidity shock. Thus, they are forced to trade in the secondary market. If Xj = 2 or Xj = −2 the market maker infers y = H or y = L, respectively, because the speculator has no  incentive to deviate from his trading strategy. In these cases, Pj Xj reveals the speculator’s information perfectly and the speculator does not benefit from trading. Only in case Xj = 0, the speculator has an informational advantage over the market because the aggregated order flow does not contain any information. The market maker sets the price corresponding to the unconditional expectation of Gj . I now separately discuss the trading games at date t = 1 for the four cases listed in Table 2. All proofs can be found in Appendix A. For case 1, in which the deal is closed and the newly merged firm (AB) has perfect information, the announcement of the deal reveals  perfectly. As a result, all market participants have perfect information. In particular, the speculator loses his informational advantage over the market maker. Moreover, neither firm ABnor  firm C learn from the price. The market maker sets the price PAB  . In case 2, the deal is closed but firm AB has no private information. In this case, the announcement of the deal is uninformative.

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model

ARTICLE IN PRESS

QUAECO-982; No. of Pages 9

M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

4 Table 2 Information structure. Case

Probability

Outcome of the deal

Peer firm j

Peer firm information zj

Announcement v

Price Pj

1

fAB

Closed

AB





PAB 

2 3 4

 (1 − fAB ) (1 − ) fA (1 − ) (1 − fA )

Closed Cancelled Cancelled

AB A A

ø  ø

ø ø ø

PAB (y) PA (y) PA (y)

Thus, in case of merger (superscript m), the speculator benefits from trading and expects to earn the following profit: E(W )(m) =

 1 2 2 H gA + gB2 . 8

(2)

In cases 3 and 4, in which the deal gets cancelled (superscript n), there is no announcement. Thus, the speculator expects to benefit from trading in both cases. His expected profit in both cases without merger is: E(W )(n) =

1 2 2 H gA . 8

(3)

Considering the probabilities of the three cases in which the speculator expects to generate trading profits, the ex ante expected trading profit of the speculator, before information acquisition is given by: E (W ) =

 1 2 2 H gA (1 − fAB ) + gB2  (1 − fAB ) . 8

(4)

Not surprisingly, the speculator’s expected trading profit increases in the level of profitability H of the technology. Furthermore, the investment profitabilities of firms A and B (before merger) respectively the profitability of the newly merged firm AB increase the speculator’s expected trading profit. It is also not surprising that the quality of private information enhances the expected trading profit. However, the internal information production of firm AB represented by fAB has a negative effect. This is intuitive because when the merged firm is likely to have perfect information to announce after the deal is closed, the probability of case 1, in which the speculator does not benefit from trading, is high, too. Put differently, when fAB increases, the probability of scenarios in which the speculator expects to generate trading profits, decreases. The most interesting parameter seems to be the probability  of closing because the impact of  is not immediately visible. Before discussing the effect of , I derive the speculator’s equilibrium information acquisition. Taking into account the cost of private learning, the speculator maximizes his net expected profit, i.e.:



max E (W ) − 

ambiguous effect on  ∗ as well. This can be seen from the derivative of  ∗ with respect to  which is given by:

 d ∗ H2  2 = gB (1 − fAB ) − gA2 fAB . 8c d From this, it is straightforward that fAB >

gB2 gA2

+ gB2

.

d ∗ d

< 0 if and only if: (6)

The speculator reduces private information production if and only if the newly merged firm’s internal information processing is sufficiently precise. Otherwise, when fAB is sufficiently small, which means that firm AB is more likely to be internally uninformed about , the speculator has an incentive to increase private information quality because this enhances expected trading profits. Apparently, there is a tradeoff determining the speculator’s private learning. The tradeoff is as follows: On one hand, when the likelihood of closing increases, the M&A becomes more likely. Since the M&A creates synergy, internal information quality of firm AB becomes more precise. As a result, the publicly traded firm raises more capital to invest which, in turn, increases expected firm value. This means, firm AB becomes more valuable. Thus, when the M&A becomes more likely, the speculator expects to trade a more valuable stock which enhances expected trading profits, i.e., E(W )(m) > E(W )(n) . I label this effect the “merger effect”. On the other hand, since the announcement is mandatory,  also represents the probability of the announcement. In case of closing and perfect information of firm A and/or B, which occurs with probability fAB (case 1 in Table 2), firm AB’s announcement reveals  perfectly and the speculator is indifferent in trading. This expresses itself in an incentive to reduce private information quality  ∗ . This effect is labeled the “announcement effect”. 4. Efficiency effects In this section, I adopt an efficiency point of view to further study the effects of M&As on financial markets. In particular, I examine the effects in terms of both market efficiency and real efficiency.



c 2  . 2

4.1. Market efficiency

Solving the maximization problem yields: ∗ =



 H2  2 gA (1 − fAB ) + gB2  (1 − fAB ) . 8c

(5)

Since profitability parameters H, gA , and gB facilitate the speculator’s opportunity to generate trading profits, he chooses to acquire more precise information as these parameters increase. Recall that fAB has a negative impact on the speculator’s expected trading profit, his optimal information production also decreases in fAB . Thus, as the merger of firm’s A and B creates synergy in learning, i.e., fAB > fA , fB , the information production of the market becomes less precise. I now revive the impact of parameter . As broached above,  does not have an obvious effect on E (W ). Consequently, there is an

“A market in which prices always ‘fully reflect’ available information is called ‘efficient.”’ (Fama, 1970; p. 383). This means, market efficiency measures to what extent prices can be used to inform about the future value of the traded assets. Thus, market efficiency can best be measured by the price informativeness (PI) of a traded stock. In cases 2 and 4 presented in Table 2, the peer firm j does not have private information. Given that firms B (in case of no merger) and C have no private information either, every firm learns from the price of the public firm. Price informareduction effect of tiveness can be expressed as the uncertainty     the traded firm’s share price, i.e., var  − var |Pj . The market is entirely efficient when the asset price completely explains the uncertainty associated with the payoff of the traded asset, which means that there  isno residual uncertainty after observation of the price, i.e., var |Pj = 0. Accordingly, equilibrium price informa-

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model

ARTICLE IN PRESS

QUAECO-982; No. of Pages 9

M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

tiveness when no firm has private information is given by:

 





PI ≡ var  − var |Pj = H2 = 4



H2 ∗ 2 ( ) 4

 H2  2 gA (1 − fAB ) + gB2  (1 − fAB ) 8c

merger and the announcement of the deal enhance real efficiency because of increased information production by the speculator. This is due to the fact that, in this scenario, both the synergy effect and the information production effect are positive.

2

However, when fAB >

.

because, in this case,

The proof is in Appendix B. The speculator’s private information is the only source of information for the firms when they learn from the stock price. Thus, there is a positive effect of private information production of the market on firm investments because price informativeness enhances investment efficiency (e.g., Bade & Hirth, 2016). This is aggravated by the merger effect introduced above. However, recall that the announcement effect causes the speculator to reduce information acquisition, M&As with mandatory announcements can also reduce the informational content of prices which is associated with lower market efficiency. Put together, the tradeoff induced by the probability of closing translates into an ambiguous effect on market efficiency. This result extends current findings in the literature. For instance, Gao and Liang (2013) state that the effect of disclosure on both price informativeness and investment efficiency is negative. In this paper, M&As with disclosure within the announcement cause an ambiguous effect on price informativeness and thus market efficiency as well as investment efficiency. 4.2. Real efficiency Real efficiency can be defined as the ex ante expected aggregate output of the economy (Goldstein, Ozdenoren, & Yuan, 2013). That is, real efficiency is measured by:





RE ≡ E Gaggr. . Thus, considering the probability of the two cases (merger and no merger), real efficiency is given by: RE = E (GAB ) + (1 − ) (E (GA ) + E (GB )) + E (GC ) .

(7)

In order to analyze the effects of M&As on real efficiency, Eq. (7) is differentiated with respect to , i.e.: dRE dE (GAB ) = E (GAB ) − (E (GA ) + E (GB )) +  d d + (1 − )

 dE

dE (GB ) (GA ) + d d



+

dE (GC ) d

E (GAB ) − (E (GA ) + E (GB )) > 0. This positive effect of  on real efficiency is labeled the “synergy effect”. The second effect is expressed by: dE (GAB ) + (1 − ) d

 dE

dE (GB ) (GA ) + d d



+

dE (GC ) , d

which can be both positive and negative, depending on the response of the speculator’s information acquisition. This effect is labeled the “information production effect”. When fAB <

sequently,

dRE d

d ∗ d

g2 B

g 2 +g 2 A

, the overall effect is ambiguous

B

< 0 and thus,

and

dE(GAB ) d

< 0. Con-

> 0 if and only if:

E (GAB ) − (E (GA ) + E (GB )) > − − (1 − )

dE(Gi ) d

 dE

dE (GB ) (GA ) + d d

dE (GAB ) d



−

dE (GC ) . d

In words, given that the internal information production of the merged firm is sufficiently precise, the effect of the merger and the announcement on real efficiency is positive if and only if the positive synergy effect is stronger than the negative information production effect. Considering only the private economy’s real efficiency, given by RE = (1 − ) E (GB ) + E (GC ), the synergy effect turns negative. This is because in case of merger, firm B joins the public economy and leaves the private sector which reduces the overall expected value of the private sector. Furthermore, the information production effect influences the private economy’s real efficiency the same way it affects overall real efficiency. The public economy’s real efficiency is given by: RE = E (GAB ) + (1 − ) E (GA ) , Again, both the synergy effect and the information production effect, affect real efficiency. This is because the public peer firm, which is either merged firm AB or firm A, learns from its own stock price. However, the information production effect is weaker in this scenario because the effect on the private sector is excluded. As a result, the synergy effect gains importance. To conclude, the synergy effect only impacts the public sector whereas the information production effect affects both the public sector learning from its own stock price and the private sector learning from the peer. Moreover, as the impact of mergers and their announcement on the information production of markets depends on exogenous parameters, such as internal information quality and profitability parameters, and thus is ambiguous, so is the effect on real efficiency. 5. No mandatory announcement

The impact of  on real efficiency can be divided into two effects. First, recall that the merger creates learning synergies, i.e., fAB > fA , fB . Thus, as shown in Appendix B,



5

g2 B 2 g +g 2 B A

,

the effect of  on  is positive, which means that the speculator increases private information acquisition because the merger effect dominates the announcement effect. As the net expected cash flow from each investment opportunity positively depends on , there is a positive effect of  on real efficiency. This means, when internal information quality of the newly merged firm is sufficiently low, the

In this section, I relax the assumption that the announcement of M&A deals is mandatory. Instead, I assume that the merging firms decide not to disclose internal information because, from the firms’ perspective, there are only positive effects associated with the information production of the speculator. As a result, case 1 in Table 2 changes. Particularly, with probability fAB , the deal is closed and firms A and B learn  perfectly. Unlike in the baseline model, there is no announcement, i.e., y = ø. Thus, neither uninformed traders nor the market maker learn about . Consequently, the price is a function of the speculator’s private information, i.e., PAB (y). Since the rest of the model remains unchanged, I directly turn to the results of this modification that are different from those in the baseline model. In this modification, the speculator expects to profit from trading in every four cases. His expected trading profit is now given by: E (W ) =

 1 2 2 H gA + gB2  . 8

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model QUAECO-982; No. of Pages 9

ARTICLE IN PRESS M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

6

Unsurprisingly, the speculator’s expected trading profit is independent from the firms’ internal information qualities because the speculator knows that there will be no announcement. Still, E (W ) depends on the probability  of closing. Most importantly, it turns out that the tradeoff disappears. In particular, the announcement effect is muted. In contrast, the merger effect is robust to this modification. The speculator must not consider the scenario of disclosure which would preempt his informational advantage. Thus, the speculator has an incentive to increase private information quality as long as the benefit exceeds the cost of private information acquisition. This simple modification shows that the mandatory announcement of M&A deals harms information production of financial markets. This in turn reduces market efficiency and investment efficiency because of a weakened information feedback effect from price to investment. Thus, from an efficiency point of view, it seems to make more sense to have the announcement optional because this mutes the negative announcement effect.

information acquisition. These findings shed an entirely new light on efficiency effects caused by M&As and their announcement. The paper abstracts from other aspects of M&A studies, such as the impact of timing of the announcement of deals (e.g., Demski & Feltham, 1994; Luo, 2005) or the endogenous cancellation of deals by firms since I assume that the merging firms are committed to the M&A deal and the probability of closing (cancellation) is exogenously given. In fact, as mentioned in the introduction section, in some situations, managers of merging firms respond to stock price movements by cancelling the deal themselves (Kau et al., 2008; Luo, 2005). Thus, looking forward, it would be interesting to endogenize the actual decision on the M&A deal of the merging firms. In addition, as mentioned above, the findings in the literature on the effects of M&As on technological synergies in R&D and profitability gains are quite mixed. Assuming positive synergies and profitability gains is central to the present model. Future research could examine the same research question under the assumption of negative synergies or profitability losses. Appendix A. Equilibrium information acquisition and the tradeoff

6. Conclusion This paper investigates the effects of M&As and their announcement on the information production of financial markets. The model is based on theoretical studies on feedback effects and is consistent with a good deal of empirical findings from M&A literature. Yet, this paper contributes to the literature on both topics as it presents some new findings and predictions. The announcement of closed M&A deals creates opposing effects on the information production of financial markets and thus market efficiency and real efficiency. First, the merger between two firms generates synergy in learning and profitability of investment which leads to greater expected trading profits of speculators. This provides an incentive to produce private information feeding back into real investments of both privately held firms and publicly traded peers. This effect is new to the literature and should be empirically tested. For example, it would be interesting to test the relation between profitability gains of merging firms and insider trading volumes around M&As as the model predicts that this relation might be positive. Second, when the announcement of M&A deals is mandatory, closing the deal is connected with disclosure of firm internal information which levels the playing field among traders and preempts speculators’ informational advantages over uninformed investors. This reduces speculators’ expected trading profits and thus motivates them to produce less precise private information. This so called announcement effect is basically consistent with theoretical (e.g., Gao & Liang, 2013; Han et al., 2014) and empirical evidence (e.g., Brown & Hillegeist, 2007; Healy et al., 1999; Heflin et al., 2005; Leuz & Verrecchia, 2000; Welker, 1995) from research on the effects of disclosure and announcements on information acquisition, informed trading, and market liquidity. The resulting tradeoff between the merger effect and the announcement effect determines the equilibrium information acquisition of financial markets. Furthermore, the paper shows that the aforementioned tradeoff translates into a similar effect on market efficiency. Disclosure within the announcement of M&A deals causes an ambiguous effect on price informativeness and thus market efficiency as well as investment efficiency. With regard to real efficiency, there is a similar tradeoff. First, there is the positive synergy effect resulting from the combination of knowledge which enhances internal information processing of the newly merged firm. This is consistent with the assumption that mergers of related firms cause profitability gains. Second, there is the information production effect that can be both positive and negative, depending on the response of the market’s

I proceed in three steps. First, I prove the speculator’s expected trading profit for cases 3 and 4 in Table 2 (no merger) as given in expression 3. Second, I derive the speculator’s ex ante expected trading profit at date t = 0. Finally, I calculate equilibrium information acquisition. The following calculations are used later. Conditional on his private signal y, the speculator has the following expectations:









 +1 H, 2

E |y = H =

E |y = 0 =

 

E  =



 +1 2

 H,

H , 2









E  2 |y = H =

E  2 |y = 0 =

 

1−

E 2 =

 +1 2 H , 2



 +1 2

1−



H2,

H2 . 2

Subcase: firm A learns  perfectly





Thus, E |zA , PA =  and GA∗ = The speculator expects: E (GA |y = H) =

E (GA |y = 0) =

gA2 2





E  2 |y = H =

gA2 H 2 2



1−

 +1 2

g2 2 A  . 2

gA2 H 2

 + 1

2

2

,

 .

From XA = 2, the market maker infers y = H, and from XA = −2, he infers y = 0. When XA = 0, he does not learn any information. This leads to the following prices: PA (XA = 2) = E (GA |y = H) =

gA2 H 2

 + 1

2

2

,

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model

ARTICLE IN PRESS

QUAECO-982; No. of Pages 9

M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

gA2

PA (XA = 0) = E (GA ) =

8



gA2 H 2

PA (XA = −2) = E (GA |y = 0) =

2

 +1 2

1−



Similar proceeding leads to the speculator’s expected trading profit in the cases of merger (case 2 in Table 2) which is given by:

,

E(W )(m) = H2.

Given the speculator’s private signal is y = H, his expected trading profit is: 1 (E (GA |y = H) − PA (XA = 2)) 2

(n)

E(W )y=H = +

7

 1 2 2 H gA + gB2 . 8

In case 1, the speculator does not benefit from trading. Thus, collecting the speculator’s expected trading profit in various cases and weighting them by the respective probabilities (see Table 2), the speculator’s expected gross profit at date t = 0 is: E (W ) =  (1 − fAB )

1 1 (E (GA |y = H) − PA (XA = 0)) = gA2 H 2 ␥. 2 8

(1 − fA )

 1 1 2 2 H gA + gB2  + (1 − ) fA H 2 gA2  + (1 − ) 8 8

 1 2 2 H2  2 H gA  = ␥ gA (1 − fAB ) + gB2  (1 − fAB ) . 8 8

Given his private signal is y = 0, the speculator expects to earn the following trading profit: 1 (PA (XA = −2) − E (GA |y = 0)) 2

(n)

E(W )y=0 = +

Equilibrium information acquisition The speculator chooses information acquisition  to maximize the net expected profit, i.e.:

1 1 (PA (XA = 0) − E (GA |y = 0)) = gA2 H 2 ␥. 2 8



max E (W ) − 

Subcase: firm A does not learn any private information about 



Thus,















E |zA , PA = E |PA = E |XA





which results in expression 5 in the text, i.e.: GA∗ =

and



gA2 E |XA − 12 E 2 |XA . Given y = H, the speculator expects: 1 1 E (GA |y = H, XA = 2) + E (GA |y = H, XA = 0) . 2 2

E (GA |y = H) =



c 2 ,  2

 H2  2 gA (1 − fAB ) + gB2  (1 − fAB ) − c = 0 8  H2  2 gA (1 − fAB ) + gB2  (1 − fAB ) . 8c

∗ =

I now calculate the two summands separately. The speculator needs to forecast the firm A’s expectation of , i.e.:





 









E |y = H, XA = 2 = E E |XA = 2 |y = H = E |y = H .

Market efficiency

Thus, E (GA |y = H, XA = 2) =

gA2 2





E 2 |y = H =

gA2 2

H2

  + 1 2 2

.

When XA = 0, firm A does not learn from price. Hence,

 





 

and E (GA |y = H, XA = 0) =

gA2 4

H2



1 + 2

 .

Put together, the speculator’s expectation is: E (GA |y = H) = 1 2



gA2 2

H2

1 1 E (GA |y = H, XA = 2) + E (GA |y = H, XA = 0) 2 2

  + 1 2 2

+

1 2



gA2 4



H2  +

1 2



.

Given y = 0, the speculator expects: 1 1 E (GA |y = 0) = E (GA |y = 0, XA = −2) + E (GA |y = 0, XA = 0) . 2 2 Analogous calculations lead to: E (GA |y = 0) =

gA2 16



As discussed in the text, I use price informativeness to measure market efficiency. Price informativeness is defined as the explained portion of uncertainty about , i.e.:

 





PI ≡ var  − var |Pj .

E E |XA = 0 |y = H = E 

=

Appendix B. Efficiency effects



H 2 2 − 4 +  2 .

It is necessary to differentiate between the cases of merger and no merger. However, it is easy to show that price informativeness is the same from every firm’s perspective. Only the cases in which the firm considered has no private information about  matter. Moreover, the price only contains information if Xj = 2 or Xj = −2. I now calculate price informativeness when firms A and B merge for the two cases of XAB . The ex ante uncertainty about  is given by:

 

 

 

var  = E  2 − E 2  =

H2 H2 H2 − = . 2 4 4

In case XAB = 2, ex post uncertainty is:









var |XAB = 2, zi = ø = E  2 |XAB = 2, zi = ø





−E 2 |XAB = 2, zi = ø = =

 +1 2 H − 2

  + 1 2 2

H2

 H2  1 − 2 . 4

Thus, the expected trading profit of the speculator is given by: (n)

E(W )y = +

1 (E (GA |y = H) − PA (XA = 2)) 2

1 1 (E (GA |y = H) − PA (XA = 0)) = H 2 gA2 ␥. 2 8

Thus, price informativeness is straightforward: PI (XAB = 2) =

 H2 2 H2 H2  1 − 2 = −  . 4 4 4

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model

ARTICLE IN PRESS

QUAECO-982; No. of Pages 9

M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

8

It can easily be shown that in case XAB = −2, price informativeness is also given by: H2

PI (XAB = 2) =

4

Completely analogous calculations lead to:

 2.

Therefore, price informativeness in case of merger of firms A 2 and B is PI (m) = H4  2 . Moreover, in case of no merger, price inforH2

mativeness is PI (n) =

In the cases of no merger, only firm A is traded. First, I derive the ex ante expected value of the peer when the peer has no private information which occurs with probability 1 − fA. The aggregate  order flow takes three different values, i.e., XA ∈ −2, 0, 2 . Conditional on XA , each uninformed market participant expects:



 +1 E |XA = 2 = H, 2



E |XA = −2 =

E (GA |XA = 2) = =

2

E

2



dE (GAB )

− H.



  + 1 2

E (GA |XA = −2) =

2

2 1 + fAB + (1 − fAB ) 2

2

2

.

+ (1 − )

 dE

dE (GB ) (GA ) + d d



+



dE (GC ) . d



8

2 1− 2



gA2 + gB2 H 2

2 1 + fA + (1 − fA ) 2



1 + fAB +

8

−

gB2 H 2



8

2 (1 − fAB ) 2

2 1 + fB + (1 − fB ) 2



> 0,

(fAB − fB ) > 0.

This proves the synergy effect. Now, I turn to the second term in square brackets. Given fAB >

2

H ,

g2 B

g 2 +g 2 B

,

d d

< 0, and thus

g2H2 d dE (Gi ) = i <0 (1 − fi ) 8 d d

H ,



d

gA2 H 2

because

gA2

,

I now analyze the two components in square brackets separately and start. First, I find that



2

gA2

8

+ 

A

8





dRE = [E (GAB ) − (E (GA ) + E (GB ))] d



|XA = 2 =

E (GA |XA = 0) =



,

2 (1 − fC ) 2

1 + fC +

gA2 + gB2 H 2



In addition, I prove the tradeoff of real efficiency. For this purpose, I differentiate real efficiency with respect to , i.e.:

  1 |XA = 2 − gA2 E 2 |XA = 2 2



gA2



E (GAB ) − (E (GA ) + E (GB )) =

Hence,



2 (1 − fB ) 2



 +1 1− 2

gA2 E 2

1 + fB +

RE = E (GAB ) + (1 − ) (E (GA ) + E (GB )) + E (GC ) .

H E |XA = 0 = , 2





Then, real efficiency is given by:



gA2

8

E (GAB ) =

Peer has no private information



gC2 H 2

E (GC ) =



In the text, following Goldstein et al. (2013), I define real efficiency as the ex ante expected aggregate output of the economy. In order to determine real efficiency, the ex ante net expected cash flow from each firm needs to be calculated.



8

 2.

4

Real efficiency



gB2 H 2

E (GB ) =

 +1 1− 2

2

and H2.



dE (GAB ) = d



gA2 + gB2 H 2 8

(1 − fAB )

d < 0. d

Then, the following is straightforward: Peer learns  perfectly  When firm A learns  perfectly which occurs with probability, the peer does not learn from the price. Thus, firm A expects:





E GA |zA =  =

gA2 2

H2.

Collecting the two expectations, the ex ante expected cash flow from GA is given by: E (GA ) = fA

+

=

1 4



gA2 H 2 2

gA2



2

gA2 H 2 8



1−

 

+ (1 − fA )  +1 2

2

1 4

H2

+

2 1 + fA + (1 − fA ) 2

gA2

  + 1 2

2

2

1 2

.



gA2 8



H

H2

2

dE (GAB ) + (1 − ) d

 dE

dE (GB ) (GA ) + d d



+

dE (GC ) < 0. d

This proves the information production effect. To conclude, the first term in square brackets (synergy effect) is positive whereas the second term in square brackets (information production effect) is negative. This suggests that there is a tradeoff between these two effects if and only if internal information processing of the peer firm is sufficiently strong. In contrast, when fAB <

g2 B

, the information production effect is reversed. In this

g 2 +g 2 B A scenario, dRE d

> 0.

References Alhenawi, Y., & Krishnaswami, S. (2015). Long-term impact of merger synergies on performance and value. The Quarterly Review of Economics and Finance, 58, 93–118. Bade, M., & Hirth, H. (2016). Liquidity cost vs: Real investment efficiency. Journal of Financial Markets, 28, 70–90.

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006

G Model QUAECO-982; No. of Pages 9

ARTICLE IN PRESS M. Bade / The Quarterly Review of Economics and Finance xxx (2016) xxx–xxx

Brown, S., & Hillegeist, S. (2007). How disclosure quality affects the level of information asymmetry. Review of Accounting Studies, 12, 443–447. Chen, Q., Goldstein, I., & Jiang, W. (2007). Price informativeness and investment sensitivity to stock price. The Review of Financial Studies, 20, 619–650. Demski, J. S., & Feltham, G. A. (1994). Market response to financial reports. Journal of Accounting and Economics, 17, 3–40. Dutordoir, M., Roosenboom, P., & Vasconcelos, M. (2014). Synergies disclosure in mergers and acquisitions. International Review of Financial Analysis, 31, 88–100. Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. The Journal of Finance, 25, 383–417. Fama, E. F., & Miller, M. (1972). The theory of finance. New York: Holt, Rinehart & Winston. Foucault, T., & Fresard, L. (2014). Learning from peers’ stock prices and corporate investment. Journal of Financial Economics, 111, 554–577. Gao, P., & Liang, P. J. (2013). Informational feedback, adverse selection, and optimal disclosure policy. Journal of Accounting Research, 51, 1133–1158. Goldstein, I., Ozdenoren, E., & Yuan, K. (2013). Trading frenzies and their impact on real investment. Journal of Financial Economics, 109, 566–582. Hagedorn, J., & Duysters, G. (2002). The effect of mergers and acquisitions on the technological performance of companies in a high-tech environment. Technology Analysis & Strategic Management, 14, 67–85. Han, B., Tang, Y., Yang, L. (2014) Disclosure and discretionary liquidity trading: Implications for market liquidity and efficiency. Unpublished working paper. Healy, P. M., Hutton, A., & Palepu, K. (1999). Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16, 485–520.

9

Healy, P. M., Krishna, G. P., & Ruback, R. S. (1992). Does corporate performance improve after mergers? Journal of Financial Economics, 31, 135–175. Heflin, F., Shaw, K., & Wild, J. (2005). Disclosure policy and market liquidity: Impact of depth quotes and order sizes. Contemporary Accounting Research, 22, 829–866. Jovanovic, B., & Rousseau, P. (2002). The q-theory of mergers. American Economic Review Papers and Proceedings, 92, 198–204. Kau, J. B., Linck, J. S., & Rubin, P. H. (2008). Do managers listen to the market? Journal of Corporate Finance, 14, 347–362. Kyle, A. S. (1985). Continuous auctions and insider trading. Econometrica, 53, 1315–1335. Leuz, C., & Verrecchia, R. E. (2000). The economic consequences of increased disclosure. Journal of Accounting Research, 38, 91–124. Luo, Y. (2005). Do insiders learn from outsiders? Evidence from mergers and acquisitions. The Journal of Finance, 60, 1951–1982. Mitchell, M. L., & Mulherin, J. H. (1996). The impact of industry shocks on takeover and restructuring activity. Journal of Financial Economics, 41, 193–229. O’Hara, M. (1997). Market microstructure theory. Blackwell Publishers. Ozoguz, A., Rebello, M. (2013). Information, competition, and investment sensitivity to peer stock prices. Unpublished working paper. Polasky, S., & Mason, C. F. (1998). On the welfare effects of mergers: Short run vs. long run. The Quarterly Review of Economics and Finance, 38, 1–24. Ravenscraft, D., & Scherer, F. (1987). Mergers, selloffs, and economic efficiency. Washington: Brookings Institution. Roll, R. (1986). The hubris hypothesis of corporate takeovers. The Journal of Business, 59, 197–216. Welker, M. (1995). Disclosure policy, information asymmetry, and liquidity in equity markets. Contemporary Accounting Research, 11, 801–827.

Please cite this article in press as: Bade, M. The effects of mergers and acquisitions on the information production of financial markets. The Quarterly Review of Economics and Finance (2016), http://dx.doi.org/10.1016/j.qref.2016.09.006