The motives of the separation of companies and the determinants of their effectiveness: a comparative analysis of developed and developing markets

Размещено на http://www.allbest.ru/

Федеральное государственное автономное образовательное учреждение высшего профессионального образования "Национальный исследовательский университет "Высшая школа экономики"

Факультет экономических наук

Образовательная программа "Экономика"

БАКАЛАВРСКАЯ ВЫПУСКНАЯ РАБОТА

Мотивы разделения компаний и детерминанты их эффективности: сравнительный анализ развитых и развивающихся рынков

Выполнил Ильюшкин Феофан Анатольевич

Научный руководитель К .э. н., преподаватель

Партин Илья Маркович

Москва

2016



Table of contents

Introduction

1. Theoretical background of studying demergers

1.1 Description of a demerger as a measure of corporate restructuring

1.2 Motives of demergers

1.3 Demerger efficiency and its determinants

2. Methodological issues of studying demergers

2.1 Determining demerger efficiency: major approaches

2.2 Quantification of demerger efficiency determinants

3. Empirical study of corporate demergers determinants

3.1 Hypotheses

3.2 Variables and model

3.3 Data description

3.4 Results

3.5 Limitations and future research

Conclusion

References

Appendix 1. List of spin-offs analyzed



Introduction

A widely studied in literature and successfully applied in practice life cycle theory suggests that as a firm is ageing and its organic growth opportunities are becoming less apparent and less profitable, in order to sustain growth and not to lose appeal to its stakeholders, the company has to take some restructuring steps. These steps are not usually concerned with the nature of the company's business, but are rather organizational, however they can be a rather powerful tool in solving strategic corporate problems.

One of these measures is aimed and expanding the business by buying other companies - i.e. conducting M&A processes. While highly popular both in practice and research literature, this kind of "maturity crisis" disease aid turns out to be relatively unsuccessful. Modern studies (Christensen et al., 2011) show that up to 70% M&A's fail due to some obstacles on the way to corporate integration.

The opposite of acquiring new companies in order to expand is divesting from some assets and splitting the initial company into pieces - demerging. Counterintuitively, but this kind of measures is more often associated with increasing corporate efficiency and improving shareholder value. Demergers or spin-offs are associated with turning a company's subsidiary or set of assets into a separate independent entity and distributing the shares of the new company among the shareholders at pro rata basis.

As a measure of corporate restructuring, demergers originated in the US in the 1st half of XX century. The first article dedicated to demerger specificity was Simon (1960). For a long time, they had only been a solely American phenomenon. However since the 1970's they have become increasingly popular in Europe and emerging markets - especially in China, India, Malaysia and South Africa. Mergermarket.com statistics show that over the last 10 years, the market for spin-offs in emerging markets rose from $15 billion to $45 billion with more than a half the market being allocated in China and India. Nevertheless, the literature on explaining demerger efficiency and its determinants by far stays quite poor, has certain contradictions, disagreements between different academics and a lot of blank spots to be filled in by researchers. Besides, a vast majority of the articles are concerned with American reality and developed markets in general. There are just a few researches on emerging markets or cross-country empirical investigation of demerger efficiency determinants.

Taking all of that into account, our paper provides its own insights into the motives of demergers, their efficiency and its determinants largely based on a compilation of existing body of scientific financial literature in this field, but also testing the conclusions against cross-country data. The ultimate purpose of our research is identifying what factors drive demerger efficiency the tasks solved in order to reach this goal include:

· examining existing research literature and real cases to distinguish possible motives for companies in developed and emerging markets to demerge;

· identifying key approaches to determining demerger efficiency employed by researchers;

· formulating and explaining major hypothesis in explaining demerger efficiency roots

· providing empirical testing for the formulated hypotheses

Accordingly, the paper is structured as follows. The first chapter deals with theoretical aspects of demerger process including the details of the process itself, its motives and main efficiency explaining models described in the literature. The second chapter is devoted to methodological issues of empirical demerger efficiency study and contains descriptions of the ways to quantify both demerger efficiency proxies and its determinants including pros and cons of each of the approaches. Finally, the third chapter provides testing existing demerger efficiency hypotheses against factual data and boils down to formulating empirical hypotheses, analyzing a data sample of demerger cases and making conclusions based on empirical results.



1. Theoretical background of studying demergers

1.1 Description of a demerger as a measure of corporate restructuring

Many companies adhere to corporate restructuring in order to improve their operational metrics, increase financial stability and overall create some extra shareholder value. A restructuring of any kind is basically a set of measures intended to benefit the company and its stakeholders without tackling with the nature of the business and intervening its very operational sides directly or, to put it another way, to develop the company non-organically. A restructuring usually takes place at stagnation stage of firm's lifecycle (Ivashkovskaya & Yangel, 2007) since any kind of restructuring is quite expensive in principle, but at that point of the company's development there are only a few ways for the company to move forward.

Corporate restructuring can take a variety of forms and all of them can be classified into several groups (Bowman & Singh, 1993):

· Expansion. A non-organic expansion of a company is associated with enlarging the company by investing in outside growth opportunities. This can include standard mergers and acquisitions and also forming joint-ventures or strategic alliances with other firms. Being non-ogranic, this kind of expansion enables the company to turn its financial resources into growth factors more swiftly than it could given that it used them to invest in its own projects;

· Ownership change. In order improve financial position, the company can sell shares in its subsidiary as a way of capital raising. This set of measures also includes share repurchases as they change the overall balance of powers in corporate ownership;

· Divestures. Instead of expanding its current business, the company can divest some of its assets in order to pursue several goals. A divesture can result in a total sell-off of the targeted asset or transferring the rights of ownership to the company's shareholders or other affiliates at a non-cash basis - a spin-off or a demerger. The difference between divestures and equity carve-outs is that a divesture requires the majority of the asset to be spun-off or sold-off

A demerger is a non-cash corporate divesture which is realized in a form of granting the ownership of a company's assets to its shareholders or other affiliates. After the demerger is completed, the shareholder of the parent company end up with shares of both the initial legal entity and it spun-off subsidiary. In case the company is public, its shares also have to be listed in a stock exchange in order for the minority shareholders to receive their stakes in the newly created firm.

The process of a typical demerger can take from weeks to months and even years and can be broken down to the following stages (Thomasson & Janusonis, 2012):

· Pre-demerger stage. This study defines pre-demerger stage as the period of time between starting an internal discussion of whether to demerge or not and publicly announcing the restructuring. The pre-demerger stage is crucial for the parent firm in terms of making the decision and calculating its future cost and benefits

· Execution stage. After the announcement and before the new entity is finally formed, its shares are transferred to the shareholders and the public listing procedure is gone through - i.e. the deal's completion, the demerging firm has to undergo a series of formal procedures including numerous shareholder and regulator approvals, legal and tax audit, finalizing the deal structure, etc. This stage can take from several days to years - sometimes, after the announcement the demerger lapses after having received a refusal from the parent company's shareholders or the authorities. For example, in 2009 Liavant Hall Corp. spin-off from Li Heng Chemical Fibre Technologies was not approved by the exchange;

· Post-demerger stage. In case the deal execution runs smoothly, the spin-off starts trading in a stock exchange and this becomes the finish of the execution stage. Afterwards several probable outcomes can take place. In the majority of cases, both the parent company and its spin-off continue their normal activity as two separate firms. Another option for the future development is an acquisition of one of the post-demerger entities by a strategic buyer. However sometimes in a short period following the completion, the spin-off can be reacquired by its parent or even the opposite - the spin-off can take over its parent. While the former case is usually a way of tackling an unfavorable or successful demerger, the latter is in most cases a planned act (Gaughan, 2007).

Now, having in mind demerger structure and its place in the universe of corporate restructuring measures, we can move on to its motives and efficiency.

1.2 Motives of demergers

Why do companies decide to split their businesses up? The motives of corporate restructuring in fact can be very diverse in terms of their nature, however in general most of the restructuring measures taken are aimed at enhancing the company's development and increasing shareholder value. In case of demergers, there exist several classifications of their motives, but we would suggest our own classification based on the dual substance of the matter. Since a typical demerger involves both a disposal of some assets and a subsequent IPO, according to our literature study (See Carhart, 1997; Danielova, 2008; Furlan & Grandinetti, 2014) and press releases of the demerging entities, all the demerger motives can be categorized into asset relocation related, IPO related and specific motives. Let us take each of the groups into a closer consideration.

Asset relocation motives

Asset relocation motives are associated with the fact that during a demerger process some of initial company's assets are transferred into a newly created entity or the shares in some of the company's subsidiaries are moved into the hands of the company's shareholders.

The first and foremost relocation related reason mentioned by researchers studying demerger process is increasing focus of the business (See Comment & Jarrell, 1995; Daley et al., 1997; Desai & Jain, 1999). As it was previously mentioned, restructuring measures are most commonly applied when the company is at the maturity stage of its life cycle. Evidently, by this stage of development some large companies overgrow with numerous divisions, subdivisions and subsidiaries that emerge from its urge for vertical integration, but eventually harm the core business since they require extra expertise from the general managers. The recent case of Metro AG split into Food specialist and Consumer electronics entities announced on March 30 this year which is, according to the press release, intended to increase their focus, flexibility and operational efficiency, serves as a perfect example of such motive.

This is why, as studies reveal, the majority of spin-offs are expected to serve the sole purpose of letting each of the business units have its own development strategy and pursue its own goals eventually having separate management systems, corporate governance and ownership structure. However, some studies argue that a sell-off might be a more preferable option for a parent company working in a virtually unrelated area than the spin-off. For instance, in Bergh et al. (2008) it is established that, when the business areas of the companies are unrelated, the parent is more likely to sell its unit out than to conduct a demerger. This could be attributed to the urge of the shareholders to dispose of the assets not in line with their investment strategy. However we did not find any other proof of this kind of shareholder preferences.

Another relocation related motive, which is very similar to focus increasing, but still is not the same one is called "fixing the mistake hypothesis". It was developed by Allen et al. (1995). The researchers capitalize on the fact that sometimes companies spin-off their business units that were previously bought, but turned out to be unfavorable, value decreasing acquisitions. Unfortunately, most of the M&A deals appear to be value destroying. Christensen et al. (2011) reports that 70% of mergers end up hurting both companies engaged. Cutting value may be attributed to a variety of reasons such as corporate culture mismatch, unsuccessful implementation of integration plans, ownership related issues, etc. Hence in a short period of time a wise decision for the company's shareholders could be a separation of the merged entities in order to at least partially restore the destroyed value.

But even though mergers and acquisitions are reposted to be unsuccessful in 70% of cases, only a small portion of corporations actually try to spin-off its newly acquired divisions. That could be a consequence of the fact that a demerger tend to be a very costly, long and disturbing process and therefore not all of the companies that experienced repercussions of an adverse acquisition are ready to undergo a spin-off process in order to increase the shareholder value back.

Another asset relocation reason for a company to demerge is for its shareholders to remove some of the most valuable assets to a new entity so that the in case of a bankruptcy they would not be expropriated in favor of the creditors. Since this motive is unlikely to be explicitly announced due to its unfairness towards the debt holders, it is only studied to the extent of how certain capital structure parameters (See Mehrotra, Mikkelson & Partch, 2003; Maxwell & Rao, 2003; Dasilas et al., 2011), however the motive as it is still makes sense. We will demonstrate how it can be beneficial for the shareholders to transfer some of their wealth into a separate debt-free entity.

And finally, a spin-off can serve as a powerful anti-takeover tool. When a hostile takeover is taking place, the predators are usually targeting some particular strengths of the company that can particularly be employed in the future merged firm. A perfect example of such a demerger is the split of Diamond Shamrock R&M from Diamond Shamrock in 1987 (Gaughan, 2007). However, when the current target shareholders decide to split these particular strengths of the company in a form of concrete assets - "crown jewels" - its value for the predator becomes far less and the takeover could be avoided. But even though this kind of spin-off can serve as an effective anti-takeover provision, deliberately splitting some of the assets can be costly in terms of both the actual legal and organizational expenses and the loss in shareholder value. That is why demergers are very rarely used as an anti-takeover provision.

IPO-related motives

While the first group of demerger motives are associated with turning the parent company better-off directly, the following reasons are all connected to the fate of the newly commenced company since after the demerger is conducted, the spin-off starts trading as a separate public company.

The first reason for a company to spin-off its division into a separately traded entity is attracting additional capital. As it is noted in Danielova (2008), listing one of the company's divisions separately and subsequently selling a part in the new company to the public can bring additional capital for investment purposes.

Another motive of demergers connected with their IPO nature is that in case the spun-off business unit does not participate in the parent's core operations, such division can make it easier for the companies to attract certain types of investors according to their industry preferences and risk aversion (Abarbanell et al., 2003). For example, if a company operating in oil and gas field spins off its machinery manufacturing division, investors what do not have any appetite for oil and gas sector, but are keen on heavy machinery and high tech are going to be more eager to invest in the spin-off stock even though they did not prefer parent's.

And last but not least motive on this side is increasing the transparency of the business (Khorana et al., 2011). This stems from the fact that in a larger and a more intricate corporation it is less obvious for external observers how the cash flows are generated, why the management chooses a certain strategy and what the perspectives of the business are. Under the circumstances of a spin-off, the investors become more aware of each company's financial situation and its management actions and thus can evaluate each investment opportunity more properly. Due to this reduction of information asymmetry, after a spin-off both companies can become less risky and draw more favorable market valuations.

Specific motives

Whereas the two abovementioned groups of demerger motives can be considered as finance-related, other motives are more associated with non-financial and even non-operational side of the businesses and thus each case when such a motive turns into reality should be investigated separately.

The first reason from this group is legal obligation to split the business. Sometimes when a company acts as a monopoly or holds a very large market share so that the antitrust authorities see it as a threat to competition and price stability in the market, they can legally oblige it to split. To this extent, the most widely known case would be the segregation of AT&T corporation in the United States in 1984 (See Gaughan, 1999). When the authorities received a formal complaint from the antitrust regulators, the company was forced to split 7 newly formed firms. Despite the fact that the action was intended to improve the position of AT&T's customers, its brought about a large amount of legal work for separation of the entities, thus overall decreasing the operational efficiency of the AT&T.

Another specific motive of splitting a company into pieces is associated with relations between the key owners of the firm. A split can be necessary because of the family-related issues. For instance, in Ramu (1999) it is well-noted that in case the owner dies and each of the heirs is willing to run a company of their own, a legal and financial separation of the companies will be preferable even regardless of any financial matters associated with the deal. The author gives an example of Indian capital market where due to the dominance of family firms, family-related demergers account for a significant portion of the spin-offs market. The same can apply to break-ups between husbands and wives.

Having discusses the three major groups of demerger motives, we are now moving on to an overview of its efficiency and factors that drive this efficiency.

1.3 Demerger efficiency and its determinants

Under the scope of this study, we define demerger efficiency as the influence that a spin-off exerts on financial and operational metrics of the group of companies - the parent firm and the spin-off - jointly. This formal approach is chosen because of two reasons. Firstly, according to our literature review, this is the approach that most researchers adhere to. Secondly we personally see it as the most sensible one compared to judging about demerger efficiency by separate influences on the parent and the spin-off since the ultimate goal of a voluntary demerger is to increase shareholder value, but this value change can only be captured when comparing the joint group of companies to its predecessor.

There are basically three approaches to quantifying demerger efficiency: event-study approach, financial reports approach and survey approach. All the three are going to be discussed in more detail in Chapter 2, but the choice of the proxy for demerger efficiency in fact matters while determining its determinants. Nevertheless, unlike M&A that do not gain any positive results in 70% cases, regardless of the chosen proxy for demerger efficiency, most of demergers in fact have a positive influence on the company's performance. The most crucial task for a researcher is to realize what drives the performance and how to manage it.

All the existing body of scientific literature on the determinants of spin-off efficiency includes 6 major hypotheses explaining the positive results of demergers. Let us take each of them into a deeper consideration.

Corporate Focus Hypothesis

As its was already mentioned, one of the key motives for a company to spin its part off is to increase corporate focus. The logic behind this concept is that increasing focus enables each unit to form an independent management system, act separately and therefore enhance its operational performance. At some extent, the corporate focus hypothesis has its reflection in every single more or less descent research of spin-off efficiency, but arguably its most noticeable proponents are Comment & Jarrell (1995), Berger & Ofek (1995), Desai & Jain (1999). In these articles, the authors investigate the effects of focus increasing on the efficiency of a spin-off. In general, they all report a positive linkage between how distinct the businesses of the parent company an the spin-off are. In addition, Daley et al. (1997) reveals that companies working in a different industry than their spin-offs benefit more from the demerger than related industries ones. Moreover, they document that only focus-increasing demergers eventually lead to significant improvements in terms of financial performance and profitability metrics.

Most of the recent studies concerned with demerger efficiency determinants consider deal with industry focus hypothesis and find supporting evidence for it in emerging markets: for example, Uddin (2010), Sudarsanam & Qian (2007),

Another aspect of business focus increase is rising geographical concentration. For example, Harris & Glegg (2008) considers spin-offs increasing geographical focus and finds out that the efficiency of such spin-offs is largely dependent on the differences between institutional environments of the parent and spin-off countries.

Managerial incentives hypothesis

Apart from enabling the management to concentrate on their specific business models and pursuing individual business goals, demergers can be a way to establish a different incentive system for divisional managers and mitigating agency problem. While Shipper & Smith (1983) suggests that a spin-off can give the shareholders an opportunity to improve monitoring process, a theoretical model of Aron (1991) suggests that a significant post-spinoff efficiency improvement can be achieved through an introduction of a new compensation system. An empirical confirmation of the hypothesis is demonstrated by Pyo (2006) which shows a role of managerial compensation packages both as a motive to spin-off and a determinant of its operational efficiency.

The most recent studies of managerial compensation to the extent of its influence of spin-off performance includes Feng et al. (2015). A curious finding of that article is the fact that high incentives firms yield lower announcement abnormal stock returns, however enjoy higher spin-off achievements in the long run.

Information asymmetry hypothesis

Taking into consideration the same treat of demerger process as the focus hypothesis - decreasing corporate complexity - information asymmetry hypothesis capitalizes on the fact that large conglomerates engaged in various types of business usually lack transparency and thus appear to be undervalued. A demerger is unlock the potential value of the whole company by exposing its processes to the investors in more detail. Such authors as for instance Habib et al. (1997) and Nanda & Narayanan (1999) develop theoretical models that predict a potential wealth effect from decreasing information asymmetry. Krishnaswami & Subramaniam (1999) report that a post-spinoff set of entities is more transparent since analysts tend to forecast their earnings more correctly and the dispersion of the forecast significantly decreases after the spin-off. Bergh et al. (2008), having compared spin-offs and sell-offs as types of corporate restructuring, come to the conclusion that spin-offs are more preferable than divestures in case the initial company is undervalued. Khorana et al. (2011) and Huson & McKinnon (2003) come to similar conclusions.

Wealth transfer hypothesis

This hypothesis stems directly from the asset relocation motive described in 1.2. In a nutshell, the hypothesis boils down to the following statement: a disproportionate distribution of debt between the post-spinoff firms can lead to gains for shareholders and losses for debtholders. Therefore a demerger can have extra efficiency for the shareholders compensated by losses of the creditors. Firstly, this effect can be brought about by discrepancies in the leverages of the post-spinoff firms that result in the creditors of usually the parent company being less secured with collateral and thus the shareholders saving a portion of the assets in a newly created firm (See Maxwell & Rao, 2003). Then, even if the leverages remain at the pre-demerger level, a wealth shift can occur due to the differences in debt coverage ratios of the parent and the spin-off (See Mehrotra, Mikkelson & Partch, 2003) which also leads to higher default probability and consequently the cost of financial distress for the parent company and lower - for the spin-off. Finally, according to Maxwell & Rao (2003), the parent company's debt can become riskier because of the disappearing diversification whereas the stock remains diversified since the shareholders remain in possession of both the parent and the spin-off.

However we see an upside of the wealth transfer hypothesis dragging shareholder benefits down. It is largely associated with the concept of cost of financial distress. While some of the assets are drained from the parent company into a new entity, a significant share of assets still remains in the parent firm and in case none of the initial company's outstanding debt contracts are not renegotiated to address the spin-off, the parent company can experience debt overhang problems described by Myers (1984). Thus the cost of financial distress can outweigh shareholder gains from wealth expropriation. Moreover, highly leveraged companies can attract extra interest from creditors and thus splitting assets can become impossible due to creditors' restrictions.



2. Methodological issues of studying demergers

2.1 Determining demerger efficiency: major approaches

In this section, we study different approaches to the quantification of demerger efficiency and decide which approach is the most applicable in our own empirical investigation of demerger efficiency factors.

Management surveys

Although using management surveys in assessing demerger efficiency is not as widespread as event studies and even financial reports analysis, the method still has its implications in scientific literature and is indeed worth noting in our paper. According to its name, the method boils down to surveying company's management on some crucial issues related to corporate governance, regular operations and organizational side of the business in conjunction with the spin-off. The most promising and accurate results in this field were obtained by Seward & Walsh (1996). The article finds out spin-offs usually result in the establishment of an ex-ante efficient corporate management system and Board of directors staffing.

Due to the fact the managerial services are more time consuming and less interpretable than any other kind of research given in this chapter, they are more frequently used in case-study analyses. For example, in Fjellstrцm & Gustafsson (2015) which considers two Swedish demergers: ICA group in 2014 and Wipcore Corporation in 2009, spin-off efficiency is based on series of interviews with the key managers of the companies.

Financial reports approach

Financial reports approach is based on comparison between financial metrics (e.g. return on assets, return on equity, EBITDA or net income margin, revenue growth rate, etc.) before and after the demerger occurs. Researchers usually compare results achieved by the initial company to the joint performance of all the post-spinoff entities and take into account several years before and after the event.

The positive features of financial reports analysis are:

· Universality. In theory, this approach can be applied both for public and private companies conducting spin-offs. Also, as for public companies, this approach does not require its stock to be liquid for the results to be valid unlike event study;

· Long-term nature. Financial reports analysis enables the researcher to assess long-term changes in financial metrics of the companies provoked by the spin-off and thus can serve as the most reliable way detect its influence over a long period of time;

· Straightforwardness. The method implies using deltas of quite simple and intuitive metrics as proxies for demerger efficiency, so it does not need any extra assumptions intended solely for their calculation.

However, the method encounters several significant obstacles:

· Inconclusiveness. Even though the metrics are straightforward, their changes may be subject to some external factors excluded from the initial model and being particularly difficult to identify, quantify or unify in a single equation. These factors can range from macroeconomic events (e.g. market seizure) to firm-specific alterations such as implication of a new ERP system;

· Market disregard. The approach is only based on the analysis of corporate financial reports and thus it does not reflect whether the capital market endorses the company's decision to demerge or not. Indeed, internal financial metrics are crucial in terms of realizing its performance, however one cannot reliably assess whether information asymmetry or wealth transfer hypotheses hold without taking into account market information.

Due to its obvious downturns, financial reports analysis is rarely applied in calculating demerger efficiency. Klein & Rosenfeld (2010) represents an example of such a research. The authors compared ROA, ratio of capital expenditures to net sales, incomes per share and several other metrics before and after the spin-offs documenting a significant growth in these metrics. Another example of such an investigation is Panda & Rao (2012) which analyzes demerger efficiency in terms of changes in ROI, ROC and net income profit margin. Ghosh (2014) reveals positive impact of demergers on EPS, share price growth, ROI and Both articles, however, only discuss whether the influence of demergers is positive or negative on the selected variables and do not suggest any causes of a certain sign or value.

A slightly different kind of research is conducted in Rose & Kiyohiko (2005). The paper gives a comparison of post-demerger performance of spun-off entities and their parents and finds out that revenue growth rate in the spin-offs is higher than in parents for the first several years. However it also reveals that profit margins and profit growth rates stay virtually the same in both post-demerger companies.

Event study

Event study analysis is the most widespread approach towards calculating demerger efficiency. The method is based on analysis of stock price behavior around a specific event. It was first suggested by Fama, et al. (1969) regarding investigation stock splits announcements. Since then, the method has not altered in essence, but has still has been developed significantly in terms of its flexibility and statistical accuracy. It is by far the most common method applied in M&A including multiple implications on emerging markets data (See Grigorieva & Grinchenko, 2013; Partin, 2015; Chirkova & Chuvstvinina, 2011).

Suppose, we have a time series of adjusted share prices from moment 0 to moment T: . For each we define daily stock return as:

Assuming that for each time period the market has it expectations about the return denoted as , we can also find out that each moment there exists some deviation of the actual stock return from its expected value called abnormal stock return ():

A typical event study boils down to the following: a researcher is to calculate a certain statistic based on abnormal stock returns over a certain time period (time window). In case the purpose of the event study is to figure out whether the event had any influence on stock prices, this statistic is assessed in terms of its significance and sign. If, on the other hand, the researcher aims to answer a question about the causes of the event's influence, a correlation or regression analysis is conducted depending on what factors the researcher considers. The most widely used statistic based on daily abnormal returns is cumulative abnormal return for the i-firm (CAR):

As it is evident from the formula and the logic of the method, the two most tricky moments for a researcher conducting an event study are determining the expected stock return proxy and the time window.

There are three main approaches in determining expected stock return: constant mean return model, matched firms or portfolios model and market return model. The first one, applied by for instance Brown & Warner (1980, 1985) assumes that stock return changes in accordance with a simple process yielding that at each given time

.

The second one suggested by Lyon, Barber & Tsai (1999) capitalizes on the idea that one can approximate normal stock returns with returns of a firm or a portfolio with similar characteristics.

The latter, stating that normal returns are a function of benchmark returns is more preferable since its assumptions are not as restrictive as in constant mean model, but even though it does not bear as many additional assumptions and calculations for finding the right firm or portfolio. For simplicity, we use market return model described by McKinlay (1997) which defines expected stock returns for a firm i at day t as:

Where is the stock index return depending on the market i-firm operates in:

denotes linear regression coefficient estimate over pre-event period which is usually taken to be 150-200 days. In our case we use 150 days estimation period.

This approach is quite popular in scientific literature on spin-off efficiency and its determinants.

The table below summarizes the results of several papers that provide multiple event studies in specific markets:

Table 1. Selected event studies dedicated to demerger efficiency
Article	# of spin-offs	Region	Period	Event window	Mean CAR
Miles & Rosenfeld (1983)	55	USA	1962-1980	(-10;10) (-1;1)	7.64%*** 3.34%***
Denning (1988)	42	USA	1970-1982	(-6;6)	2.58%*
Desai & Jain (1999)	155	USA	1975-1991	(-1;1)	4.45%**
Kirchmaier (2003)	48	Europe	1989-1999	(-1;1) (-10;10)	5.4%*** 5.6%***
Veld-Merkoulova (2004)	156	Europe	1987-2000	(-1;1)	0.76%***
Sudarsanam & Qian (2007)	157	Europe	1987-2005	(-1;1)	4.82%***
Harris & Glegg (2008)	58	Cross-country	1990-2006	(-1;1)	2.23%**
Vyas, Pathak & Saraf (2015)	51	India	2012-2014	(-10;10)	1.74%**
*** - significant at 1% level; ** - significant at 5% level; * - significant at 10% level

As it is evident from the table, researchers find significant positive abnormal returns on spin-off announcements. However, one cannot make any particular conclusion about spin-offs in general from this table, since it represents studies conducted in different market over different historical period. The (-1; 1) time window turns out to be the most popular one. This can be attributed to the fact that the time period including 1 day before and after the spin-off announcement are supposed to capture only market reaction to the spin-off itself, but not the type and quality of rumors occurring before the announcement and not the way the demerger is executed which becomes evident shortly after the deal is announced.

To conclude, the advantages of using event studies in demerger efficiency investigations are:

· Market orientation. Event study reveals the market perception of a given event which directly links demerger to the changes in shareholder welfare. Therefore an event study enables the researcher to judge about the spin-off in terms of how it pursues its ultimate goal - increasing shareholder value;

· Ease of attribution. Due to the fact that the time frame can be selected to reflect changes in stock prices associated with the demerger, the CAR can be seen as a refined measure of demerger efficiency and attributed directly to the fact of the spin-off announcement.

Nevertheless, any event study has some significant flaws that can hardly be mitigated due to the specificity of its procedure:

· Short-term results. The study can only reveal market reaction to the spin-off over a short period of time which is reliant current market expectations, but does not reflect any actual improvements in the company's operational performance. There is in fact a developed methodology on long-term event studies (see Kothari & Warner, 2008), but they require more advanced calculations and restrictive assumptions and thus can hardly be seen in demerger literature;

· Volatility. Since the method is very reliant on market performance, the results can be significantly altered in case some market irregularities such as bubbles exist. Therefore one must carefully select the market and time frame of the study in order for the abnormal return figures and adjust them to reflect only the actual results from the demerger.

2.2 Quantification of demerger efficiency determinants

After we defined an accurate proxy for demerger efficiency, the next task in analyzing how certain factors influence it is to build a model describing a theoretical dependency. In order to do that, first we need to quantify demerger efficiency determinants in a form of figures and numbers. This section is devoted to the quantification of demerger determinants.

Focus variables

As it was previously noted in section 1.3, one of the main determinants of demerger efficiency is the extent to which it improves focus concentration of the mother and spin-off companies. In principle there are three main approaches in quantifying concentration:

· Simple dummy variable. Such researches as Whetly et al. (2005), Pickering (2002), Leventis et al. (2011) put into the regression a dummy variable, which equals 1 in case the companies operate in different industries and 0 otherwise. Although simplistic, this approach enables the researcher to test focus hypothesis at least resistance way;

· SIC code based variable. This approach is more sophisticated than the previous one, but can yield more valuable results since it can show a variety of industrial distances between different enterprises. SIC codes are applied in Daley et al. (1995) in calculating the focus variable.

· Correlation between cash flows. As John (1986) proposes, the distance between business activities can be measured as the correlation between cash flows of the parent company and its spin-off. Despite this measure makes sense, its complexity of calculation and lack of additional interpretability relative to other methods resulted in its very rare implementation in literature.

Information asymmetry variables

Since information asymmetry most commonly results in undervaluation of the company and therefore a gap between the book and market values of its assets, Tobin's q or market-to-book figures are commonly applied in financial studies in order to account for information asymmetry. For example, in Grigorieva & Fomenko (2012) MV/BV variable is used as a proxy for the degree of information asymmetry associated with investing in company's stocks.

MV/BV can also be used as a proxy for growth opportunities (Kogan & Papanikolaou, 2010) - a research by Krishnaswami & Subramaniam (1999) also suggests that firms experiencing high growth rate usually urge more information transparency in order to gain access to additional capital for investing in their growth and attract more investors. Following this logic, for the sake of testing whether information asymmetry hypothesis holds one can also include other growth proxies in the regression. For instance, the ratio of capital expenditures to total assets total net PPE or revenue gives a notion of how heavily the company invests and thus to what extent it requires extra external investments and consequently desires information transparency as a result of the spin-off.

Wealth transfer variables

The most straightforward way to determine whether transfer of wealth from creditors to shareholders influences the efficiency of a demerger is the one suggested by Maxwell & Rao (2003) which involves a direct comparison between gains on stock and losses on bonds. In theory, a positive relationship between the two means that wealth transfer hypothesis holds. However, most companies, especially in emerging markets, do not have any tradable debt, so there is no way to calculate debtholders' losses directly.

A possible solution to this problem might be using capital structure related metrics that can indicate that wealth shift hypothesis holds indirectly. In Uddin & Hamid (2010), it is found that there is a significantly positive correlation between abnormal stock returns on demerger announcement and parent company's debt to total assets figure. The logic behind this is that companies with higher indebtedness are more likely to go bankrupt and thus a transfer of some of the assets can result in more significant benefits for the shareholders from the demerger.

Another indirect wealth transfer variable is relative size of the spin-off. Such articles as Zakaria (2014), Veld & Veld-Merkoulova (2008) indicate strong positive influence of spin-off size on demerger efficiency. Such relationship can be easily explained since the more assets of the parent company are spun-off, the more assets are saved in the new entity from a potential bankruptcy, so the larger the shareholder gain. Turning the spin-off size relative dividing it by the amount of total assets enables the researcher to compare spinoff sizes of different observations and estimate regression models.



3. Empirical study of corporate demergers determinants

This chapter presents the framework and results of our empirical investigation of demerger efficiency determinants. Overall, our empirical analysis is based on the theoretical research described in chapters 1 and 2. We use event-study obtained cumulative abnormal returns on mother company stocks as a proxy for demerger efficiency and regress it on several variables corresponding to each of tested theoretical hypotheses under consideration.

3.1 Hypotheses

Following the authors of previously mentioned articles, in our research we test the next hypotheses regarding demerger efficiency determinants.

Hypothesis 1. Industry focus increasing demergers yield higher abnormal stock returns on announcement date

This hypothesis itself or with minor alterations is more or less discussed in most of the examined research literature on demerger efficiency and its determinants. For example, Desai & Jain demonstrates that industry focus increasing demergers on average enjoy 4.62% of abnormal parent stock returns in 1-3 years after the spin-off compared to -0,59% for those that do not increase focus. A more detailed description of this hypothesis and industry focus concentration as both a motive to demerge and a demerger efficiency determinant is given in sections 1.3 and 2.2 of this paper, therefore we do not concentrate on this hypothesis here.

Hypothesis 2. Geography focus increasing demergers yield higher abnormal stock returns on announcement date

Increasing geographical focus during a demerger and its influence on company's stock performance cannot be listed among the most popular demerger efficiency determinants considered in the literature. However, we suggest that geographical specialization of businesses can also result in certain difficulties for a company to act as a single unit. Under these circumstances, spinning-off a geographically distinct entity would serve for the best of the parent one. Besides, operating in multiple countries just as operating in multiple industries, can lead to undervaluation of the stock since different countries have different business risks.

Hypothesis 3. If the demerging company announces the demerger to be motivated by operational issues (e.g. increasing business focus), the spin-off yields higher abnormal returns than is case the company announces raising capital as its primary concern in the demerger

After a preliminary study of corporate press-releases and public demerger announcements, we found that when a company announces its spin-off, it mentions the key motives of why this measure of restructuring is taken and these motives can be divided into two groups (in line with the Section 2.2 classification): business structure reassembling and raising additional capital. Going through a highly complex demerger process just in order to attract additional funds can serve as a negative signal for the investors. According to pecking order theory by Myers & Majluf (1984), equity finance is listed as the most expensive and therefore least preferable type of finance resorted to in case internal funds and debt capital are not available. One can also find analogy with seasoned equity offerings that are reported to exert negative influence on companies' stock prices (For instance, in Toktoanliev & Chirkova, 2012).

Hypothesis 4. Abnormal stock returns experience a reverse U-shaped dependency from the company's relative indebtedness

In studying corporate capital structure and the influence it exerts on demerger announcement abnormal stock returns we came to two contradictory conclusions. On the one hand, as most researchers note, but still not all the empirical results suggest, due to the fact that after a demerger shareholders remain in control of all the initial company's assets and all the cash flows arising from them, whereas debtholders typically can only rely on those that are left in the parent company, a demerger increases the welfare of shareholders at creditors' expense. In case of a bankruptcy, after the demerger shareholders will be left with more assets than in case there was no spin-off at all. This is why higher amount of debt carried by the initial entity will presumably lead to higher abnormal stock returns. On the other hand, an asset withdrawal would leave the parent company more likely to bankrupt - especially under the circumstances of a high indebtedness of the business and, since bankruptcy probability is not a linear function of leverage, this factor can drag shareholder value down. Section 2.2 describes this issue in more detail.

Therefore, due to the theoretically dual action of higher leverage on company's stock performance we propose a U-shaped kind of dependency to be tested against factual data. At relatively low levels of leverage, when the default probability is poor, but still evident and the company remains financially stable, so that an increase in leverage stemming from the demerger execution will not result in a sharp rise of default probability. In parallel, at low levels of leverage, the creditors are still going to agree with the asset transfer. However, high levels of leverage will either force creditors to engage in disputing the fairness of the asset transfer and result in transferring the debt along with the assets or increase the default probability level. This is why we advocate reverse U-shaped type of dependency and will test this hypothesis in Section 3.3.

Hypothesis 5. A higher relative size of the spin-off leads to higher demerger announcement abnormal returns

As it was already discussed, this hypothesis is highly applauded in scientific literature on demerger efficiency determinants. For example Veld & Veld-Merkoulova (2003) finds a positive relationship between abnormal stock returns and relative demerger size and explains this phenomenon by wealth transfer effect. The logic behind that is associated with the previous hypothesis and boils down to the notion that a demerger is a transfer of assets from the joint "possession" of shareholders and creditors into the one of shareholders alone.

...