St. Petersburg State University
Graduate School of Management
Master in Corporate Finance
Reaction of the bondholders to the M&A deals:
evidence from American oil & gas market
Master’s thesis by the 2nd year student
Vladimir Kadnikov
Research advisor:
Vitaly Okulov, Associate professor
St. Petersburg
2016
АННОТАЦИЯ
Автор
Название магистерской
диссертации
Факультет
Направление подготовки
Год
Научный руководитель
Описание цели, задач и
основных результатов
Ключевые слова
Владимир Андреевич Кадников
Рекция держателей облигаций на сделки слияния и поглощения
на примере американского рынка нефти и газа
Высшая Школа Менеджмента
Корпоративные финансы
2016
Виталий Леонидович Окулов
Целью данного исследования ставился анализ реакции долгового
рынка на сделки слияния и поглощения, а также исследование
реакции с позиции гипотезы перетекания ценности между
участниками сделки. Мы использовали метод разность разностей
для оценки реакции долгового рынка и метод оценки
собственного капитала как опциона кол для оценки эффекта
перетекания ценности между держателями облигаций и
акционерами поглощающей компании. На данных американских
компаний нефтегазового сектора была обнаружена реакция
кредиторов: стоимость облигаций поглощающих компаний в
среднем упала на 4,37 единиц сильнее, чем в среднем упала
стоимость облигаций компаний, не участвующих в сделках
слияния и полглощения. Для 84% поглощающих компаний,
имевших негативную тенденцию стоимости облигации, был
найден эффект перераспределения ценности от держателей
облигаций к акционерам компании. Этот эффект может
объяснять снижение стоимости облигаций: держатели облигаций
стремились компенсировать потери посредством предъявления
больших требований к доходности облигаций, что впоследствии
снизило цену облигации. Принимая эффект перераспределения
ценности между акционерами и кредиторами поглощающей
компании за объясняющий фактор реакции долгового рынка на
сделку, авторы выявили влияние риска на сам эффект:
обнаружено повышение общего риска для 91% компаний с
эффектом перераспределения цености. Повышение риска
происходило посредством покупки более рисковой компаниицели. Данная цепочка «риск-перераспределение ценности»
согласуется с теорией оценки оционов.
Реакция долгового рынка, держатели облигаций, сделки слияния
и поглощения, рынок нефти и газа, эффект перетекания
ценности, разность разностей, модель Мертона, оценка опциона
3
ABSTRACT
Master Student Name
Master Thesis Title
Faculty
Main field of study
Year
Academic Advisor’s Name
Description of the goal,
tasks and main results
Keywords
Vladimir A. Kadnikov
Reaction of the bondholders to the M&A deal: evidence from the
American oil&gas market
Graduate School of Management
Management in Corporate Finance
2016
Vitaly L. Okulov
The aim of this study is to analyze of debt market reaction to M&A
deal as well as to identify the debt market reaction from the
perspective of the hypothesis of the wealth redistribution between the
participants of the deal. We used the difference in differences method
for the evaluation of the debt market reaction and option pricing
theory for estimation of the effect of the wealth redistribution
between bondholders and shareholders of the acquiring company. On
the basis of data of US oil&gas companies the reaction of creditors to
the M&A deal was detected: the value of the bonds of acquiring
company on average falls by 4.37 units, which is greater than the
average bond price drop of the companies that did not participate in
any M&A deal. For 84% of the acquiring companies that have
negative trend in their bond prices the effect of wealth redistribution
from bondholders to shareholders was found. This effect can explain
the decline in the bonds prices: bondholders seek to compensate for
the loss by claiming for higher yield, which subsequently reduces the
bond price. Considering the effect of the wealth redistribution
between shareholders and creditors of the acquiring company the
explanatory factor of the debt market reaction to the deal, we found
the impact of the risk on the effect itself: the risk increases for 91% of
the companies with the wealth redistribution effect. The increased
risk occurs due to the purchase of riskier target. Such link "risk –
redistribution effect" is consistent with the option pricing theory.
Debt market reaction, bondholders, mergers and acquisitions, oil&gas
market, wealth redistribution, difference in differences, Merton
model, option pricing
4
Contents
Introduction..................................................................................................................................6
Chapter 1. Theoretical grounds of the market reaction on M&A deal........................................9
1.1 Motives of M&A..............................................................................................................9
1.2 M&A deal: the consequences for stakeholders of bidding and target companies..........11
1.3 Market reaction to the M&A deal.................................................................................. 13
1.4 Conflict of interest between shareholders and creditors................................................ 17
1.5 Research gap.................................................................................................................. 20
1.5 Summary and important considerations.........................................................................21
Chapter 2. Research methodology............................................................................................. 22
2.1 Research problem...........................................................................................................22
2.2 Research design..............................................................................................................23
2.3 Research method............................................................................................................ 24
2.3.1 Difference-in-Difference method............................................................................25
2.3.2 Merton model..........................................................................................................29
Chapter 3. Measuring the reaction.............................................................................................36
3.1 Data collection................................................................................................................36
3.2 Empirical analysis results...............................................................................................39
3.2.1 Bond market reaction..............................................................................................39
3.2.2 Wealth redistribution effect.....................................................................................42
3.2.3 Factors influencing wealth redistribution............................................................... 45
4.3
Managerial implications............................................................................................ 47
3.4 Research limitations....................................................................................................... 48
Conclusion................................................................................................................................. 50
References..................................................................................................................................52
Appendices................................................................................................................................ 55
5
Introduction
Mergers and acquisitions (hereinafter M&As) field was always important in the economic
and financial science and even nowadays, in a time of financial crisis, it attracts the attention of
researchers. Generally, researchers are interested in positive effects of M&A activity as a source of
the growth for companies. In particular, such effect as creation an additional value for shareholders
is the most important factor for being M&A bargain successful. The performance of each private
company in capitalistic world must bring profit and, consequently, create value for shareholders.
Thus M&A activity, as an essential part of the performance of growing companies, should also
create additional value. In reality, even though there are many concerns on whether M&As bring or
destroy the real value it remains the main source of companies’ growth.
Alongside the creation of value in M&A deal the effect of value redistribution from
shareholders to creditors or vice versa can be observed. Managers are supposed to increase
shareholder value by leading the company in a proper way. Thus, such value redistribution during
M&A deal should be thoroughly tracked by managers in order to protect shareholders’ interests. The
debt market reacts to M&A deal, in particular – to the characteristics of the target in the deal, by
changing creditors’ required yield. Such yield volatility enables creditors to capture additional profit
or at least hold the wealth. For managers it is essential to know what target’s characteristics are
important for creditors to be able to predict the credit rate for the current obligations of the parent
company.
The object of this study is bond prices of the companies participating in M&A deal. Followed
by the research object, the following research subject is identified: the influence of M&A deal
announcement on the bond prices
Thesis main goal is to analyze the debt market reaction to the M&A deal and investigate the
reaction from the perspective of wealth redistribution effect
To reach the main goal of this research the following objectives are stated:
1.
To analyze theoretical background of stock and debt market reaction to M&A deals
and identify factors, which influence the marker reaction to the deal;
2.
To analyze prior researches on conflict of interest between shareholders and
creditors;
7
3.
To analyze the empirical methods of measuring the reaction of stock and debt
4.
markets to M&A deals;
To collect the data for empirical analysis on the American market and restructure that
5.
data for proper application of the chosen methodology;
To conduct an empirical study using Difference-in-differences approach to identify
6.
debt market reaction to the M&A deal;
To estimate the effect of welfare redistribution between the stakeholders using
7.
Merton model;
To identify the interconnection between welfare redistribution in M&A deal and the
8.
bondholders reaction to M&A deal;
To identify the factor influencing the welfare redistribution between shareholders and
bondholders.
To complete the analyses the following research questions are set:
What are the factors that influence the market reaction to the M&A deal?
In what way the conflict of interest between shareholders are creditors affects the
gains and losses of both counterparties?
In what way the debt market reacts on the M&A deal on the oil&gas market?
To what extent the acquiring companies are exposed to welfare redistribution effect?
How the welfare redistribution effect and the debt market reaction are interconnected?
What factors do influence the welfare redistribution effect?
The primary method for the analysis is Difference-in-Differences method. This statistical
method is widely used for catching an effect difference between different groups. Option pricing
model (Merton model) is the second method used in this study for catching the welfare
redistribution effect. The complete methodology of our research is described in detail further in this
paper.
As a final result of our research we will consider the answers on all research questions as
well as explicit managerial implications. The results of the empirical part can be useful for the
agents of the deal – both strategic management, who decides on the essentiality of the potential deal
in terms of potential effects for the company in relation to its bondholders, and bondholders of the
acquiring company in respect to their interests.
The paper is divided into 5 main parts: introduction, 3 chapters, and conclusion. The first
chapter is devoted to the theoretical issues related to M&A deal, bondholders yields, conflict of
interest between shareholders and bondholders that influences their positions in the M&A deal, and
literature review of key papers and previous researches in this field as well as determination of the
8
research gap, which we are aiming to fulfill by our study. In the second chapter we explicitly define
the research problem and described the research design and the chosen research method. The third
chapter is a description of the empirical part of our research – data collection process as well as
main findings are shown there. Finally, in conclusion we summarized the whole logic of the paper –
there we derive with main managerial implications on the bondholders’ reaction on M&A deal.
9
Chapter 1. Theoretical grounds of the market reaction on M&A deal
1.1 Motives of M&A
Mergers and acquisitions are extremely popular as a growing strategy for business both
nowadays and in the past century: from the very beginning of the XX century there were 6 so called
M&A “waves” (Bruner, 2009). The last M&A wave occurred just before the international financial
crisis in 2007-2009 due to easy access of the multinational firms as well as SME 1 to the capital and
funding.
Graph 1.1.1 M&A waves2
Definitely, there should be reasonable motives for companies to engage in mergers and
acquisitions. Then, the question “Why do companies participate in M&A?” arises. We define 3
major motives why firms engage in M&A activity: the potential for synergy, the agency motive, and
the hubris motive (Berkovitch and Narayanan, 1993).
The synergy motive for M&A is possible because in some cases the combined entity
(acquirer) have greater market value than the sum of premerge market value of counterparties. In
other words, combining the two companies creates synergies. Under this motive it is assumed that
managers of both bidding and target firms aim to increase the shareholders’ wealth and then engage
in M&A only in case of win-win deal, where the shareholders of both entities gain additional wealth.
To make the abovementioned statement reasonable, the hostile acquisitions are not considered for
1 SME – small and medium enterprises
2 M&A activity: Where are we in the cycle
10
explaining the synergy motive. However, the synergy resulting from the deal does not mean that
both parties get the fair amount of wealth (or that the so called “synergy” wealth, which is the
difference between the combined entity market value and the sum of the market values of both
M&A parties just before the deal): the bargaining power of one of the firms engaging in the deal
leads to a bigger amount of wealth gaining.
All synergies can be divided into two categories: operating and financial synergies. Operating
synergies is that that is achieved by obtaining the economy of scale and/or economy of scope by
creating advantage of combined entity in production, distribution and marketing activities, the
transfer of skills and expertise by acquirer’s management team, or the acquisition of new technology
or intangible assets (e.g. acquisition of knowhow of new markets). Operating synergies tend to arise
primarily when both firms are from the same industry or their business somehow connected: either
in horizontal or vertical value-chains. Financial synergies are those that are achieved basically in
conglomerate acquisitions, when the businesses of the deal parties are not interconnected. Mostly
they arise as a part of diversifying strategies. The good results in such deals are possible due to
cheaper access to capital, an internal capital market, cash flow stability or a lower bankrupt
probability (Martynova and Renneboog, 2006). Financial synergies aim to lower the cost of capital:
9 out of 10 of the acquirers or the target used to achieve the lower cost of capital (Frommelt, 2004).
The second motive is the agency motive, which suggests that it is the self-interest of the
bidding firm management that is a major driver and the main motivation for M&As. According to
this theory, M&As primarily take place when the bidding managers are willing to increase bidding
firm market value and then the bidding firm management welfare at the cost of target’s value and its
shareholders. There are several studies supporting this theory. Amihud and Lev (1981) mentioned
the aim of diversification of management’s personal portfolio. Jensen (1986) showed that managers
use free cash flow in order to increase the size of the bidding firm instead of the increasing of firm
value because of the mismatch of the private benefits of the management and the shareholders of the
bidding company. Shleifer and Vishny (1989) concluded that acquiring assets can increase the
dependence on the bidding management and that is a motive for M&A incentives from the
management perspective. For example, specialist managers acquire assets in their own line of
business so that the company depends even more on them (Berkovitch and Narayanan, 1993). The
idea is that the bidder’s management extract the value from the bidder’s shareholders. Target
shareholders in such deals try to gain as much welfare as possible by exploiting their bargaining
power. Therefore, the more severe the agency problem is the higher the target shareholders gain.
11
Such behavior of the acquirer’s managers leads to high agency costs, which in turn reduce the total
shareholder value.
The third motive is the hubris hypothesis, which suggests that the target valuation mistakes of
the bidding firm management are the reason of many mergers and acquisitions. For this reason many
companies engage in M&As and get no benefits including synergies. According to this theory,
M&As are the result of overvalued synergies and overconfident managers (Frommelt, 2004). Under
the assumption that there is an equal probability of overestimation and underestimation of the
synergy from the bidder’s management perspective, transaction is a case of overestimation. If not,
then the target shareholders will accept underestimated synergy and corresponding underestimated
offer, but it does not hold in the reality as they rationally reject such economically disruptive deals.
As the synergies under hubris hypothesis are zero, the deal itself is basically a transfer of wealth
from the bidder to the target:the higher the gain of the target the greater the loss to the bidder.
Consequently, while bidders lose their value, the net economic gain of the transaction is zero.
According to theory, mergers and acquisitions are value increasing events for target
shareholders in all three motives. However, for acquirers, both hubris and agency motivated
acquisitions will have a negative impact on their shareholder value.
1.2 M&A deal: the consequences for stakeholders of bidding and target companies
Even though M&A activity is very popular among relatively large companies and
corporations, still the conclusions on its benefits, both short-term and long-term, for the shareholders
are controversial. On the one hand, there are many researches that prove the existence of value
creation during the M&A, on the other hand, some empirical researches catch the value disruption in
the M&A deals. As it was discussed in the previous paragraph, the shareholders of bidding and the
target companies may have different results from the deals. In this section we observe literature on
the field of benefits and loses for the stakeholders of the M&A deal parties.
Every time when economic agents face with information, the problem of information
asymmetry arises: the same is true for M&A deals. Information asymmetry can destroy the market
by making good agents leave the market (Akerlof, 1970). When two counterparties have different
information, the market is stagnating – Akerlof called such markets as “markets of lemons”. Even
though the M&A market is also exposed to the information asymmetry, still this market is very
active: probably shareholders and managers eager to succeed with the deal, thus, overconfidence and
12
belief in the M&A success make the market of mergers and acquisitions active. Anyway, the success
of the deal depends on various factors: the growth prospects of the counterparties, the nature of their
business and their interdependence, type of the deal, mean of payment, capital structure of acquiring
and target firm, management decisions, financial stability, the economic cycle, etc.
Kirchhoff and Schiereck (2011) on the basis of pharma market show that combined entity
does not have significant announcement effects and moreover the acquirers destroy the value of their
shareholders. Moeller, Schlingemann and Stulz (2005) massively analyzed M&A deals with no
reference to a specific industry: they found that for the period 1998-2001 the shareholders of
acquiring companies on average lost 12 cents around acquisition announcement per dollar spent on
acquisition. A bunch of papers concludes that the shareholder value resulting in M&A is negative:
Lyroudi et al. (1999), Eckbo and Thorburn (2000), Bruner (2002).
Morck, Schleifer and Vishny (1990) analyze the M&A from the view of managers and state
that value destroying is caused by managerial objectives of acquiring firm. Their results corresponds
with the agency theory (agency motive for M&A from previous paragraph) when managers and
shareholders have different objectives and that results in value-destroying deals for the bidding
companies.
Datta et al. (1992) conducted meta-analysis of M&A deals to answer the question what
factors influence the wealth creation from mergers and acquisitions. Using multivariate framework
they concluded that while the target’s shareholders gain significantly from mergers and acquisitions,
those of the bidding firm do not. They found that mean of payment, the presence of multiple bidders,
and the type of acquisition have a direct impact on the wealth creation prospects but still majority of
the deals are value destroying from the viewpoint of acquirers’ shareholders.
This finding about benefits for targets firms support many researchers: Jensen and Ruback
(1983), Bruner (2002), Eckbo and Thorburn (2000), Jarrell (1988).
Fuller et al. (2002) found the conditions of the deal, which on average lead to the gain of the
bidding firms: bidder shareholders gain in case of buying a private firm whereas public target can
lead to their losses. This conclusion is important for us because we can compare our findings on
wealth gaining/losing/redistribution wealth in case of private/public target. Moreover, authors state
that the return is greater if the bidder uses stocks as a mean of payment.
13
As for the combined entity, targets benefits lead to the overall significantly positive effects
according to Bruner (2002) and Andrade el al. (2001). However, these positive effects remain in
magnitude below the targets’ gains.
So far shareholders’ wealth changing in M&As was discussed. At the same time the interests
bondholders are considered important for our research as we further study the reaction of the
bondholders of the acquirer on the M&A deal. Decisions favorable for shareholders do not always
increase the value of the company and can cause economic damage for bondholders. The reason for
that is different objective functions of agents. Hilscher and Sisli-Ciamarra (2013) supported this
view and stated that some announcements of M&A are associated with lower shareholder value,
higher creditor value, and lower overall firm value when a creditor is present. In other words,
conflicts of interest or if not conflict then just one of classical agency problems “creditors vs
shareholders” can result in destroying the value (from the viewpoint of shareholders) of acquisitions.
The authors developed the idea of different objective functions of creditors and shareholders – the
former prefer diversifying acquisitions whereas the latter are interested in cash-financed industry
specific mergers and acquisitions. Even though authors primarily looked at the conflicts on
corporate boards of directors, they found significant causal link between the existing of creditors in
the company and its success in M&A in terms of value generating – the corporate governance
mechanism works in the interest of those who are the part of the board of directors: creditor-director
approves the deals with negative value for shareholders while shareholder-director does the same for
creditor. The same conclusion is made by Jensen and Meckling (1976).
We can conclude that value creation potential of a certain M&A for the bidder, target and the
combined entity is under the great uncertainty taking into account issues of information asymmetry
and conflict of interest between shareholders and creditors. Still, researchers try to find the
determinants of the success deal like mean of payment, organizational form, industry belonging, and
different characteristics of the bidding and target companies. We discuss these characteristics on the
next paragraph.
1.3 Market reaction to the M&A deal
In order to analyze the effect of M&A announcement on the debt market certain
characteristics of the deal and target involving in the deal should be specified. To do so a bunch of
research papers, both theoretical and empirical, devoted to the market response on M&As were
14
analyzed. First, we need to distinguish the stock market reaction and the debt market reaction.
Previously we discussed that shareholders and creditors have different objective functions. Thus, we
conclude that in some cases the shareholders and the bondholders can react differently on the same
M&A deal.
Many researchers evaluate the stock market reaction or the synergy effect of M&A by
calculating abnormal returns. Generally, abnormal returns are the returns that are greater than those
predicted by the statistical methods (often Event studies approach) on the basis of the historical data.
The existence of abnormal return means that the particular stock beat prediction of own price
behavior based on the historical data and the market index, which is usually taken in the models as a
benchmark.
Jansen and Stuart (2014) analyzed the factors that can help CEOs to predict the stock market
reaction on the deal announcement. According to them, the acquisitions can be positive net present
value projects, especially those leading to the economies of scale/scope or lowering the cost of
capital, still can cause a negative stock market reaction. They used the concept of CAR (cumulative
abnormal returns) in estimation of the reaction on the announcements and stated that average M&A
announcement has a positive CAR but the volatility of stock reaction is very big. According to the
researchers, the great portion of such reaction variation can be explained by 3 factors:
Size of the bidder: defined by firm market value, small firms have the market
capitalization below 25 th percentile of all firms trading on NYSE (at the end of 2012
this cutoff was around $335 million);
The ownership status of the target: public or private;
The method of payment: cash or shares of stocks.
The authors stated that these factors have immediate influence on the company’s market
value when the acquisition is announcing. They reported evidence on the stock price reaction on the
M&A from the point of bidder: the data set of almost 17 M&A 000 announcements for the period
1980 – 2008 was used. Regarding the abnormal stock returns measured with CAR methodology, the
factor analysis was used:
Small bidders had average CAR of 1,74$%, while large firms gained negative return
-0,08%;
Prior studies provided several possible explanations for why firm size matters and concluded
that the most likely reason is managerial overconfidence: CEOs and executives of the acquirers
15
generally overestimate their ability to manage the target and thus overpay for it. In 1986 Roll
introduced the hubris hypothesis of corporate takeovers: he stated that acquiring firms infected by
hubris first overvaluate and then overpay for the target. Roll empirical evidence supported his view
of hubris hypothesis. Moreover, he argued that overconfidence of the bidders is as much significant
as other explanation factors such as synergy, inefficient target management, and taxes. Other authors
support the view of managerial overconfidence as an important factor for M&A announcement
success – Jansen et al.(2013), Moeller et al. (2004).
Private targets bring much more for the bidder in contrast with those of publicly
traded: acquisitions of private targets result a positive abnormal return of 1.37%,
while acquirers of public targets lose due to a negative abnormal return of −1.20%;
Certain market expectations about the private companies may lead to inadequate or biased
valuation of them, which in turn lead to overpayment and negative return during M&A
announcement. Fuller et al. (2002) provided the most recent and complete explanation of the
relation between ownership status of the target and the abnormal returns. They argued that because
of much less liquid market for the shares of private companies, their negotiation position is weaker
than that of the public companies. Public targets use their relatively stronger negotiation power to
get the higher bidding price: on average the value left to the bidders’ shareholders is less than the
bidding price, thus the are left with the negative CAR.
Cash acquisitions on average bring 1,01% CAR, the mix of cash and shares results in
0,8% CAR, stock deals bring negative -0,01% CAR.
Here signaling hypothesis works: the issuing of equity for the deal is a bad signal to the
market that means for investors that the share price is too high and the bidder is trying to take
advantage from such overvaluation. The existence of such signal is possible due to the information
asymmetry: the shareholders and managers know more about their business than the external agents
(investors, analytics, creditors, etc). Myers and Majluf (1984) described this signal hypothesis in
detail.
Shams et al. (2013) confirmed the findings of Jansen and Stuart by researching the influence
of organizational form and the method of payment on the possibility of gaining abnormal returns in
M&A in case of public, private, and subsidiary acquisitions on Australian market. They concluded
that both factors are relevant for determine abnormal returns.
16
Apart from the abovementioned factors, other influential factors, which can help to predict or
explain the market reaction , exist. Such characteristic as company efficiency is also valuable for
catching response effect. Al-Khasawneh and Essaddam (2012) found that mergers combining low
efficiency acquirers and targets create significant market returns following the merger event, while
mergers combining the least efficient acquirers with moderately efficient targets diminish the
acquirer's wealth more than any other type of merger.
Jansen and Ivo (2015) studied the volume reaction to M&A announcements. They found that
for acquiring firms such factors as method of payment, target ownership, firm size and the relative
size of acquisitions are statistically significant for market response.
Shah and Arora (2014) found that the target and the bidding firms are affected by the reaction
differently: the target firms depict that the post announcement returns are significantly greater than
the pre-announcement returns, indicative of the immediate market reaction to the information
disclosure, while the bidding firm do not show statistically significant abnormal returns.
Bouzgarrou and Louhichi, (2014) aimed to fill the gap of research of distinguishing between
the method of payment and the means of financing in M&A deals and tests if the financing means
has incremental information beyond that contained in the payment means. One of the findings of
this research is the fact that market reaction depends on legal environment (common law vs. non
common law) on acquisition characteristics such as deal size and on acquirer specific factors such as
size and growth opportunities.
Along with the choosing of characteristics of the target and the deal in general, the proper
using of methodology is important. Palmucci and Caruso (2011) analyzed Italian market of M&A of
banks and they found that event period should be extended by the “rumor date” – this is important to
catch full effect of market response. They showed that not all the effect of market reaction can be
measured if the event date is taken as announcement date. It was stated in the research paper that
using wider event window including so called “rumor date” bigger portion of market reaction to
M&A is captured.
So far we observed the stock market reaction on mergers and acquisitions. In fact, most of
researchers’ attention is devoted to stock reaction rather than bond market reaction. The reason of
such massive interest to the market reaction is clear shareholder and managerial implication – by
17
determine the factors of market reaction on the M&A deal companies’ governors can have a power
to predict the reaction and/or influence it and/or decide whether to engage or not in a particular deal.
Still, the debt market reaction (in our case bondholders’ reaction) exists and it in our interest
to observe the relevant literature.
Penas and Unal (2004) analyzed mergers in the banking sector and pay attention fully to the
bond market reaction. They used cross-sectional analysis and found out that the determinants of the
bondholder gains during M&A process are diversification gains, gains associated with achieving
too-big-to-fail status, and synergy. Corporate banking mergers can influence bondholders differently.
In case of synergistic merger, both bondholders and shareholders win because of the possibility of
achieving economies of scale and scope by the combined entity through M&A. Another reason of
synergistic M&A participants gain is elimination of less-efficient management. The second reason is
well developed by Jensen and Ruback (1983). Further, in nonsynergistic mergers, bondholders gain
in case of reducing the cash flow volatility resulting from M&A. The lower the cash flor volatility
the lower the default risk. Other researchers support this idea as well: Higgins and Schall (1975),
Galai and Masulis (1976).
Some researchers argue that bondholder may gain wealth in M&A deals through coinsurance
effect: Shastri (1990) analyzed cases of different risk levels, leverage ratios, and debt maturities of
the bidders. He argues that the acquirer’s bondholders either gain from coinsurance effects or lose
from expropriation effects: the resulting effect depends on the deal, bidder and target characteristics.
The author shows that wealth redistributions from stockholders to bondholders (or vice versa) or
within securityholder classes occur frequently, depending upon the covariance between the returns
of the merging firms.
1.4 Conflict of interest between shareholders and creditors
In this section we discuss the issues related to the interest of different stakeholders of the
company. We argue that shareholders and creditors (as well as shareholders and managers) have
different objective functions. Thus, their decisions are biased by their personal preferences and
interests, which can lead to a potential loss of the counterparty’s welfare.
There are several types of conflict of interest. In business these types are:
Managers Vs Shareholders
Shareholders Vs Managers
18
Conflicts of interests arise in the firm when the incentives of counterparties are different and
mutually exclusive. For example, generally managers tempt to not participate in risky projects to
preserve the job position whereas shareholders can be interested in risky projects that bring higher
returns. On the other hand, in leveraged firms managers tend to maintain high level of risk by
investing in risky projects. This phenomenon is known as asset substitution, when less risky assets
are substituted by the assets with higher risk. Managers choose the riskiest investment alternative
and that is not in the interest of shareholders – again we observe the conflict of interest. In general,
any separation of control and managing rights creates the agency problem or conflict of interests
between shareholders and managers.
Shareholders and creditors have different incentives as well. Bondholders may suffer from
aggressive investment politics of the company, because such politics brings additional risk to the
company’s profile, which in turn is not in the interest of the creditors. Here and after by conflict of
interest we mean the conflict between shareholders and creditors (bondholders).
We exploit the idea of conflict of interests and base on that idea the possibility of welfare
redistribution in M&A deal. Option theory gave the researchers the powerful technique of pricing
assets. Damodaran (1995) showed how option theory can be applied to illustrate conflict of interest
between shareholders and creditors (agency problem) in case of M&A. He stated that decisions
favorable for shareholders do not always increase the value of the company and can cause economic
damage for bondholders. Presenting the firm value as the sum of the equity and the debt market
values, we can argue that even if the value of the firm is reducing the value of the equity can
increase by conquering some value of debt.
Conflicts of interest between shareholders and creditors can result in value-destroying
acquisitions, when stockholders invest in negative NPV projects (M&A deal) that leads to higher
volatility of the firm free cash flow and, as a consequence, higher value of equity. So, it appears that
investors in equity or stockholders win due to higher equity value, while overall firm value reduces
due to negative NPV of a project. The enterprise value is a sum of equity and outstanding debt
values. So, in our example, the firm value decreases and the equity value increases at the expense of
bondholders, because the debt value reduces. Such simultaneous increase of equity and decrease of
debt is the welfare redistribution from bondholders to stockholders. Below an example of wealth
redistribution is provided.
Table 1.4.1 The numerical example of wealth redistribution between stakeholders
19
Auto corp
Costmetics corp
Joint entity (result of M&A)
The value of equity
75,94
134,48
207,58
The value of debt
24,06
15,52
42,42
Enterprise value
100
150
250
Here we can see that two companies participated in M&A deal. The value of combined entity
is just the sum of premerge values of counterparties (250=100+150). At the same time, we cannot
say the same about the equity and debt values:
Combined entity’s equity value is lower than the sum of premerge values of equities:
207,58 against 210,42 ( the absolute change is -2,84);
Combined entity debt value is higher than the sum of premerge values of companies’
debt: 42,42 against 39,58 ( the absolute change is +2,84);
So, the welfare of creditors has increased by the same amount as the welfare of shareholders
has decreased. In this case we conclude that the deal led to welfare redistribution from shareholders
to bondholders. In order to catch such welfare redistribution effect, option pricing model is used for
estimation equity and debt values.
Even though option methodology for pricing equity has its limitations – the calculated price
is relevant basically for the distressed business – it is still a good instrument for explaining why
M&A can make creditors more wealthy and shareholders – less. Damodaran contends that good
M&A deal leads to a less variable cash flow and, as a circumstance – less variable value of the firm.
The lower the dispersion the lower the price of option and the value of equity. Thus, during M&A
process new firm has a value of equity lower than the sum of parent’s and target’s equities.
1.5 Research gap
In the previous paragraphs certain aspects of M&A deal were discussed: in particular,
motives of M&A deal, wealth gains and losses of shareholders of both the acquirer and the target,
and finally the stock and bond markets reactions on the M&A deal. The topic of this research paper
is the credit market reaction on the deal. The choice of the topic was determined by the number of
20
relevant studies that is much less than the number of studies devoted to the market reaction on the
M&A deal. As we mentioned earlier, such split of researchers’ attention is reasonable – the most
concerns regarding M&A success or failure are about shareholder’s welfare creation or destruction.
It is indeed very important to realize the M&A success determinants to predict the M&A effects on
the shareholders, to predict the market reaction, which can contribute to the gain or losses of the
shareholders, and finally to decide whether engage or not in the deal taking into account all essential
information (target characteristics, potential results, potential for synergy, etc.), information
asymmetry, and uncertainty. From the perspective of managerial implications all mentioned above is
clear. However, the debt market reaction is interesting for us from the perspective not the wealth
creation and destruction, but the wealth redistribution. By redistribution we mean the conflict of
interest between shareholders and bondholders, which is a classic example of agency problem. In
case of zero synergy effect, we predict the potential welfare redistribution between shareholders and
bondholders: the former can gain at the expense of the latter or vice versa.
Here is what makes our research especially valuable and topical – the field of research lies on
the intersection of debt market reaction and wealth redistribution between shareholders and
bondholders.
First, we not only aim to find out how debt market reacts on M&A deal but in additions try to
check the hypothesis whether the debt market reaction can be explained from the perspective of
welfare redistribution between stakeholders of the acquiring firm.
Second, we use a novel method of study – a mix of difference-in-differences (DD) method
and option pricing method or Merton model of equity valuation. Almost all research papers that aim
to find and explain the market reaction, both of stocks and of bonds, base their empirical part on the
Event studies, which can be biased by inappropriate estimation of parameters and thus give
inadequate results, biased by external factors. DD method eliminate such problems. Next, option
pricing theory can help to explain the peculiarity of the welfare redistribution. More detailed
information is provided in the paragraphs related to the methodology in the second chapter.
1.5 Summary and important considerations
So far we analyzed the questions regarding M&A deal and its participants. Namely, we
defined the motive of M&A: synergy effect, agency and hubris motives. We found out that
depending on the motive of M&A the benefits of the bidding and the target companies’ shareholders
21
vary: in case of synergy motive both the acquirer and the target gain additional wealth, while in case
of agency and hubris motives most probably only the target gains considerably at the expenses of the
acquiring company.
We found out that the potential for wealth creation or destruction varies according to the the
characteristics of the deal: counterparties characteristics, their business interrelation, motive of
M&A, etc. Important observation is that bondholders can gain or lose in the same manner as
shareholders gain or lose. Thus, it is possible to research debt market reaction.
The review of relevant literature reveals the determinants of the market reaction such as
mean of payment, organizational form, size of the entities and business interrelation, motive of
M&A and others.
This study aims to analyze debt market reaction on M&A deal in the American oil&gas
market by using the methodology of difference-in-differences and option pricing methods.
22
Chapter 2. Research methodology
2.1 Research problem
In this research we aim to find out whether the debt market on behalf of the creditors of the
acquiring company reacts to the M&A: namely, how the prices of bonds issued before the deal’s
announcement are changing due to the potential redistribution of the wealth between the parent’s
shareholders and bondholders. We study American oil&gas industry just for the ease of data
collection: in order to conduct a study we observed 270 companies that made 935 bond issues from
2000 to 2015. We had an access to CBONDS database where we observed the behavior of bond
prices of the chosen companies.
The oil&gas U.S. industry was chosen inventively: America is located far from the main oil
consumers such as China and Europe. Exporting crude oil from the U.S. for decades was largely
illegal due to legislation ban for export. Only in 2016 U.S. oil was introduced on the European
market (Italy) after the 4 decades of the strict ban. Thus, probably U.S. oil&gas industry was
independent from world crude oil price fluctuations and crisis. Such independence is an essential
factor to be sure that at least world’s market features do not influence the prices of stocks in the
American oil&gas sector: in our case potential debt market reaction on the M&As is “cleared” from
the exogenous factors outside the U.S. But still market and different macro factors, which are the
features of U.S. market itself, can bias our results – in order to prevent result biasness we will use
Difference-in-Differences method accompanied with rigorous data collection in terms of companies’
similarities.
At the current stage of research, we are focusing on the bond prices in 2 time periods: 1 year
before and 1 year after the M&A deal announcement. There is no theoretical justification of such
choice of observation. Rather, prior empirical studies shown the market reaction several months in
advance of the deal announcement: such reaction is based on the rumors and inside information. To
mitigate such influential factor we have decided to take observation in a year before the event.
Our objective is to investigate whether the bond prices before and after the announcement of
M&A are statistically different within a sampling frame.
23
2.2 Research design
The type of this research is quantitative: final conclusion on research questions is made by
working with financial data and building statistical model. The research design is explanatory one
because empirical part aims to set casual links between factors and dependent variable. The
empirical part of the research is based on Difference-in-Differences method and Merton model of
option pricing. Two methods have been chosen due to the nature of the problem: first, we would like
to observe the debt market reaction on the deal; second, we would like to prove the hypothesis that
the bondholders react in response to the wealth redistribution due to the M&A deal from the
prospective of changing the riskiness of the assets.
To conduct our study we accomplished the following steps:
1. The collection of the data on M&A deals of oil&gas American companies for the period
of 2000-2015 from Thompson Reuters Eikon database;
2. The collection of the data on the bond issues of oil&gas American companies for the
period of 2000-2015 from CBONDS database;
3. The creation of new dataset by matching the two datasets from previous points so that the
final set meets the following requirements:
a. All M&A deal announcements took place during the same year;
b. The acquiring company has a bond debt, which was issued at least a year before
the deal announcement
c. The acquiring company has the same bond debt (from the point b), which will be
repaid not earlier than a year after the M&A deal announcement;
d. The number of companies that satisfy a-c criteria is the biggest possible from the
initial dataset;
This dataset is a treatment group in term of DD methodology.
4. The creation of new dataset, which is a control group in DD methodology: companies
that had bond issues for the same period as companies from the treatment group;
5. The collection of bond prices of the companies from both groups mentioned above for
two periods: 40 days before the earliest M&A announcement and 40 days after the latest
M&A announcement of the companies in the treatment group – from the CBONDS
database;
6. The statistical analysis of bond prices fluctuation with Difference-in-Differences
methodology:
a. Calculating the average bond prices for both groups, treatment and control, for
two time periods, before and after M&A announcement (described in the 5 th
point);
24
b. Calculating the bond prices difference between the groups for 2 period;
c. Calculate the changes in bond prices over a time period for each group;
d. Calculating the treatment effect;
7. The checking of the hypothesis of wealth redistribution due to the M&A deal with
Merton model of option pricing for each chosen company separately:
a. The collection of the enterprise value;
b. The collection of the par value of all debt outstanding;
c. The collection of risk-free rate;
d. The calculation the firm value standard deviation on the annual basis;
e. Estimation the average duration of outstanding debt;
f. Calculation of the value of equity before the M&A deal as a call option;
g. Calculation of the debt value before the M&A deal;
h. Repeat steps f and g for the period after M&A deal;
i. Compare the changes of enterprise value with the changes of equity and debt
value in order to try to catch the redistribution effect between the bondholders and
the shareholders.
2.3 Research method
In our research two methods were used in order to achieve main goal of the thesis:
Difference-in-Differences method and Merton model. The nature of the research problem and the
research design make us use two methods consequently: first, we need to obtain results on debt
market reaction in general, second, we try to understand the nature of that reaction – namely, we are
trying to explain the reaction as redistribution of the wealth between stockholders from the
prospective of option price theory. Thus, there are 2 stages of the empirical part and correspondingly
2 different research methods.
2.3.1 Difference-in-Difference method
Difference in differences (DD) is a statistical technique used in econometrics and quantitative
research in the social sciences that attempts to mimic an experimental research design using
observational study data, by studying the differential effect of a treatment on a 'treatment group'
versus a 'control group' in a natural experiment 3. Generally this method is used during the drug
approval stages when the effectiveness of a new drug is under analysis. The explanatory or
independent variable in this method is the effect of a so called treatment and the response or
dependent variable is an outcome of the experiment. Basically, DD method compares the average
changes over time in the outcomes of the treatment and the control groups. Even though this method
3 Angrist, J. D.; Pischke, J. S. (2008)
25
helps to mitigate the exogenous effects and selection bias, it is still a subject of the certain biases
such as reverse causality. Unlike the time-series estimation of the effect over a time and the crosssection estimation of the effect between groups at a certain time point, DD method uses panel data
and thus measures the differences between the treatment and control group of the changes in the
outcome that occur over time.
Difference in differences requires data set for both a treatment and a control groups for two
or more different time periods, which are time before and after a 'treatment'. Below is the graphical
representation of the basic logic behind DD method.
Graph 2.3.1.1. Graphical representation of DD logic
On the graph above the basic idea behind DD method is provided. There are two groups: S –
control group, P – treatment group. Before the treatment, at time 1, group P had a group average P 1,
while group S – S1. DD method does not provide any explanation why the averages of the groups are
different – this information is taken as granted. The focus of the methodology is to explain the future
changes of averages. From the graph above, with time and after the certain treatment that occurred
in-between of time 1 and time 2, S 1 increased up to S 2, P1 – up to P2. The differences (P2-S2) and (P1S1) are not the same just because of the treatment effect P 2Q – the treatment group was exposed to a
some treatment, while the control group did not, thus P 2Q exists.
So, having the outcomes for 2 time periods, before and after treatment, and 2 groups, control
and treatment, we can determine the treatment effect by comparing the differences in the outcomes
of the groups at time 1 and time 2.
26
Not all of the difference between the treatment and control groups at time 2 (P 2S2) can be
explained as an effect of the treatment: partly the difference is explained by initial difference P 1S1 at
time 1. Assuming the parallel trend 4, DD calculates the "normal" difference in the outcome variable
between the two groups by generating normal outcome for the treatment group P 1Q: the treatment
effect is the difference between the observed outcome and the "normal" outcome (P 2Q).
Formal definition is as follow:
yistststistD=+++
ulde
(1),
which is the main equation of the model, where:
yist is the dependent variable for individual i, given s (group) and t (time);
us and lt are then the vertical intercept for s and t respectively;
Ds
d
t a dummy variable indicating treatment status;
is
is the treatment effect;
e i s tis an error term.
Taking into account that:
1 n
ystist=y �
n i =1
1 n
ggg
�
s s==
s
n i =1
1 n
lll
�
s s==
s
n i =1
1 n
DDD
ststst== �
n i =1
(2)
,
(3)
,
(4)
,
(5)
4 In addition to all OLS assumptions, DD requires a parallel trend assumption – the difference of group
averages is stable over a time.
27
1 n
eestist= � .
n i =1
(6),
δ in statistical terms can be interpreted as the treatment effect of the treatment (
Ds
).t
Ds t is a dummy variable of treatment status, which means that it is a binary variable with 2
meanings: 1 for the treatment group at time 2 (after treatment), and 0 for all other cases (treatment
group at time 1, control group at times 1 and 2).
For our purposes we use DD method to catch the differences in pricing fluctuations for the
bondholders of two groups of the companies: the first group aggregates the firms with bond issues
that participate in the deal (not as a target), the second group aggregates the firms with bond issues
but that did not engage in any M&A deal. The “treatment” as an essential part of DD methodology is
M&A deal announcement, while the treatment effect – the exceeding scale welfare change of the
acquirer’s bondholders in comparison to those of ordinary firms (with no M&A during the analysis
period). The welfare of bondholders directly reflects in their bonds’ prices – thus, we use bond
prices adjusted for the accrued coupon income in DD method.
Table 2.3.1.1 The implementation of DD method
Yst
T=1
T=2
Change
S=1
Y11
Y12
Y11-Y12
S=2
Y21
Y22
Y21-Y22
Difference
Y11-Y21
Y12-Y22
(Y11-Y21)-(Y12-Y22)
Let us go back to our paper goal –estimation of the debt market reaction. In our case, T1 is
time before the M&A deal announcement, while T2 – the time after that. The treatment effect
between these two time periods is the M&A deal announcement itself. Two potential states of the
companies in our case are (S1) non-engagement in the M&A deal but having corporate bonds
issued, and (S2) engagement in the M&A deal and having corporate bonds issued. Y st is a averages
28
for the bond prices: for example, Y11 is an average of bond prices in the group of companies, which
did not take part in any M&A deal, for the time period 1.
So, we will have two groups of the companies, and the bond prices observations for all those
companies for two time periods. By comparing the groups’ average bond prices along the time we
can estimate the effect of M&A deal announcement on the bond price or, in other words, how the
bondholders of the acquiring firm react to the M&A deal.
DD method is a strong statistical method but still it has its own imperfections and limitations.
The great appeal of DD estimation comes from its simplicity as well as its potential to circumvent
many of the endogeneity problems that typically arise when making comparisons between
heterogeneous individuals, according to Bertrand et al. (2004) who quoted Meyer (1995).
Nevertheless, the problem of the method is the possible endogeneity of the treatment, when ideally it
should be random and exogenous. Along with biases in estimating the treatment effect, some
researchers claim about statistical imperfection as well: Bertrand et al. (2004) argue that the main
equation (1) in practice contains the serial correlation problem.
Another limitation – is the method assumption of “parallel trend”, which means that within 2
periods 2 periods model the changes of the treatment group over time would have the same as the
changes of the control group in case of no treatment exposure. We agree with this limitation, but still
think that the advantages of the model (biasness to the external factors, which are not caught by the
classical techniques as Event studies or CAPM, and relative easiness of the usage) still outweigh the
disadvantages. Moreover, in this paper DD method is not the only empirical method, thus we can at
some extent be tolerant to its imperfections.
2.3.2 Merton model
The second method of our empirical research is option pricing model or Merton model. The
model itself is just a formula for calculation of the fair value of the European call option. The
researchers of finance Fisher Black and Myron Scholes first published their model in their paper
“The Pricing of Options and Corporate Liabilities” of 1973. They introduced the formula, or a
partial differential equation, which enables to estimate the option price over time. The idea behind
the formula is that the option itself can be presented as the combination of long and short positions
of underlying and risk-free asset. Such strategy of replication is called hedging strategy or delta
hedging.
29
Merton was the one who generalized the model, introduced the concepts of risk-neutral
probabilities, no-arbitrage bounds, and made the usage of the option-pricing model very popular on
the Wall-Street.
In this paragraph and further in the paper we will use the terms “option-pricing model”,
“Black-Scholes model”, “Merton model” and “option-pricing theory/formula” interchangeably.
Before we introduce the concept of option pricing applicable to the equity valuation, we
would like to answer the question “Why the option pricing model can be applicable in such cases as
corporate valuation”. The classic model of free cash flows discounting as a method of asset
valuation has its own limits: when the cash flows are negative for a long period of time the company
can still have positive economic value, but according to free cash flow method, it should not. So,
discounting of future or expected cash flows can lead to inadequate valuation in some cases. Option
pricing can solve this problem. In fact, option pricing of distressed companies as a method has an
advantage over a DCF method. By distressed companies we imply those with high leverage ratio
and negative cash flows.
Investors in equity of distressed firms 5 have the call option on liquidation and paying the
debts. Such call option with stike of the debt amount is able to enlarge the value of equity: for
example, in case of big uncertainty about the value of assets.
This phenomenon, equity value reflection in call option, is possible due to two characteristics
of the equity of public companies.
First, investors in the equity or shareholders are able to manage the company and anytime
can make a decision of selling the assets out and repaying the debt obligations.
Second, shareholders in public companies have limited liabilities. Their liabilities cannot
exceed the amount of their investment in the company (the size/amount of equity). So, if the
company goes bankrupt, the shareholders will cover the debt only by the amount of the equity and
thus they do not risk by their personal welfare.
Such combination of the option of liquidation of the company and limited liabilities of the
shareholders gives the equity of the features of call option. From the perspective of option theory we
can derive the value factors of the equity. Below we will shortly summarize them.
5 Basically, the theory can be applicable on any firm, distressed or not, from the methodology perspective we
will use original conceptual link between option pricing and distressed companies as Brealey, Myers, and Allen (2011)
and Damodaran (2008) do
30
Under the DCF method it can be argued that the company is worth nothing if its liabilities are
greater than its assets. But at the same time, the first conclusion from consideration of the equity a
call option is that equity has positive value even if the firm value is less than the nominal debt value.
Similarly, options that are deep "out of the money" (the price of underlying asset is much less
than the strike price) have value just for the reason of non-zero probability of underlying asset to
grow in price in the future – it is a time value of option. So, equity have its value for a time
determinant of option price, which is a time of corporate bond expiration, and the probability of
increasing of assets’ value more than the nominal value of the debt (bonds) before the maturity date.
Another interesting phenomenon of equity as a call option is the direct relation between the
risk (uncertainty) and the option (equity) price. In DCF method, abnormal risk leads to a reduced
cost of investing in equity. In option theory, when equity has the characteristics of the option, things
are opposite – the higher risk leads to benefits of equity investors. The fluctuation of the firm value
results in two variants for shareholders: either they lose the fix amount of their initial investment in
the equity or gain significantly because the upside brunch is unlimited.
Application of option theory to pricing the equity includes several assumptions (Damodaran,
2011):
1. There are only two types of claims to the company: stock and bond;
2. There is only one issue of debt (corporate bonds), which can be repurchased at its
nominal value before the maturity date;
3. The issued bonds have no coupon payments (zero-coupon bonds) and there are no
specific characteristics of the bond (such as convertibility, covenants, etc.);
4. The value of the firm and its dispersion are estimable.
Each of these restrictions or assumptions of the model have its reason. Mostly they are taken
just for the ease of calculating and maximizing the accuracy of the estimation under the option
pricing model.
So far we discussed the theoretical background behind the option pricing theory. Now we
would like to discuss how we are going to apply it in our paper.
Most firms do not fit the above mentioned severe restrictions, such as the presence of only a
single issue of bonds with zero coupons. So, certain compromises are necessary for proper use of
the option model for pricing the equity.
31
1. The firm value. There are 3 methods for obtaining the firm value:
a. To sum up the market values of the firms’ equity and debt. Then, the firm
should be public; its bonds should be tradable. In this case the option model
redistributes value of firm between debt and equity. This approach is simple in
its implementation, but it has internal contradictions: it starts with one
set of the market values of debt and equity and completes completely different
values of debt and equity as a result of option pricing.
b. To use DCF for proper firms’ assets pricing, as a discount rate WACC 6 should
be used. One important consideration here that we need to keep in
mind is that the value of company obtained by the option is the value that
shareholders will get after potential firm liquidation. That means that we
should only consider existing investments if we estimate the value of the
company using a DCF model.
c. Use value multiples. Need to consider the healthy firms in the same business
or industry, of comparable size and state of development.
For example,
applying the revenue multiple to the revenue of the target firm. Here the value
estimation is based on the implicit assumption that in case of the firm
liquidation the potential buyer will pay the exact amount of the calculated
estimation of the firm value.
2. The firm value volatility. There are several possible ways of calculating the firm
value dispersion:
a. To calculate it from the variants of stock and bond, if both are publicly
tradable.
Define the variance of stock as σ e2 and the variance of bond as σ d2, we need to use the
following formula:
2
2
2
2
2
❑
❑
❑
❑
σ firm =w e σ e +w d σ d +2 ρ ed w e w d σ e σ d
(1),
Where:
we is a weight of the market value of equity;
wd is a weight of the market value of bond issue;
ρed is the correlation between the prices of stock and bond;
6 WACC – weighted average cost of capital
32
When companies begin to experience financial difficulties, this approach can give incorrect
results, because the volatility of prices of stocks and bonds increases.
b. To use the industry average volatility of the firm values from one sector;
This approach compared to the previous one usually gives less biased estimations.
c. To use the variance of similar stocks and bonds with similar ratings on the
market if the firm is private and there is no public estimations of its stock and
bond values.
d. To use the historical approach and calculate the variance of the firm’s assets
and imply that the firm value variance is equal to the firm assets variance.
3. The maturity of the debt. The majority of firms has more than one issuing debt,
much of which goes with the coupon. Because the option pricing model allows only
one item of input data to time before the expiration of the term, we must transform
these several bond issues and coupon payments into one equivalent bond with a zero
coupon. Below the potential ways of such debt “transformation” are described:
a. Take into account both the coupon payments and the bonds maturity by
estimating the bonds’ duration and calculating the weighted average duration,
which is implied as the maturity term for the option;
b. Use the option model and weight the obtained estimation from the model by
the par value of duration of zero-coupon bonds (Damodaran, 2008).
4. The par value (nominal) of debt. In case of multiple bond issues there are 3 ways to
determine the nominal debt value for option model:
a. To sum up the par value of all bond issues and consider this sum a nominal for
debt in option model. The restrictions of this approach is that we do not take
into account all debt service payments that company makes before the
maturity date in a form of coupon payments.
b. To sum up all the payments for the debt: both the par value and the coupon
payments within the whole period of bond term. By doing so, we are mixing
cash flows related to different time periods. Nevertheless, this way of
considering the interim interest payments is the easiest one.
c. To sum up the par value of all bond issues and consider this sum a nominal for
debt in option model, while the coupon payments for the bonds are determined
as percentage of the firm value and considered a dividend yield in a option
model. By doing so, we get the decreasing of the firm value by the amount the
annual coupon payments.
33
So far we discussed the questions “Why” and “How technically” we should use option
pricing model. Now we would like to concentrate on the question how the option theory relates to
the agency problem and the conflict of interests of stakeholders. We exploit the idea of conflict of
interests and base the possibility of welfare redistribution on that idea.
We will use option pricing model (OPM) and Merton model as interchangeable titles of the
method. Under the OPM the value of equity is a fair price of a call option with a strike equals to the
level of debt – the shareholders get the residual value after paying all creditors of the firm.
The formal representation of the model is as following:
CSN
tK
(,)(d)(),
eNd =d1 =
--r (T t )
(1)
1 2
ln(/K)(r/2S)() T t ++- s 2
s T -t
,
(2)
ddTt
1 s.
2 =--
(3)
where:
C(S,t) – the fair value of the European call option, which is estimation of the equity value;
S – the spot price of underlining asset, which is the current value of the firm;
N(X) – distribution function of standard normal distribution,
K – strike price of the option, which is a debt level;
r – risk-free rate, which is a 30 years T-bills yields;
T – t – time of option expiration;
Ϭ – return volatile of the underlining asset, which is the volatility of the firm value;
So, in our study option value C(S,t) is a value of equity of the acquiring firm, strike (K) is the
level of debt of the acquiring firm, and Ϭ is a volatility of the firm value (or its assets, depending on
the case).
34
By calculating the value of equity of the combined entity we can compare, we can derive
with the estimation of the company’s fair value from the perspective of OPT. Equity value can be as
greater as well as smaller of the initial number (the equity before the deal) – any difference may be
attributed to the wealth redistribution between shareholders and bondholders of the bidding firm.
For example, the shareholders have the opportunity to invest in the project with the negative
net present value. If the project is very risky it makes the standard deviation the firm's value to
increase. In this case the equity value calculated as a call option increases. Thus, equity increases,
the firm value decreases by the amount of the negative NPV of the project, while the debt value also
decreases. In other words, by investing in the risky projects (or participating in the risky M&A deal)
the shareholders can increase their wealth at the expense of debtholders of the firm (acquiring
company).
35
Chapter 3. Measuring the reaction
3.1 Data collection
In order to conduct the empirical research financial data is needed. We are trying to set a link
between M&A deal and the reaction of the debt market: namely, in what way the wealth of the
bondholders of the bidding firm is changing due to the announcement of the M&A deal.
Secondary financial data from financial databases is used. As the empirical work of this
research consists of two parts then the data collection is made for each part consequently.
First, we collected the data about all mergers and acquisitions from Thompson&Reuters
Eikon database for the period from 2000 till 2015. The further selection process is based on the
following criteria:
1.
2.
3.
4.
5.
6.
Region – USA
Deal type – merger, acquisition
Deal status – completed
Minimum deal value – $ 100 ml
Industry for both companies – Oil&Gas
Date of announcement is available
There are 1273 deal, which meet the abovementioned requirements.
For debt market reaction we need the data of corporate bond prices, issued in the period from
2002 to 2014. We use the CBONDS database. The selection process is based on the following
criteria:
1.
2.
3.
4.
5.
6.
Region – USA;
Industry – Oil&Gas;
Type of security – bond, eurobond;
Rating – the issue is rated by at least 1 rating agency;
Maturity – no more than 10 years;
Status of issue – in circulation, repaid;
There are 935 issues, which meet the abovementioned requirements.
Following the logic of the research, we created 2 groups of companies:
1. Companies participating in M&A and having bonds issued (102 companies);
2. Companies with bonds issued but not participating in any M&A during the period of
our analysis (833).
102 companies engaged in M&A deals in the period 2000-2015, but for the purpose of DD
method use we need to have a group of companies, M&A deals of which took place in one year. To
avoid the disturbing effects of financial crises 2007-2009 we were looking for announcements in
36
post crisis period 2010-2015 or precrisis period 2000-2006. This is done intentionally for less biased
results, but still the period 2007-2009 was not rejected from the analysis.
The most numerous group of companies participating in the M&A is that one of 2014 year. In
the table below those companies and the date of their M&A deal announcement are presented.
Table 3.1.1 The distribution of the M&A announcements of the oil&gas American firms
Firm’s name
QEP Resources
Diamondback Energy
DCP Midstream Partners
Gulfport Energy
Baker Hughes
Martin Midstream
Cimarex Energy
Devon Energy
Legacy Reserves
Rice Energy
Whiting Petroleum
SM Energy
Tesoro
Linn Energy
Murphy Oil
Boardwalk Pipeline
SandRidge Energy
Vanguard Natural Resources
National Oilwell Varco
Enterprise Products
EnLink Midstream Partners
BreitBurn Energy Partners
Forum Energy Tech
ONEOK Partners
Western Gas Partners
Halliburton
Paragon Offshore
Superior Energy Services
Memorial Production
Southwestern Energy
M&A announcement date
30.01.2014
18.02.2014
25.02.2014
19.03.2014
24.03.2014
05.05.2014
06.05.2014
06.05.2014
06.05.2014
07.07.2014
13.07.2014
29.07.2014
30.07.2014
04.08.2014
06.08.2014
03.09.2014
04.09.2014
16.09.2014
30.09.2014
01.10.2014
22.10.2014
24.10.2014
27.10.2014
27.10.2014
28.10.2014
14.11.2014
17.11.2014
11.12.2014
18.12.2014
23.12.2014
As we can from the table above, the earliest M&A deal announcement was on 31st of January
while the latest – on 23 rd of December. Thus, for the DD method the time periods will be determined
in relation to these dates: comparing of the bonds’ prices should be conduct for 2 time periods,
37
before 31st of January and after 23rd of December. In the relevant literature where abnormal returns
are found the estimation window for a potential rumor effect is 1-40 days, thus we should eliminate
this potential bias by excluding these days. Finally, we compare the bonds’ prices for 2 days:
20/11/2013 (40 days earlier the date of the first M&A announcement on 31st of January) and
02/02/2015 (40 days later the day of the last M&A announcement on 23 rd of December). The overall
time difference is 439 days.
Bond prices were taken from CBONDS database7. The prices of the bonds of two groups of
the companies and for 2 time points, 20/11/2013 and 02/02/2015, are provided in tables in
Appendices 1 and 2.
For Merton model we need collect the data for each company individually. We have 21 out of
26 companies from the list presented in the Table 3.1.1 above, because 5 companies announced
M&A deal and thus were included in DD dataset but have not participate in the deal yet and thus are
excluded from Merton model dataset.
Each piece of company specific data was gathered or calculated for two time periods, which
are exactly the same as those used in DD: 20/11/2013 and 02/02/2015. The following company
specific data was obtained:
1. Enterprise value (EV):
Directly gathered from Thompson & Reuters Eikon database for two time periods;
2. The par value of outstanding debt:
Balance value of outstanding debt is taken for 2014 and 2015, where the pre-merge
debt par value was equal to the balance debt value of 2014 and the for post-merge
debt par value the number is calculated as:
Debt par valuepost-merge = Debt par value2 0 1 4 + (Debt par value2015 –
Debt par
value2014)*439/365. So, by using this formula we make an assumption that the debt
value is linearly changing over the 2015 year;
3. Average duration of outstanding debt:
Calculated in basis of debt term structure of companies from Thompson & Reuters
Eikon databases;
4. Standard deviation of the firm value:
Calculated for each company on the basis of its belonging to a particular oil&gas
subindustry. We divided oil&gas industry into 4 subindustries: producing and
extraction (176 firms in the subindustry), oilfield equipment and services (81),
distribution (12), and integrated subindustry (26). For each subindustry we have
7 The internet site is cbonds.ru
38
estimations of industry averages of firm value standard deviations. These estimations
are calculated and regularly updated by Damodaran at his website8.
The subindustries mentioned above have the following averages of standard deviations of the
firm value9:
1.
2.
3.
4.
Oil&gas integrated: 43,13%;
Oil&gas production and exploration: 43,96%;
Oil&gas distribution: 24,60%;
Oilfield services/equipment: 50,06%.
In addition, non-company specific data was needed:
Risk-free rate: as a proxy the yield of 10 years T-bills in 2014 year was taken.
The data for Merton model is presented in Appendix 4.
3.2 Empirical analysis results
3.2.1 Bond market reaction
After thorough data extraction from Thompson database and CBONDS we face the problem
of companies’ likeness. Indeed, for DD method the groups of companies, on the basis of which the
differences are calculated, should be comparable. In our case, the company and bond issues should
possess certain characteristics.
For the companies these characteristics are:
1. Belonging to one industry – American oil&gas industry;
2. Credit rating: all companies have credit ratings from at least 1 credit rating agency;
For the bond issues such characteristics are:
1. The term of bonds: all bonds are 10 years ones;
2. All bonds have investment grades from at least 1 credit rating agency;
We neglect the differences of coupon payment.
Below we present the table of DD results. For simplicity we call the group of companies that
were engaged in M&A deal and had bond issue on a time of the deal announcement as
“M&A+bonds”, the group of companies, which were not engaged in any M&A deal during the
8 http://pages.stern.nyu.edu/~adamodar/
9 Estimations of Damodaran for the end of 2015 year
39
period of analysis, we call “Only bonds”. “M&A+bonds” group is our treatment (experimental
group), while “Only bonds” group is a control group.
Table 3.2.1.1 DD results: the averages of the bond prices
20/11/2013
02/02/2015
Change
Average bond price in
“M&A+bonds” group
101.93
97.52
-4,41
Average bond price in
“Only bonds” group
100,87
100,83
-0,04
Difference
-1,06
3,31
-4,37
The results are as follows:
The treatment group had higher group average price on 20/11/2013, before all M&A
deals announcements: 101,9 against 100,87 of the control group.
The treatment group had lower group average price on 02/02/2013, after all M&A
deals announcements: 97,52 against 100,83 of control group;
Even though both groups showed the downward trend in bond prices, if the control
group’s average price was pretty stable over a analysis period – only 0,04 drop –then
the treatment group’s averages plunged in a greater scale – 4,41 price drop over time;
Initial difference between the groups on 20/11/2013 were 1,06. Under an assumption of
parallel trend, which is essential for DD methodology, the difference should be the same over a
period of time without external interruption (experiment, treatment). So, we assume that if no M&A
occurred between 20/11/2013 and 02/02/2015 then the treatment group’s average price should be
just 1,06 higher than the control group’s average price, which was 100,83 on 02/02/2015. Thus, we
have calculated average for the treatment group 101,89 as if no M&A occurred and real average
97,52. The difference between these two numbers, empirical and implied ones, is 4,37 – this is a
treatment effect of the M&A deal announcement.
This means that the bonds of the companies, which were engaged in the M&A activity in
2014, were underperformed by 4,37 each. This is how bond market reacts to the deal:
1. The M&A deal can lead to a welfare redistribution: if the creditors (bondholders) lose
their welfare they will claim higher yields to compensate the ;
2. The higher yield drives the bond prices to decline;
3. The higher the welfare loses from the perspective of the bondholder, the higher the
yield claims and then the more significant a price drop.
40
The drop of price of the bonds can be explained by various factors. We do not have objective
to provide the full set of explanations, but we hypothesize that such bond price drop can testify the
welfare compensation for bondholders. They could compensate the lost welfare in three cases:
1. It corresponds the overall market trend: all bonds on the American market of oil&gas
companies could lose their value.
We believe that this is not the case because we have a control group where the price drop is
almost insignificant – only minus 0,04. So, the treatment effect is pretty significant, thus we reject
the first possible reason for bondholders compensation.
2. The M&A deal occurred to be value destroying one, both shareholders and
bondholders lose their value.
3. The shareholders captured the greater amount of value than the deal was able to
generate, as a result of these shareholders gain at the expenses of bondholders.
4. The shareholders gained at a higher extent than the bondholders did.
We cannot distinguish the last 3 variants at this point. To do so we apply option pricing
model further.
3.2.2 Wealth redistribution effect
After calculating the values of equity as a call option and the value of debt, we found the
differences in these values over a period of time from 20/11/2013 to 02/02/2015 (equity change and
debt change in the table below). The results are presented below.
Table 3.2.2.1 Results of Merton model application by companies
#
Company name
Bond price change
1
2
3
4
5
6
7
8
9
10
11
12
Baker Hughes
Boardwalk Pipeline
BreitBurn Energy Partners
Cimarex Energy
DCP Midstream Partners
Devon Energy
EnLink Midstream Partners
Enterprise Products
Gulfport Energy
Legacy Reserves
Linn Energy
Linn Energy 2
3,67%
-5,33%
-36,14%
-0,77%
2,73%
6,92%
1,10%
1,83%
-6,54%
7,30%
-22,13%
-22,13%
Equity value
change
-1,33%
2,07%
49,19%
0,11%
4,76%
-1,21%
2,82%
-2,16%
4,24%
-2,43%
2,38%
-0,40%
Debt value
change
-0,13%
0,00%
5,88%
0,00%
-0,44%
3,33%
0,02%
2,03%
0,07%
3,52%
-1,34%
-1,13%
41
13
14
15
16
17
18
19
20
21
22
Martin Midstream
Memorial Production
National Oilwell Varco
ONEOK Partners
Rice Energy
SM Energy
Southwestern Energy
Tesoro
Western Gas Partners
Whiting Petroleum
-5,85%
-12,17%
4,94%
0,51%
-3,09%
-7,75%
-5,33%
5,69%
-10,32%
-6,99%
-0,05%
0,42%
-2,28%
-4,03%
22,13%
-10,54%
5,64%
1,95%
2,79%
-39,54%
-4,17%
-1,36%
11,21%
-0,42%
6,55%
-0,30%
-4,42%
-0,65%
0,92%
-9,97%
Having the results of prices dynamics, we can compare them and find the cases of wealth
redistribution between shareholders and bondholders. According to option pricing theory, such
phenomenon of value stream from bondholders to shareholders or vice versa is possible in 2 cases:
1. Negative dynamics of debt value (debt value change in the table above is less than
zero) and positive dynamics of equity value (equity vale change is greater than zero):
this is the case for the companies 5, 11, 14, 19, and 20 from the table above;
2. The growth of debt value is relatively lower than the growth of equity value: this is
the case for the companies 2, 3, 4, 7, 9, 12, 13, 16, 17, and 21 from the table above.
Graph 3.2.2.1 Equity and debt values dynamics
Debt value dynamics
10.00%
5.00%
0.00%
-40.00% -30.00% -20.00% -10.00% 0.00% 10.00% 20.00% 30.00% 40.00% 50.00%
-5.00%
-10.00%
Equity value dynamics
On the graph above each blue dot represents the acquiring company’s equity and bet
characteristics. The dynamics of values are the change of the value occurred due to the M&A deal.
Graph 3.2.2.2 The scale of equity and debt values dynamics
42
21
19
17
15
13
11
9
7
5
3
-40.00% -30.00% -20.00% -10.00%
1
0.00%
Equity value dynamics
10.00% 20.00% 30.00% 40.00% 50.00%
debt value dynamics
On the graph above the wealth redistribution is easily noticeable: every case where the debt
grows slower than the equity. For each company of the sample such comparison can be made from
the graph above.
In total 15 companies out of 22 have the conditions for welfare redistribution from
bondholders to shareholders. Creditors who are losing their wealth response to that by increasing the
required yield of return. The yields for creditors and the bond price have an inverse relationship.
Thus, increasing yields for bondholders lead to decreasing of bond price. We can check whether this
effect appears in our data. The data of 11 companies (2, 3, 4, 9, 11, 12, 13, 14, 17, 19, and 21 in the
table above) shows the existence of this effect: bond prices dropped meaning that bondholders yields
increased, while at the same time bondholders wealth estimated with Merton model decreased. So,
bondholders react and try to compensate their loses in 73,3% cases (11/15) by asking for higher
yields, which lead to the price drop of bonds.
On the other hand, there are 15 companies of the whole group, the bonds of which showed
the reduction of price for the period of analysis. As we just discussed above, we found 11 companies
with wealth redistribution effect. Thus, we can conclude that in 84,6% of cases of corporate bond
price reduction the wealth redistribution occurs and may be the main reason of such price drop.
Let us interpret the results on the case of one of those 11 companies with wealth
redistribution effect. Southwestern Energy (19th in the table above) is a growing independent energy
company primarily engaged in natural gas and crude oil exploration, development and production.
The initial data as inputs in Merton model are presented below:
43
Before the deal
Enterprise value (in option pricing model - Stock price, S 0) is equal to 12 002 $ mln;
Outstanding debt par value (Strike price, K) is equal to 6 967 $ mln;
Average duration of outstanding debt (option expiration period, t) is 3 years;
After the deal
Enterprise value (in option pricing model - Stock price, S0) is equal to 12 104 $ mln;
Outstanding debt par value (Strike price, K) is equal to 6 552 $ mln;
Average duration of outstanding debt (option expiration period, t) is 3 years;
The calculation of option gave the following results:
1. The option price before the deal C(S0,t)=4 681,66 $ mln, which is the value of equity
at time 5/09/2014
2. The option price after the deal C(S 1,t)=4 311,6 $ mln, which is the value of equity at
time 20/11/2014.
Below the table of results for Southwestern Energy is presented.
Table 3.2.2.2 Results summary for Southwestern Energy
5/09/2014
20/11/2014
Change
Value of equity as a
call, $ mln
6 287,9
6 642,6
354,7 (5,64%)
Value of outstanding
debt, $ mln
5 713,6
5 461,3
-252,3 (-4,42%)
Bond price, $
119,78
113,40
-6,38 (-5,63%)
Here we see that shareholders gain, while the creditors lose. Because we attribute these
wealth fluctuations to a certain event – M&A deal in which Southwestern Energy participated as an
acquirer firm – we argue that that M&A deal led to the benefits of shareholders at the expense of the
bondholders. The gain of shareholders due to the deal is greater in absolute than the loss of the
bondholders: 354,7 against -252,3. That means that shareholders also captured the full amount of
synergy value created due to acquisition. So, here we see the wealth redistribution effect:
simultaneous impoverishment of bondholders and enrichment of shareholders. The former
responsed to that by claiming for higher yields for the bonds. Higher yields made the bond price
decrease from 119,78 before the deal to 113,40 after the deal: this is the way how bondholders react
and try to compensate their loses.
44
3.2.3 Factors influencing wealth redistribution
So far we have estimated the effect of bondholders reaction to M&A deals: DD approach
showed us that on average bonds from the treatment group (“M&A and bonds”) decrease in price
from 101,93 to 97,52 which gives us –4,41 change. At the same time, the bonds from the control
group (“Only bonds”) on average stay almost at the same price level –100,87 before the deal against
100,83 after the deal, which gives us –0,04 change.
Then concentrate our attention to the treatment group (“M&A+bonds”). At this stage, we
used Merton model to calculate the equity value as a call option and estimate debt value for each
company from the group individually and for two periods – before and after the deal completed.
There are 22 out of 26 companies left in the dataset, because 4 companies announced the M&A deal
and thus were included in the treatment group for DD analyses, but these 4 companies still have not
completed the deal, thus we cannot apply Merton model for them and they excluded from the
dataset. 13 out of 22 companies experienced bond price reduction for the period of analyses (2014
year): 11 out of these 13 companies (84,6%) showed the effect of wealth redistribution from
creditors to shareholders.
According to option pricing theory, if managers decide to invest in risky project they can earn
money for shareholders even in case of negative NPV project – in this case wealth redistribution
arises and shareholders gain at the expenses of creditors, while the overall value of the firm reduces
for the amount of negative NPV of the project. The reason for this is added risk of the project: it
increases the firm’s cash flows volatility10, which leads to the equity value increase from the
perspective of pricing the equity as a call option (Merton model).
In our analysis, all firms are from one industry, but still there are discrepancies in the risk
profiles for several reasons:
1. Different subindustries;
Oil&gas industry is traditionally divided into extraction, distribution, oil&gas
equipment and services, and integrated subindustry (includes the companies that participated
in every step of value chain). The volatility of the firms from different subindustries of
oil&gas industry significantly differs from each other (see the data provided in data
collection section). Thus, if the acquirer buys the company from different subindustry and the
10 In fact, volatility of the firm’s cash flows is the same as the volatility of the firm value
45
average firm value volatility of target’ subindustry is higher than that of acquirer’s
subindustry, then this deal is risky for the acquirer: more volatile assets of target make the
combined firm’s assets also more volatile compared to the volatility of the acquirer value.
6 companies (2, 4, 9, 11, 14, 21 in the table 3.2.2 above) out of 11, which experience
wealth redistribution effect, have acquired the firms from different subindustry with higher
average firm value volatility. It led to a higher firm value volatility of the acquirer itself.
2. Different capital structure.
The acquirer and the target may have different debt structures. If the company
acquires the target with significantly greater leverage ratio, then the credit risk increases for
the acquirer. 4 companies (3,12, 13, 17) acquired the firms with higher leverage, which
increased the risk and thus increased the combined firm value volatility.
Higher risk profile of acquirer as a result of M&A transaction makes the creditors of the
acquirer impose higher yields. This corresponds with the option theory and Merton model: option
value increases with the rise of the volatility of the underlying asset. In corporate world, such option
feature leads to a conclusion that the firm’s equity has a greater value in cases of high volatility
(risk) of the firm value. So, we found that 10 out of 11 companies with proved wealth redistribution
effect have had the risk profile increased due to M&A deal.
4.3 Managerial implications
This paper examines the debt market reaction and welfare redistribution between
shareholders and bondholders. So far, we found the effect of market reaction by using DD method,
and support the found reaction with the idea of welfare redistribution by using Merton model (equity
as a call option): shareholders of 84,6% of the companies, the bonds of which experienced the price
drop, relatively gained at the expense of bondholders. In other words, the hypothesis that wealth
redistribution from bondholders to shareholders leads the bondholders to compensate their losses
with higher yields fits 84,6% of cases.
The debt market reaction in a form of bond price reduction leads to increasing of yield to
maturity – the indication of intention of the debt market to compensate the losses, occurred during
the M&A deal. We hypothesize that such losing of value can be a signal of welfare redistribution:
wealth flow from bondholders to the shareholders.
46
For the managers our paper reveals the effect of bond price reduction due to M&A deal in the
oil&gas industry. We found that risk factor was essential in 91% cases of wealth redistribution: the
acquiring company increases its risk profile in M&A deal by participating in a relatively risky
project of acquiring the company with higher volatility of the firm value.
Managers should consider potential debt depreciation before the decision whether to enter the
M&A deal or not is made. Strategic management responsible for such high-level decision-making
needs to reconsider the valuation of the deal from the prospective of the consequences for
shareholders’ welfare: value destroying M&A as well as welfare outflow to bondholders are possible
ways of losing the wealth in the deal. At the same time, they need to estimate the riskiness of the
target – how it can influence the overall riskiness of the acquirer in terms of firm volatility. Risky
target can bring value to the acquirer's shareholders from bondholders through redistribution effect.
The debt market response to such effect results in decreasing of the bond market price – so, in
addition the company may repurchase the bond issue at a lower price to reduce the debt level and
financial burden in a form of coupon payments in the future.
For the creditors and the bondholders of acquiring firms in particular, it is useful to know the
existence of debt market reaction and to be ready to take part into decision-making process of M&A
negotiation. We found the debt market reaction, other researchers, for example Hilscher and SisliCiamarra (2013), found that the existence of creditors in the board of directors reduce the
probability of company’s engagement in value destroying M&A deals. In fact, a private bondholder
probably will not become a member of the board of directors just to influence the decision regarding
M&A deal, which is potentially beneficial only for shareholders and not for creditors. Still, the
knowledge of negative effects of M&A deal for the bondholders of the acquiring company may
prevent the potential creditors to buy the bonds of company that actively participate in acquisitions
of risky targets.
Creditors should always take into account the idea that value-destroying deals can be
purposefully taken by shareholders: redistribution of the wealth from bondholders to shareholders
can outweigh the losses from value distortion M&A.
3.4 Research limitations
This research has several objective imperfections, which limit the explanation power of the
results. Below we describe the limitations of our research conclusions.
47
First and foremost, the sample of the companies for DD analysis is not ideal from qualitative
and quantitative perspectives:
Qualitative flaws. For DD method the dataset should consists of relatively similar
objects. The similarity of the companies in our case was proven by: a) belonging to
oil&gas industry, b) the same credit rating of companies and bond issues, c) the same
bond issues maturity of 10 years. Still, there are factors that make the sample of
companies not ideally homogenous:
1. Different businesses within one industry: oil producing, oil distribution,
oil equipment producing, and oilfield services;
2. Different coupons;
3. Different level of business development: mature Vs developing
companies;
Quantitative flaw: the number of companies in the sample is small (26), thus the
results and conclusions can be biased and thus further research on the same topic and
with the same goal but on the different dataset is needed in order to check the validity
of the results obtained in this paper.
Second, the calculation of equity and debt values within the option pricing model is partially
based on the factors, which were estimated by us or taken from outside sources. These factors are:
The firm value volatility of the firms from dataset for Merton model: we take average
firm value volatility for different subindustries of oil&gas industry and by proper
weighting11 of these averages we calculate for each company individually; By proper
weighting of these factors, the total firm value volatility was obtained for each firm of
the sample individually. So, such industry averages also bias the results;
The par value of outstanding debt: this is a balance indicator, calculated at the end of
financial year, but the deals occurred within the 2014 year, thus we estimate the
interim outstanding debt values for Merton model by using the balance values;
The sample for Merton model is a part of dataset for DD method and is also short – only 22
companies. The calculations were done for each company from the sample individually, the results
on individual basis are not biased, however, the conclusions about the welfare redistribution effect
on the basis on 22 companies may be not accurate.
11 Weight of a particular oil&gas subindustry firm value volatility average is calculated according to the
company’s business operations: the weight of revenue attributable to the subindustry is the weight of this subindustry in
the total volatility of the firm value
48
Still, we believe that the quality of dataset, gained by rigorous elimination of the companies
from initial data set of 102 M&A deals, enables us to argue that the conclusions made from the
analysis of such short datasets are significant.
49
Conclusion
M&A activity sometimes results in losing value of the deal participants. There are many
papers analyzing the success factors of the deal. Such papers identify features of the deal, the
acquirer and the target, which can help to predict the value creation during the M&A deal. Apart
from studying the value formation in the deal, many researchers are interested in the market reaction
– how the markets, debt and stock ones, react to the deal. There are plenty papers about the stock
market reaction, but a limited number of debt marker reaction studies. The focus of this research
paper is the reaction of the bondholders of acquiring company to the M&A deal on American
oil&gas market. The main goal we set at the very beginning of the research is to analyze the debt
market reaction to the M&A deal and investigate it from the perspective of wealth redistribution
effect.
We have conducted two parts of analysis. First, we have estimated the acquirer bondholder’s
reaction to the M&A deal. Using DD methodology, we gathered data for two groups of the
companies: the control group of companies that did not participate in any M&A deal but had bond
issues and the treatment group of companies that participated in the deal as acquirer and had bond
issues. The period of analysis was 2014 year: we compare the bond prices before and after the M&A
deal. We found that on average the bond price from the treatment group fell in price at 4,41 points,
while the bond price from the control group fell insignificantly at 0,04 points. The companies from
both groups were relatively similar: all are from oil&gas industry, the credit rating and the tenor of
bonds are the same. In addition, DD method mitigates all exogenous factors such as market
exposure, external shocks, etc. Thus, we concluded that the bondholders react to M&A deal and the
reaction is measured as bond price change.
Second, we have identified the welfare effect in those companies of treatment group, the
bond’s prices of which showed downward trend. We used Merton model, which presents the equity
of the firm as a call option to the firm value with a strike equal to the level of outstanding debt. For
each company we calculated the option price individually. We found that 84% of the firms with
downward trend in bond’s price have the effect of welfare redistribution: the shareholders gain at the
expense of the bondholders or value loss of the shareholders is relatively smaller compared to that of
the bondholders.
50
So, negative trend in bond’s price of the companies from the treatment group was mostly
driven by 60% of the bonds, 84% of which are the bonds of companies with wealth redistribution
effect from the bondholders to the shareholders.
We analyzed the companies with proved wealth redistribution deeper and found that 91% of
these companies enhanced their risk profile due to M&A deal by acquiring target with either higher
volatility of the firm value or more leveraged capital structure. This additional risk made the
acquirer firm value more volatile and it led to higher equity value – this result directly corresponds
with the theory of option pricing: the higher the volatility of the underlying asset is the greater the
price of the option is. So, the risk factor is essential for wealth redistribution effect, which in turn
influence the debt market reaction to the M&A deal: because of losing the wealth in a welfare
redistribution process, the bondholders try to compensate their losses by claiming for higher yields,
this in turn leads to a lower prices of the bonds.
There are clear managerial implications for strategic management, which is responsible for
decision-making process regarding M&A activity of the firm. Managers should reconsider the
valuation of the deal from the prospective of the consequences for shareholders’ welfare: value
destroying M&A as well as welfare outflow to bondholders are possible ways of losing the wealth in
the deal. They always need to pay attention to riskiness of the deal – risky target in terms of its debt
structure or highly volatile cash flows may cause increase in the risk profile of the company. To
some extent, such risk adding is beneficial to shareholders because they can gain at the expense of
the bondholders exploiting the welfare redistribution effect.
Creditors should always take into account the idea that value-destroying deals can be
purposefully taken by managers in order to satisfy the shareholders: redistribution of the wealth
from bondholders to shareholders can outweigh the losses from value distortion M&A.
The results provided in this paper are obtained on the basis of American oil&gas market for
the year 2014. To prove the consistency of these results further studies are needed with different
sample and time of analysis.
51
References
Akerlof G.A., The Market for "Lemons": Quality Uncertainty and the Market Mechanism,
The Quarterly Journal of Economics, Vol. 84, No. 3. (Aug., 1970), pp. 488-500
Al-Khasawneh, J. A., & Essaddam, N. (2012). Market Reaction to the Merger
Announcements of US Banks: A Non-parametric X-Efficiency Framework. Global Finance
Journal, 23(3), 167-183;
Angrist, J. D.; Pischke, J. S. (2008). Mostly harmless econometrics: An empiricist's
companion. Princeton University Press
Berkovitch, E., & Narayanan, M. P. (1993). Motives for Takeovers: An Empirical
Investigation. Journal Of Financial & Quantitative Analysis, 28(3), 347-362
Bertrand, M., Duflo, E., & Mullainathan, S. (2004). How Much Should We Trust
Differences-in-Differences Estimates?. Quarterly Journal Of Economics, 119(1), 249-275
Bouzgarrou, Houssam, and Wael Louhichi, "Does the Financing Decision Help to
Understand Market Reaction around Mergers and Acquisitions?" Journal Of Applied
Business Research 30, no. 2: 465-478, 2014;
Bruner R (2002), “Does M&A Pay?” Journal of Applied Corporate Finance, Vol. 12, No. 1,
pp. 48-68.
Damodaran, A. (2012). Investment valuation: Tools and techniques for determining the value
of any asset (Vol. 666). John Wiley & Sons;
Datta, D. K., Pinches, G. E., & Narayanan, V. K. (1992). FACTORS INFLUENCING
WEALTH CREATION FROM MERGERS AND ACQUISITIONS: A META-ANALYSIS.
Strategic Management Journal, 13(1), 67-84.
Fuller K, Netter J and Stegemoller M (2002), “What do Returns to Acquiring Firms Tell Us?
Evidence from Firms that Make Many Acquisitions”, Journal of Finance, Vol. 57, No. 4, pp.
1763-1793;
Hilscher, J., & Sisli-Ciamarra, E. (2013). Conflicts of Interest on Corporate Boards: The
Effect of Creditor-Directors on Acquisitions. Journal Of Corporate Finance, 19(1), 140-158;
James A D (2002), “The Strategic Management of Mergers and Acquisitions in the
Pharmaceutical Industry: Developing a Resource Based Perspective”, Technology Analysis
& Strategic Management, Vol. 14, No. 3, pp. 299-313.
Jansen, Ivo P. 2015. "Trading Volume around Announcements of Mergers and Acquisitions."
Applied Economics Letters 22, (13-15: 1204-1208, 2016;
Jansen, I. P., & Stuart, N. V. How to Predict the Market's Reaction When You Announce an
Acquisition. Journal Of Corporate Accounting & Finance (Wiley), 25(2), 43-49, 2014;
52
Jensen, Meckling, Theory of the firm: managerial behavior, agency cost and ownership
structure
Jansen, Sanning, & Stuart, 2013; Rani, Neelam, Surendra S. Yadav, and P. K. Jain. "Impact
of Mergers and Acquisitions on Shareholders’ Wealth in the Short Run: An Event Study
Approach." Vikalpa: The Journal For Decision Makers 40, no. 3 (July 2015): 293-312;
Kirchhoff, M., & Schiereck, D. (2011). Determinants of M&A Success in the Pharmaceutical
and Biotechnological Industry. IUP Journal Of Business Strategy, 8(1), 25-50;
Lukas, E., J.J. Reuer, and A. Welling, 2012. Earnout in mergers and acquisitions: A game-
theoretic option pricing approach, European Journal of Operational Research, 223, 256-263.
Lyroudi K, Lazaridis J and Subeniotis D (1999), “Impact of International Mergers and
Acquisitions on Shareholder’s Wealth: A European Perspective”, Journal of Financial
Management & Analysis, Vol. 12, No. 1, pp. 1-14.
Moeller S B, Schlingemann F P and Stulz R M (2005), “Wealth Destruction on a Massive
Scale? A Study of Acquiring-Firm Returns in the Recent Merger Wave”, Journal of Finance,
Vol. 60, No. 2, pp. 757-782.
Morck R, Schleifer A and Vishny R W (1990), “Do Managerial Objectives Drive Bad
Acquisitions?” Journal of Finance, Vol. 45, No. 1, pp. 31-48.
Myers, S., and N.S. Majluf, 1984. Corporate financing and investment decisions when firms
have information that investors do not have, Journal of Financial Economics 13, 187-221.
Officer, M.S., A.B. Poulsen, and M. Stegemoller, 2009. Target-firm information asymmetry
and acquirer returns, Review of Finance 13, 467-493
Palmucci, F., & Caruso, A. (2011). Measures of Market Reaction in Mergers and
Acquisitions in Italy. Corporate Ownership And Control, 8(2), 131-144;
Penas, M. F., & Unal, H. (2004). Gains in bank mergers: Evidence from the bond markets.
Journal Of Financial Economics, 74(1), 149-179
Roll, 1986: the corporate hypothesis of corporate takeaways;
Shah, P., & Arora, P. (2014). M&A Announcements and Their Effect on Return to
Shareholders: An Event Study. Accounting And Finance Research, 3(2), 170-190;
Shams, S. M., Gunasekarage, A., & Colombage, S. N. (2013). Does the Organisational Form
of the Target Influence Market Reaction to Acquisition Announcements? Australian
Evidence. Pacific-Basin Finance Journal, 2489-108;
53
Appendices
Appendix 1. The bond price of “M&A+bonds” group
#
Company name
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Baker Hughes
Boardwalk Pipeline
BreitBurn Energy Partners
Cimarex Energy
DCP Midstream Partners
Devon Energy
Diamondback Energy
EnLink Midstream Partners
Enterprise Products
Forum Energy Tech
Gulfport Energy
Halliburton
Legacy Reserves
Linn Energy
Martin Midstream
Memorial Production
Murphy Oil
National Oilwell Varco
ONEOK Partners
QEP Resources
Rice Energy
SM Energy
Southwestern Energy
Tesoro
Western Gas Partners
Whiting Petroleum
Bond price on
20.11.2013
100,42
111,99
101,00
105,75
101,72
95,17
106
114
111,21
104
107
119,1
27,96
101,06
102,5
102,75
99,77
92,45
104,72
106
97
108
119,78
97,5
110
103,22
Bond price on
01.02.2015
104,11
106,02
64,5
104,94
104,5
101,76
104,25
115,25
113,24
93,75
100
118,62
30
78,7
96,5
90,25
97,1
97,02
105,25
104,96
94
99,63
113,4
103,05
98,65
96
54
Appendix 2. The bond prices of “Only bonds” group
#
Company name
1
2
3
4
5
Air Products and Chemicals
Airgas
Albemarle
ATP Oil & Gas
Berry Plastics
Black Elk Energy Offshore
Operations
CenterPoint Energy Resources
CMS Energy
Commonwealth Edison
DuPont
Eastman Chemical
EOG
Hess
Hillenbrand
Huntsman International
Laclede Gas
Mosaic
Noble Corp
Phillips 66
PolyOne
Praxair
Rockwood Specialtie
RPM International
Unit Corp
Williams Cos
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Bond prices on
20.11.2013
92,64
95,73
104,86
65
100,19
Bond prices on
01.02.2015
99,58
101,71
100,03
66
102,76
93,38
82,56
107,53
108,74
101,63
105,45
102,01
117,32
99,88
106,66
98,96
98,93
99,7
104,17
101,62
98
105,83
102,75
114
105,25
91,73
111,94
115,18
107,17
107,98
102,14
115,36
92,87
108,1
77,75
101,74
109,73
88,32
108,83
102,62
110,41
105,62
116,7
94,25
91,48
55
Appendix 3. Inputs for Merton model: acquiring companies before the deal
Company
EV
outstanding DV
EV volatility
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
33 821 718 223
7 999 787 121
4 223 502 804
13 491 371 486
7 193 135 829
31 972 338 000
9 707 298 059
85 888 635 628
5 381 074 995
2 580 547 457
19 800 051 729
19 969 233 645
1 748 017 712
2 464 369 834
29 514 935 947
17 719 073 659
3 436 976 613
7 266 493 488
12 001 615 416
12 083 341 029
9 191 672 850
11 600 594 531
4 381 000 000
3 683 000
3 352 160 000
1 500 000 000
2 424 000 000
11 262 000 000
2 022 500 000
21 363 800 000
703 564 000
938 876 000
10 295 809 000
10 295 809 000
888 887 000
1 574 147 000
3 166 000 000
7 067 178 000
900 680 000
2 332 445 000
6 967 000 000
4 161 000 000
2 422 954 000
5 602 389 000
50,06%
24,60%
43,96%
43,96%
24,60%
43,96%
24,60%
33,87%
43,96%
43,13%
43,96%
43,96%
37,33%
43,96%
50,06%
43,13%
43,13%
43,96%
43,96%
43,96%
24,60%
43,96%
average
duration
7,44
3,91
4,44
4,5
4,48
7,63
7,84
7,1
4,24
4,06
3,98
3,99
3,5
5,42
7,47
7,2
4,81
5,27
3
4,33
6,56
3,67
56
Appendix 4. Inputs for Merton model: acquiring companies after the deal
Company
EV
outstanding DV
EV volatility
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
33 412 778 117
8 165 179 305
5 328 348 705
13 504 636 795
7 422 731 357
31 939 972 000
9 934 381 631
84 764 832 238
5 583 040 548
2 565 005 279
19 969 233 645
19 829 907 369
1 715 840 787
2 455 786 237
29 176 091 327
17 183 340 310
4 079 010 422
6 691 811 676
12 104 018 751
12 228 523 170
9 409 527 098
8 367 455 849
4 133 000 000
3 683 000
3 245 150 811
1 500 000 000
2 411 890 411
11 746 931 507
2 022 500 000
21 854 534 247
703 564 000
978 722 795
10 090 154 315
10 165 561 033
888 298 079
1 640 740 490
3 562 945 205
7 094 230 986
949 472 723
2 361 417 397
6 551 783 562
4 121 339 726
2 444 771 216
5 438 297 548
50,06%
24,60%
43,96%
43,96%
24,60%
43,96%
24,60%
33,87%
43,96%
43,13%
43,96%
43,96%
37,33%
43,96%
50,06%
43,13%
43,13%
43,96%
43,96%
43,96%
24,60%
43,96%
average
duration
7,44
3,91
4,44
4,5
4,48
7,63
7,84
7,1
4,24
4,06
3,98
3,99
3,5
5,42
7,47
7,2
4,81
5,27
3
4,33
6,56
3,67
57
Appendix 4. The pivot table of results+
Output
Bond prices dynamics
Equity value dynamics
Company
name
Bond price
before
M&A deal
after
change
(abs)
change
(rel)
before
after
1
100,42
104,11
3,69
3,67%
30 449 238
232
2
111,99
106,02
-5,97
-5,33%
7 996 441 809
3
101,00
64,50
-36,5
36,14%
1 977 622 285
4
105,75
104,94
-0,81
-0,77%
12 156 432
754
5
101,72
104,50
2,78
2,73%
5 029 403 846
$30 044 578
139,58
8 161 833
993
2 950 332
329
12 169 676
363
5 268 554
854
6
95,17
101,76
6,59
6,92%
24 152 334
402
7
106,00
104,25
-1,75
-1,65%
8
114,00
115,25
1,25
1,10%
8 043 979 866
9
111,21
113,24
2,03
1,83%
68 498 556
328
10
104,00
93,75
-10,25
-9,86%
11
107,00
100,00
-7
-6,54%
12
119,10
118,62
-0,48
-0,40%
13
27,96
30,00
2,04
7,30%
1 788 397 899
14
101,06
78,70
-22,36
14
101,06
78,70
-22,36
15
102,50
96,50
-6
11 6 8 6 0 2 4
570
11 9 6 3 7 8 8
437
988 361 827
16
102,75
90,25
-12,5
17
99,77
97,10
-2,67
22,13%
22,13%
-5,85%
12,17%
-2,68%
4 751 873 276
1 399 285 500
23 859 950
119
8 270 663
684
67 022 143
280
4 953 377
027
1 744 982
168
11 963 788
437
11 791 578
523
959 065 615
1 362 006
368
debt value dynamics
change
before
after
change
-1,33%
3 372 479 991
3 368 199 977
-0,13%
2,07%
3 345 312
3 345 312
0,00%
49,19%
2 245 880 519
2 378 016 376
5,88%
0,11%
1 334 938 732
1 334 960 432
0,00%
4,76%
2 163 731 983
2 154 176 503
-1,21%
7 820 003 598
8 080 021 881
3,33%
2,82%
1 663 318 193
1 663 717 947
0,02%
-2,16%
17 390 079 300
17 742 688 958
2,03%
4,24%
629 201 719
629 663 521
0,07%
-2,43%
792 149 558
820 023 111
3,52%
2,38%
8 114 027 159
8 005 445 208
-1,34%
-1,44%
8 005 445 208
8 038 328 846
0,41%
-2,96%
759 655 885
756 775 172
-0,38%
-2,66%
1 065 084 334
1 093 779 869
2,69%
0,44%
27 037 481
721
12 782 318
514
18
92,45
97,02
4,57
4,94%
19
104,72
105,25
0,53
0,51%
20
106,00
104,96
-1,04
-0,98%
21
97,00
94,00
-3
-3,09%
2 675 390 823
22
108,00
99,63
-8,37
-7,75%
5 396 591 516
23
119,78
113,40
-6,38
-5,33%
6 287 963 832
24
97,50
103,05
5,55
5,69%
8 613 590 585
25
110,00
98,65
-11,35
10,32%
7 137 596 625
26
103,22
96,00
-7,22
-6,99%
7 021 697 151
101,89
96,82
26 420 872
513
12 267 151
977
3 267
844
4 827
977
6 642
501
8 781
387
7 336
867
4 245
179
537
557
684
437
484
238
-2,28%
2 477 454 226
2 755 218 814
11,21%
-4,03%
4 936 755 145
4 916 188 333
-0,42%
22,13%
761 585 790
811 472 578
6,55%
-10,54%
1 869 901 972
1 864 253 699
-0,30%
5,64%
5 713 651 584
5 461 334 250
-4,42%
1,95%
3 469 750 444
3 447 085 783
-0,65%
2,79%
2 054 076 225
2 073 042 231
0,92%
-39,54%
4 578 897 380
4 122 217 670
-9,97%
59
Отзывы:
Авторизуйтесь, чтобы оставить отзыв