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

Hypothesis 6. Higher company growth judging by the relative amount of capital expenditures results in substantially higher abnormal stock returns following the demerger announcement

This hypothesis is intended to be associated with information asymmetry theory of explaining demerger announcement abnormal stock returns. Heavy investments and high growth usually mean extra need in investment capital and thus higher valuation by the market. Therefore using demerger as a way to mitigate information asymmetry problems is considered to be a more favorable step for high growing and heavily investing companies.

This hypothesis is, however, different from the 3rd one since the 6th specifically targets not just any capital raising need, but only ones that result in capital expenditures in future growth. The 3rd one is associated with all the capital raising related demerger situations including cases when additional capital is required in order to stabilize the financial situation in the company or pay out some matured outstanding debts.

3.2 Variables and model

Overall, the regression model we use in determining the impact of each factor on demerger efficiency can be described in a single equation:

Under the scope of our regression analysis, we use the following variables:

· - cumulative abnormal return of i-company over a certain time period. As it was previously mentioned, the most widespread measure of demerger efficiency in existing demerger studies is cumulative abnormal return calculated over (-1;1) time window. However some studies (for instance, Miles & Rosenfeld, 1983) also take into consideration longer time periods such as (-10;10). We also looked at (-20;20) period since this kind of time frame is popular in emerging markets M&A studies (see Grigoriyeva, 2013). Regressing different CAR's on different independent variables will supposedly yield substantially different results due to the fact that each type of event windows is unique in terms of its CAR volatility and the amount and quality of information absorbed by the financial market. We will further comment on the possible outcomes of these discrepancies after we discuss independent regression variables;

· is a dummy-variable which equals 1 if the parent company and the spin-off operate in different industries and 0 otherwise. In case the sign of this variable's coefficient is significantly positive, it can be concluded that the focus hypothesis holds;

· is a dummy variable which equals to 1 if the parent company and the spin-off operate in different countries and 0 otherwise. The logic behind including this variable is the same as with industry focus one;

· is a dummy variable which equals to 1 if the demerging company announces increasing focus, developing single strategies for each of the businesses as it primary motive in the demerger and 0 if the demerger is said to be intended to act as a source of additional capital. In explaining the reason why capital attraction motive is considered as a negative sign to the market we adhere to pecking order theory and the logic is described in paragraph 3.2 in more detail;

· means the ratio of total assets intended for the spin-off to the total assets of the initial company. The data on the amount of assets to be spun-off is obtained from companies' press releases and analyst reports in case the company does not provide such an information in its public announcements. A positive sign of is supposed to support wealth transfer hypothesis since the larger the amount of assets saved from the old creditors in a newly commenced entity, the better-off the shareholders are;

· and are the demerging entity's book value of debt to total assets ratio as of the last conceivable accounting period and its square respectively. We use book values since this proxy of leverage can be directly attributed to a potential bankruptcy probability.

· A correlation analysis revealed no significant linear dependence between CAR and debt-to-assets ratios, however our logic and literature review suggest that, if any, there should be a reverse U-shaped connection. Therefore, is expected to have a negative sign and will supposedly be positive concluding that there is an "optimal" capital structure which characterizes a demergers as the most efficient in terms of debt-to-assets ratio;

· is a ratio of LTM pre-demerger capital expenditures to the demerging company's value of net PPE according to the last conceivable balance sheet. A heavy investment in the company's development is supposed to indicate high growth opportunities and therefore result in a higher actuality of the information asymmetry hypothesis. A significantly positive will presumably mean that information asymmetry hypothesis holds to the extent of growing companies;

And, last but not least, is a dummy variable which is needed to control for the type of the market: 1 for emerging markets and 0 for developed ones. Perhaps it would have been useful to suggest some kind of difference between the ways some factors influence CAR in emerging and developed capital markets or i.e. to include an interaction variable in the model. However we did not find any evidence supporting this approach having studied the whole body of demerger efficiency literature available and failed to construct our own logic of suggesting such a difference.

Therefore we use emerging market dummy just as a control variable in order to account for inevitable differences that stem from the fact that emerging and developed markets have different return volatilities.

is a regression error which is assumed to be distributed normally. We suppose the errors to be distributed with different variances due to the fact that, even though there is an emerging market control variable we do not control for industry, country or any other source of exogenous shocks - therefore errors can vary in their distribution parameters from case to case. However since all the announcements took place in different industries, different countries and different time periods there is no reason to assume autocorrelation.

3.3 Data description

In our data analysis, we used Mergermarket.com database to identify all the potential spin-offs to be take into account while assessing the regression model. We considered the period from January 1, 2000 to December 31, 2015 (excluding 2007-2009 recession period) in all countries of the world. Having obtained the total of 277 demerger cases we then selected those that were not subject to any governmental antitrust programs or family-related. Then we controlled for stock liquidity in order for the stock price dynamics to reflect the actual assessment of companies by the capital markets yielding a sample of 113 companies. After that we calculated CAR's based on stock performance data obtained from Thompson Reuters Eikon terminal and Capital IQ. Finally for each i-th company we calculated , , and based on the company's financial reports from Thompson Reuters Eikon, press-releases and other public information and assigned the dummies for each of the companies according to press-releases, SIC codes and geographies.

Having done all the abovementioned calculations, we controlled for irregularities in the data and deleted the outliers. The total sample available for a further investigation consisted of 98 demerger announcements. You can see the list of these cases in Appendix 1.

In Table 2, you can see summary statistics for each of the CAR measures sorted by emerging and developed markets:

It is evident from the table that abnormal returns in emerging markets are significantly higher than in developed countries which is compensated by significantly higher volatility of the returns. However, in both cases they are on average positive and lie in general in line with the findings of other researchers. A curious result is that medians of all types of CAR for each group of cases are very close to each other, although the volatilities increase along with the lengths of event windows. This can be associated with the fact that market reaction to a spin-off can depend on the reaction detected at earlier periods, once triggered a process of post-spinoff abnormal return of 5% can result in 20% at a 10-day horizon. A poorly though performing spin-off will most likely stay poorly performing even in 10 or 20 days, which is implied by the fact that CAR medians are pretty much the same for all kinds of event windows.

The next diagram illustrates the distribution of debt-to-assets ratios sorted by the type of market with vertical axes depicting distribution frequency. Both ones are skewed to the left, however developed markets reveal a more harmonic, one-peaked distribution while emerging markets clearly show two peaks.

Figure 1. Debt-to-assets ratio distribution charts by developed (on the left) and emerging (on the right) markets

Nevertheless, a simple t-test mean comparison shows no significant difference in values of debt-to-assets ratio between emerging and developed markets with the former experiencing a mean of 25% and the latter having 22%. The result can indicate that the selected emerging markets (India, Greater China, South Africa, Brazil, Malaysia, etc.) are moving towards the parameters of developed ones. Despite that the difference between variances in developed markets (12% in developed and 19% in emerging markets, significant at 5%) suggests that there is still a way to go for emerging markets to converge with developed ones.

The picture seems virtually alike when we look at relative spin-off size distribution. However in this case neither means (31% for developed versus 26% for emerging) nor variances (20% versus 22%) show any significant difference in values. Perhaps, this could be attributed to the fact that an overwhelming majority of demergers take place in the US, so the country acts as a trendsetter for both other developing and emerging markets in terms of the specific demerger parameters.

Figure 2. Spin-off size ratio distribution charts by developed (on the left) and emerging (on the right) markets

The pie charts present distribution of industry and geography focusing demergers in comparison between emerging and developed markets. Although visually it seems that in emerging markets industry-focusing spin-offs account for a greater portion of the market than in developed countries, chi-squared test cannot reject the null hypothesis of similar distribution neither for geographical focus nor for industry one. In fact, p-value for geographical focus appears to be 11%, which is slightly over the critical value of 10%. However, it seems like the significance of the difference is only a matter of the number of observations.

After examining the collected data with standard statistical and graphical tools, we can now move on to assessing the regression model proposed in 3.2 and figure out whether the hypotheses mentioned in 3.1 are consistent with the data or not.

3.4 Results

After assessing the proposed regression model under standard Hauss-Markov assumptions and conducting a Breush-Pagan test for all the three types of CAR, we found out that, as expected, homoscedasticity hypothesis is rejected, therefore robustness adjustments have to take place. A regression analysis, assuming HC3-type heteroscedasticity consistent coefficient errors estimates, yields the following results:

As it can clearly be seen from the table, all the 4 proposed hypotheses hold at 10 and even 5% level if one measures demerger efficiency using 10 and 20 days CAR. The motives difference hypothesis is consistent with the data even taking into account the (-1; 1) event window. In addition, Let us now consider each of the results in more detail.

Business focus concentration

In terms of industrial focus concentration, the hypothesis that focus-increasing demergers yield achieve higher results judging by the abnormal returns over the announcement period, is consistent with the data at (-10;10) and (-20;20) event windows. On average, industrial focus increasing spin-offs have 5-6 p.p. more in 10 and 20 days CAR than their non-increasing comparables. For 1-day CAR, the coefficient is not significant, but still positive which also supports our industry focus hypothesis.

However increasing geographical focus appeared to be insignificantly negative. This can be attributed to the fact that geographical focus was increased only in 26% cases out of which almost a half accounted for increasing both industrial and geographical focus, so a simple regression analysis was unable to distinguish between those two types of concentration at the same time. A t-test of sample means shows that H0 hypothesis of equal means cannot be rejected for any of the CAR's. This is why geographical focus requires a deeper investigation to be done.

Motive

As it was presupposed, the motive dummy, unlike other variables, appeared to be significantly positive for each type of CAR taken into account. It turned out, that on average focus-motivated demergers achieve 4.2 p.p. more than their capital raising comparables at (-1;1) event window and over 11% more at larger windows. Therefore we can conclude that signaling hypothesis is coherent with the data. A twofold difference in coefficients in (-1;1) and other CAR's can be explained simply by the fact that 10-days CAR's are more volatile and show higher values on average.

Debt-to-assets ratio

Reverse U-shaped dependency hypothesis is consistent with the data obtained at 10 and 20 days event windows since simple DA coefficient appears to be significantly positive while DA-squared one is significantly positive. The results hold at 5% level. Whereas (-1;1) event window shows both figures to be insignificant, the actual values are still in line with our hypothesis, therefore one can argue that a large sample of data could result in turning these coefficients significant.

The dependency shape suggests that there exists an "optimal" level of leverage at which the abnormal stock returns are maximum. Keeping in mind the estimates of regression coefficients, we can obtain this "optimal" value to be around 40% for 10-days CAR and 30% for 20 days and taking into account that descriptive statistics analysis revealed mean leverage to be 25% for developed countries and 22% for emerging markets, one can conclude that on average companies initiating a demerger appear to have a significantly lower leverage than the optimal one for its shareholders to have maximum benefits.

Relative spin-off size

The size of the spin-off appeared to have no significant influence on its announcement performance. Nevertheless, its coefficient was positive for both (-1; 1) and (-20; 20) event windows. Even though it was negative on 10-day CAR, this result suggests that a larger data sample could in fact lead to an increase in significance and thus a confirmation of the hypothesis. A correlation analysis reveals very weak and insignificant, but still positive relationship between each of CAR's and relative spin-off size, so overall a more profound examination could result in a different conclusion about the factor's statistical influence.

Capital expenditures in company's net fixed assets

The significantly positive sign for 10 and 20 day CAR regressions suggest that there exists a positive relationship between abnormal returns and the degree of how much the initial company invests in its growth. Even (-1;1) period shows insignificant, but still positive result. Supposing that in general fast-growing companies have lower information transparency, it supports information asymmetry theory. However the logic behind this proposition is not straightforward, so in case we encounter a company that has low growth figures, but heavily invests in its fixed assets, the theory may fail.

CAR terms

As it was expected, the signs and significance of certain factors in the regression estimates varies between different types of CAR. Overall, the estimation revealed that longer period types of CAR are vulnerable to changes in such financial metrics as debt-to-assets measure and ratio of capital expenditures whereas for (-1; 1) CAR the only factors that matter are the announced demerger motive and type of market. The logic behind this difference can be associates with the notion that at longer periods stock market analysts take into account deeper fundamental characteristics while during the (-1; 1) event window the only new information that occurs in the market is contained in the company's press-release related to the demerger.

3.5 Limitations and future research

Unfortunately, just as any kind of empirical research, our paper has some that overall make it less valid it terms of the accuracy of the conclusions made and practical applicability of the results. To some extent, these flaws are still negligible, but however a future research is needed to resolve all the issues arisen from it.

The most significant limitation of our empirical results is rooted in the problems of sample selection. Due to the fact that we were strictly limited by the availability of data and its quality, the sample of roughly 100 events from emerging and developed markets cannot be considered absolutely representative in terms of, especially, the countries analyzed. India and China accounted for about 70% of the emerging markets sample, so the results for emerging markets can more or less justly be applied to them or maybe Asian countries in general, but not to each and every one - especially to CEE markets. In order to increase the degree of accuracy, one shall construct a more harmonic data set using a broader specter of data mining tools so that the principles of sample building will only be dictated by the urge to make it representative.

The next problem associated with the sample is the fact that our research is not concerned with suggesting any kind of differences in the way certain demerger efficiency factors work in emerging markets compared to developed ones. Actually we made an attempt to add interaction variables in the model such as the product of emerging dummy and capex and this variable appeared to have a significantly negative coefficient in (-10; 10) regression. However, since we have not found any way to explain this difference or any lead in the literature, this result cannot be presented in our study along with the others.

Another particular flaw of this research is associated with the chosen demerger efficiency proxy. While event study analysis is by far the most widely accepted way to tackle the measurement problem, it still gives the researcher a narrow field for making any conclusions since it is a: (1) short-term measure of demerger efficiency, (2) vulnerable to any market instability at the time of the event, (3) depends on the chosen event window and proxy for the expected stock return. These upsides of using event studies are explained in Section 2.1 in more detail. Therefore a deeper investigation of demerger efficiency determinants requires an analysis of long-term post-spinoff performance. A possible option for future studies would be testing the same or at least similar hypothesis using changes in financial performance metrics after the spin-off instead of abnormal returns. demerger corporate restructuring determinant

Additionally, in our empirical study we basically touched upon all the demerger efficiency determinants hypotheses but one - manager incentives. Since the data on compensation packages is mostly difficult to obtain, we did not make any attempt to test this hypothesis in our research. However as modern corporations are becoming more and more concerned with human capital management, increasing incentives for managers can become a more significant motive for corporates to demerge and therefore a need for such examinations is becoming more apparent. A possible way to deal with this hypothesis would be tracing the dependency between stock and option compensation practices in a company with demerger efficiency. Assuming that after a demerger corporate culture and the way of designing manager contracts stays the same in the spin-off, we can suggest that a company with higher ratio of stock-based compensations may have higher abnormal stock returns on spin-off announcement.

Conclusion

All in all, we can conclude that our research reached its ultimate goal and solved all the tasks given on the way to it. Firstly, we investigated that all the corporate demerger motives can be divided into several groups: those associated with asset disposal, those associated with capital raising and specific ones. Then, accordingly, we discovered and formulated four major hypotheses on what drives demerger efficiency. Next, we gave a description of data tools typically employed in such investigations with their pros and cons. Eventually, we formulated our own empirical hypotheses and tested them against factual data.

In brief, our findings suggest that focus concentration is the most prominent and motive of demergers significantly increasing announcement abnormal returns - both in terms of the announced demerger motive and the factual data on the industry differences between the parent and the spin-off. Another worthwhile finding is the confirmation of the reverse U-shaped kind of dependency between leverage and CAR which at some extent advocates wealth transfer theory, however in dramatically modified way. We can also see that information asymmetry hypothesis holds in terms of the dependency between CAR and initial company's capital expenditures. However, geographical focus and relative spin-off size, despite the conclusions in the existing literature, did not show any significant influence on spin-off performance. A future research in this field can be dedicated to improving data samples employed, theoretical modelling and testing other hypotheses such as the managerial incentives one.

In terms of its practical implications, this research can serve as a guide for corporations that are going to engage in spin-off activity, their stakeholders and stock analysts for a more efficient participation of them in the planned demerger. It also can provide some insights into scientific theories of capital structure, information asymmetry and signaling.

References

1. Григорьева С. А., Гринченко А. Ю. (2013) Влияние сделок слияний и поглощений в финансовом секторе на стоимость компаний-покупателей на развивающихся рынках капитала. Журнал "Корпоративные финансы"

2. Григорьева С. А., Фоменко Н. В. (2012) Детерминанты метода платежа в сделках слияний и поглощений на развивающихся рынках капитала. Журнал "Корпоративные финансы"

3. Ивашковская И.В., Янгель Д.О. (2007) Жизненный цикл организации и агрегированный показатель роста. Журнал "Корпоративные финансы"

4. Партин И. М., Масленникова М. В. (2015) Детерминанты эффективности международных сделок по приобретению компаний из стран Европейского союза. Журнал "Корпоративные финансы"

5. Токтоналиев А. Р., Чиркова Е. В. (2012) Эффекты дополнительных размещений акций. Журнал "Корпоративные финансы"

6. Чиркова Е.В., Чувствина Е.В. (2011) Реакция рынка на объявление о приобретении компаний открытого и закрытого типов. Журнал "Корпоративные финансы"

7. Abarbanell J.S., Bushee B.J., Raedy J.S. (2003) Institutional Investor Preferences and Price Pressure: The Case of Corporate Spin?Offs. The Journal of Business

8. Allen, W.A., Lummer, S.L., McConnell, J.J. and Reed, D.K. (1995). Can takeover gains explain spin-off gains? Journal of Financial and Quantitative Analysis

9. Aron D. J. (1991) Using the capital market as a monitor: corporate spinoffs in an agency framework. The Rand Journal of Economics

10. Bergh D., Johnson R. A., Dewitt R. (2008) Restructuring through spin-off or sell-off: Transforming information asymmetries into financial gain. Strategic Management Journal

11. Bowman E.H. & Singh H. (1993) Corporate restructuring: Reconfiguring the firm. Strategic Management Journal

12. Brown S.J., Warner J.B. (1985) Using daily stock returns: The case of event studies. Journal of financial economics

13. Carhart, M. (1997). Long-Run Performance after Stock Splits: 1927 to 1996. Journal of Finance

14. Christensen C., Alton R., Rising C., Andrew Waldec A. (2011) The Big Idea: The New M&A Playbook. Harvard Business Review

15. Comment R, Jarrell G.A., (1995) Corporate focus and stock returns. Journal of financial Economics

16. Daley, L., Mehrotra, V. and Sivakumar, R. (1997) Corporate focus and value creation Evidence from spinoffs. Journal of Financial Economics

17. Danielova A.N. (2008) Tracking stock or spin-off? Determinants of choice. Financial Management

18. Denning K.R. (1998) Spin-offs and sales of assets. Accounting and Business Research

19. Zakaria N. (2014) Spin-Off and Value Creation: The Case of Malaysia. Working paper

20. Desai, H. and Jain C.P. (1999) Firm performance and focus: long run stock market performance following spinoffs. Journal of Financial Economics

21. Feng et al. (2015) Executive compensation and the corporate spin-off decision, Journal of Economics and Business

22. Furlan A, Grandinetti R. (2014) Spin-off performance in the start-up phase - a conceptual framework. Journal of Small Business and Enterprise Development

23. Gaughan P.A. (2007) Mergers, Acquisitions and Corporate Restructurings Wiley: New York.

24. Ghosh (2014) Gaining Synergy by Spinning Off. Globsyn Management Journal

25. Habib, M.A., Johnsen, B.D., Naik, N.Y. (1997) Spin-offs and information. Journal of Financial Intermediation

26. Harris O, Glegg C (2008) The wealth effects of cross-border spinoffs. Journal of Multinational Financial Management

27. Hite GL, Owers JE. (1983). Security price reactions around corporate spin-off announcements. Journal of Financial Economics

28. Huson, R.M. and MacKinnon, G. (2003) Corporate spinoffs and information asymmetry between investors. Journal of Corporate Finance 9John, 1986

29. Khorana A., Shivdasani A., Stendevad C., Sanzhar S. (2011) Spin-offs: Tackling the conglomerate discount. Journal of Applied Corporate Finance

30. Kirchmaier T. (2003) The performance effects of European demergers. Centre for Economic Performance, LSE.

31. Klein A., Rosenfeld J. (2010) The long-run performance of sponsored and conventional spin-offs. Financial Management

32. Kogan & Papanikolaou (2010) Growth opportunities and technology shocks. The American Economic Review

33. Krishnaswami, S. and Subramaniam, V. (1999) Information Asymmetry, Valuation, and the Corporate Spin-Off Decision. Journal of Financial Economics

34. Leventis S., Sismanidou M, Koulikidou K, Dasilas A. (2011) Wealth Effects and Operating Performance of Spin-Offs: International Evidence. Working paper

35. Maxwell & Rao (2003) Do Spin?offs Expropriate Wealth from Bondholders? The Journal of Finance

36. MacKinlay, A.C. (1997) Event studies in Economic and Finance. Journal of Economic Literature

37. Mehrotra V, Mikkelson W, Partch M (2003), The Design of Financial Policies in Corporate Spin-offs. The Review of Financial Studies

38. Myers S.C., Majluf N.S. (1984) Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics

39. Myers (1984) The capital structure puzzle. The Journal of Finance

40. Nanda V & Narayanan M. P. (1999) Disentangling value: financing needs, firm scope, and divestitures. Journal of Financial Intermediation

41. Panda B., Rao P. (2012) Corporate restructuring: Demerging impact, SCMS. Journal of Indian Management

42. Pickering M. (2002) Clean break. Divestitures and demergers. Journal of the Australian Institute of Chartered Accountants

43. Pyo, U. (2006) Enhancing Managerial Incentives and Value Creation: Evidence from Corporate Spinoffs, Working Paper

44. Ramu S.S. (1999) Restructuring and Break-ups: Corporate Growth Through Divestitures, Splits, Spin-Offs and Swaps. Response Books

45. Rose, E.L. and Kiyohiko, I. (2005) Widening the Family Circle: Spin-off in the Japanese Service Sector. Long Range Planning

46. Seward J.K., Walsh J.P. (1996) The governance and control of voluntary spin-offs. Strategic Management Journal

47. Simon, S. (1960) Spin-offs vs. dividends in kind. Accounting Review

48. Schipper & Smith (1983) Effects of recontracting on shareholder wealth: The case of voluntary spin-offs, Journal of Financial Economics

49. Sudarsanam, P.S. and Qian, B. (2007) Catering theory of corporate spin-offs: empirical evidence for Europe, working paper. Cranfield University, UK

50. Thomasson O., Janusonis A., Managing a demerger process-A case study of a corporate divorce. Goteborg Universitet

51. Uddin H. (2010) Corporate spin-offs and shareholders' value: Evidence from Singapore. The International Journal of Business and Finance Research

52. Veld, C. and Veld-Merkoulova, V.Y. (2004) Do spinoffs really create value? The European case. Journal of Banking and Finance

53. Veld, C. and Veld-Merkoulova, V.Y. (2008) An empirical analysis of the stockholder-bondholder conflict in corporate spin-offs, Financial Management

54. Wheatley, C., R. Brown and Johnson, G. (2005) Line-of-business disclosures and spin-off announcement returns, Review of Quantitative Finance and Accounting

Appendix 1. List of spin-offs analyzed

Размещено на Allbest.ru