Corporate Acquisitions and the Operating Performance of Malaysian Companies

Transcription

1 See discussions, stats, and author profiles for this publication at: Corporate Acquisitions and the Operating Performance of Malaysian Companies ARTICLE in JOURNAL OF BUSINESS FINANCE & ACCOUNTING MAY 2004 Impact Factor: 0.69 DOI: /ssrn Source: RePEc CITATIONS 29 READS 78 2 AUTHORS, INCLUDING: Rashidah Abdul Rahman King Abdulaziz University 85 PUBLICATIONS 390 CITATIONS SEE PROFILE Available from: Rashidah Abdul Rahman Retrieved on: 19 February 2016

2 Journal of Business Finance & Accounting, 31(3) & (4), April/May 2004, X Corporate Acquisitions and the Operating Performance of Malaysian Companies R. ABDUL RAHMAN AND R.J. LIMMACK* 1. INTRODUCTION One of the issues facing policy-makers in developing economies is to identify the extent to which they should adopt marketbased policies that are perceived as acceptable by developed economies. One particular area in which this issue arises relates to takeover activity. Takeover activity may be perceived as one aspect of the market for corporate control that provides a disciplinary mechanism on management (Jensen and Ruback, 1983). In addition the takeover market provides an exit strategy for owner managers of growing firms. In a less-well developed economy, however, it may not always be possible to follow a liberal approach to takeover activity. Other policies, such as that of transferring a greater proportion of ownership of productive assets to the indigenous population (as with the bumiputra policies adopted in Malaysia) may be perceived as more important. 1 Even though corporate acquisition activity is relatively * The authors are respectively Lecturer at UiTM, Shah Alam, Malaysia and Retired: formerly Professor of Accountancy and Finance, University of Stirling, Scotland. They wish to thank participants at the EFMA Conference, Athens, 2000 and an anonymous referee for their valuable comments. (Paper received May 2002, revised and accepted October 2002) Address for correspondence: R. Abdul Rahman, Head of Consultancy (Finance), Institute of Research, Development and Commercialization Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia. rash@pc.jaring.my # Blackwell Publishing Ltd. 2004, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. 359

3 360 RAHMAN AND LIMMACK new in Malaysia, it is a significant aspect in the growth of business sectors and the development of the Malaysian economy. With an increase in the level of acquisition activity, corporate leaders need to be aware of the difficulties in achieving value maximising behaviour in corporate acquisition decisions. It is also important that legislators do not make hasty decisions about the need for regulation in a field which economic theory suggests may have the potential for major benefits in a marketbased economy. The current study aims to contribute to the debate on takeover activity by examining the financial performance of a sample of Malaysian companies that made acquisitions in the period Rather than focusing on shareholder wealth effects, however, we are more concerned with identifying whether takeovers in Malaysia lead to an improvement in corporate operating performance. Our approach is therefore based on analysis of the operating cash flow performance using a sample of Malaysian companies involved in takeovers between Jensen (1984) argued that shareholders wealth increases in takeover situations are derived from improved operating performance and increased efficiency. Other authors have identified a wide range of additional sources of potential improvements (see for example Weston et al., 1998). Recent research, however, has provided contradictory evidence on the presence of gains to bidding company shareholders and indeed on the existence of net wealth gains. This conflicting evidence has been observed in a number of countries with developed capital markets, including the USA, the UK and Australia. One possible way of reconciling the contradictory findings is to suggest that takeovers do indeed improve economic efficiency on average, but acquirers simply pay too much for the benefits. Researchers who have attempted to address the question as to whether takeovers actually lead to an improvement in operating performance, however, have generally found limited evidence to support this thesis. One possible policy argument is to propose that the onus is on advocates of takeover activity to demonstrate that the benefits outweigh what are perceived to be potentially harmful effects, including the development of monopoly power. The danger in this argument for developing markets is that politicians may favour legislation that

4 PERFORMANCE OF MALAYSIAN COMPANIES 361 limits not only the negative but also the positive aspects of takeover activity. From the regulatory viewpoint, it is therefore important that researchers establish a prima facie case for the presence of operating efficiencies in takeover activity. In this particular context it is less important to determine which group of shareholder, if any, benefits from improvements in operating performance. In smaller capital markets, including Malaysia, takeovers of private companies predominate. The opportunity for improved efficiency that may arise from disciplinary takeovers will occur less frequently than improvements from other potential sources of gain, including synergistic benefits. 2 The majority of research to date has focused on takeovers of publicly quoted companies with relatively few studies examining takeovers of privately held firms. Recent examples of the latter, however, include studies by Chang (1998) and Da Silva Rosa et al. (2001). The latter two studies have concentrated on the short-run wealth impact on shareholders of acquiring firms. Neither of these studies addresses the more contentious issue of long-run postacquisition performance (Loughran and Vijh, 1997, and Lyon et al., 1999) and both investigated takeovers in well-established markets. The main contribution of the current study is to investigate long-run post-acquisition performance of companies involved in takeover activity in one developing market, that of Malaysia. The selected sample, reflecting acquisitions of private companies, allows us to test whether opportunities for gain are present even in the absence of a disciplinary role. The method of analysis is to examine operating performance of Malaysian companies involved in acquisitions during the period The performance metric used is based on control-adjusted operating cash flows of the bidder and target companies prior to acquisition and of the combined corporate entity after acquisition. In addition, tests are undertaken to identify the source of any change in performance. The results reported in the current study suggest that acquisitions in Malaysia during the period lead to improvements in the long run operating cash flow performance. The improvement in performance results from both increases in return on sales (cash flow per dollar of sales) and in asset turnover (sales per dollar of assets). These improvements

5 362 RAHMAN AND LIMMACK are not achieved at the expense of the long-term viability of the combined firms, as they are also accompanied by an increase in the level of capital investment. The results demonstrate that Malaysian acquisitions do lead to improvements in operating performance that provide the potential for benefits to bidding company shareholders and for the economy as a whole. Both our prior assumptions and preliminary analysis of target and bidding companies suggest that our sample is unlikely to contain many, if any, disciplinary bids. Our results are therefore consistent with the presence of other sources of operating improvement. One caveat that must, however, be placed on the results is that most of the target companies in Malaysian acquisitions and, all of the targets in the current sample are privately owned companies. It is not possible to state conclusively that the results reported here would necessarily apply to acquisitions of public quoted companies. 3 The rest of this paper is organised as follows. Section 2 summarises the recent empirical evidence. Section 3 describes the data selection procedure and the methodology used to compute acquisition-induced changes in performance. Section 4 presents results on changes in operating cash flow and some of its components including return on sales, asset turnover, capital expenditure and cash flow from operating expenses. The final section draws conclusions from the study. 2. PREVIOUS EVIDENCE There are two research approaches normally employed in addressing the question of the efficiency of the takeover mechanism. One approach is to focus on the profitability of companies involved, using accounting data. The other approach is to employ share price data to establish the distribution of gains and losses to shareholders. Both measures should be related to post-acquisition cash flows, thus it might be assumed that they would lead to consistent results. 4 However, previous studies that have used both measures suggest otherwise. Studies based on analysis of accounting data have attempted to assess the economic impact of acquisitions by testing for changes in profitability of the combined firm. Most accounting-

6 PERFORMANCE OF MALAYSIAN COMPANIES 363 based studies, whether based on UK or US data, tend to lend support to the view that acquisitions do not result in improved performance and that acquisitions are non-value maximising to shareholders. Examples of such studies include those of Mueller (1980), Ravenscraft and Scherer (1987), Clark and Ofek (1994), Philappatos and Baird (1996) and Denis, Denis and Sarin (1997) for the USA; and Singh (1971), Utton (1974), Meeks (1977), Holl and Pickering (1988) and Dickerson, Gibson and Tsakalotos (1997) for the UK. However, studies by Lorie and Halpern (1970) and Lev and Mandelker (1974) based on US data reported that merging firms perform significantly better than non-merging firms. While these results are fairly consistent, concerns have been expressed over the use of accounting measures of performance, including the choice of time period over which to measure performance and the need to adjust for accounting effects of the takeover (Appleyard, 1980). An alternative approach to the above focuses on security returns together with an identification of the wealth gains or losses to the various shareholder groups. Market-based studies that have focused on announcement period returns tend to find that acquisitions provide positive abnormal returns to bidding company shareholders. 5 While there remains some debate about the long-run wealth impacts of takeovers, a recent survey of the literature by Agrawal and Jaffe (2000) concludes that the long-run performance is negative following mergers, though performance is non-negative (and perhaps even positive) following tender offers. Magenheim and Mueller (1988), Lahey and Conn (1990) and Clark and Ofek (1994) find negative post-acquisition performance for merged firms in the USA. 6 Only Franks, Harris and Titman (1991), and Loderer and Martin (1992) find positive long-run post-acquisition returns. Loughran and Vijh (1997) provide evidence that significant negative excess returns are obtained by the shareholders acquiring firms that made merger bids using share financing whereas the shareholders of firms that made cash tender offers experience significantly positive excess returns over the five-year period following acquisition. 7 Agrawal, Jaffe and Mandelker (1992) showed that the results obtained by Franks et al. (1991) were time specific ( ) and a function of the sample of

7 364 RAHMAN AND LIMMACK acquisitions examined. 8 Agrawal et al. (1992) also reported that acquisitions undertaken in the time period 1955 to 1987 are followed by significant negative returns over a five-year period after the outcome announcement date. A number of UK studies have also reported negative long-run post-acquisition returns for the acquiring firms, including those by Limmack (1991), Kennedy and Limmack (1996), Sudarsanam, Holl and Salami (1996), Gregory (1997) and Baker and Limmack (1999). Only two papers report that acquiring firm shareholders gain in the long-run from acquisition, Franks, Broyles and Hecht (1977) for the brewing industry and Franks and Harris (1989), with the latter providing conflicting results depending on the benchmark control used. 9 In addition, Higson and Elliot (1998) report no evidence of negative abnormal returns three years post-acquisition for bids conducted in the period Over the period , however, the latter authors report evidence of significant positive abnormal return during the two years after acquisition. Higson and Elliot conclude that postacquisition returns are sensitive to the observation period. In summary market-based studies have produced a conflict in results between the announcement period share price reaction and subsequent negative long-run post-acquisition behaviour. This conflict has led a number of authors (Healy et al., 1992; Jarrell, 1995; and Gregory, 1997) to suggest that the results may be a reflection of methodological errors, including the use of inappropriate control models, rather than the acquisition per se. This view is partly reinforced by the length of time over which the negative post-outcome returns are observed as this appears to contradict the notion of an efficient securities market (Ruback, 1988). Possible methodological errors include the choice of inappropriate control models or some form of selection bias either in the control model or the sample being examined. Attempts to provide what are deemed to be more appropriate models in long run studies by Franks et al. (1991) and Agrawal, Jaffe and Mandelker (1992) in the US; and Gregory (1997), Higson and Elliot (1998) and Baker and Limmack (1999) in the UK, have eliminated some of the more obvious potential sources of bias. Gregory (1997) and Baker and Limmack (1999) have shown that the negative pattern of post-outcome abnormal returns is not a function of size or book-to-market effects. Nor

8 PERFORMANCE OF MALAYSIAN COMPANIES 365 are the negative returns found in these studies caused by survivorship or prior performance bias. However, the continued presence of a long run negative drift in post-outcome returns does leave a worrying question over the appropriateness of this particular methodology. In summary, there is some conflict between the results reported in short-term and long-term market-based studies. Although there is some agreement between results reported for long-term market-based studies and accounting-based measures of performance, both sets of research have been subject to the criticism of inappropriate methodology. There remains therefore a fundamentally unresolved question of the long-term effects of acquisitions on firm performance. In the current study an alternative research design is adopted, which uses cash flow analysis to measure the performance of companies involved in takeover activity. The approach adopted is similar to that used in studies by Healy et al., (1992 and 1997), Manson et al. (1994 and 2000), Clark and Ofek (1994), Anand and Singh (1997) and Ghosh (1998). These latter studies provide evidence suggesting that acquisitions are economically efficient in the long run and therefore create the opportunity for wealth increases. (i) Data 3. DATA AND METHODOLOGY The current study focuses on takeovers of Malaysian companies that were initiated and completed during the period January 1, 1988 to December 31, The sample period selected provides a focus on recent acquisitions and also ensured that sufficient pre-and post-acquisition performance data was available for our tests. A list of takeover bids was identified from the Kuala Lumpur Stock Exchange s (KLSE) monthly Investors Digests. Initially, 160 proposed acquisitions, involving 160 bidders and 213 targets were identified. 11 The list was then cross-checked with the Annual Companies Handbook published by the KLSE and the respective companies files in the KLSE to identify the bid outcome. Restrictions imposed on the initial sample are as follows:

9 366 RAHMAN AND LIMMACK (i) The acquiring firm must be quoted in the KLSE and acquire more than 50% voting rights of targets in order to create the holding company-subsidiary relationship identified in Section 5(1)(a) of the Malaysian Companies Act Lapsed bids are excluded. (ii) The purchase price for the target must exceed RM5 million. 12 (iii) Acquisitions involving financial firms and investment trust are excluded because of their specific accounting and regulatory requirements (iv) At least two years of pre-acquisition financial data was required for target and bidder (excluding year of acquisition). In addition five years of post-acquisition data was required for the newly combined company. The requirement for five years post-acquisition data stems from the view that value increasing improvements in efficiency might not materialise for several years (Healy et al., 1992 and 1997; Manson et al., 1994; Jarrell, 1995; and Ghosh, 1999). 13 Although imposition of this requirement would potentially have introduced a survivorship bias into the sample, no acquisition was actually excluded for this specific reason. The final sample included 94 quoted acquiring and 113 private target companies. 14 Details of the relationship between the initial and final sample are reported in Table 1 while size statistics are reported in Table 2. Financial reports on the quoted acquiring and control companies were obtained from the KLSE library while those on non-quoted target and control companies were Table 1 Number of Acquisitions and Final Sample by Year Total Initial bids identified Financial Lapsed Dormant targets Target company accounts not available Non-listed bidders Public listed firms Final sample

10 PERFORMANCE OF MALAYSIAN COMPANIES 367 Table 2 Size Characteristics of Sample of Bidders and Targets (RM 000) a Targets Bidders Relative Size of Target to Bidder Mean 47, , Median 20, , Standard Deviation 58, , Maximum 433,117 1,783, Minimum 1,267 5, Note: a Size is measured as the book value of shareholders funds (share capital plus reserves) plus net debt less cash and marketable securities, measured at the end of the financial year prior to the acquisition. obtained either from the companies themselves or from the Malaysian Registrar of Companies (ROC) in Kuala Lumpur. The preponderance of privately owned targets in Malaysia suggests that many of the acquisitions are likely to be agreed bids and that these are therefore unlikely to portray the characteristics of disciplinary bids. Results of prior studies based on analysis of security returns suggest that it is the latter type of acquisition that is most likely to produce net wealth gains. Hence our sample is likely to exhibit some bias against a finding of improvement in operating performance. The relatively small size of target to bidder, reported in Table 2, is also likely to mitigate against any finding of performance change. Frequent or active bidders (bidders making more than one bid in subsequent years) are included in the current study. It is possible that acquiring firms undertaking frequent acquisitions do so as a result of perceived benefits from earlier acquisitions. Thus, excluding them from the sample may bias the results. We therefore include companies that have made multiple acquisitions in our sample from the date of the first takeover identified in the period and adjust both combined firm and control firm measures for all subsequent acquisitions. Multiple acquisitions in the same financial year are treated as a single acquisition. If a further acquisition is made in a subsequent year then pre-bid performance of bidder and control companies are based on the combined performance of the bidder plus prior target. Performance in the post-acquisition period is based on that of the

11 368 RAHMAN AND LIMMACK newly combined company while that for the control is based on a pseudo combination that includes the additional control company. Previous studies have adopted a variety of control benchmarks. Healy et al. (1992), Manson et al. (1994) and Harford (1999) use the median industry performance as their control. Barber and Lyon (1996) and Ghosh (1998) suggest that control firms selected on the basis of industry and size are likely to serve as better benchmarks than median industry performance. Ghosh (1998) argued that as large firms, on average, are more profitable than small firms the use of industry medians as a benchmark may introduce bias into the analysis. In the current study industry-matched control companies were selected from the population of non-acquiring and non-target companies. 15 Bidders were also matched on the basis of size. Size matching was also undertaken as far as possible for targets although the absence of available information on the full population of potential controls made this more difficult. We base the definition of size on the book value of total operating assets at the end of the year prior to acquisition (represented by share capital plus reserves plus total debt, less cash and marketable securities). We were unable to match using market values as the targets were mainly private companies. (ii) Performance Measurement As asserted by Healy et al. (1997), security returns around the takeover announcement represent investor s expectation of acquisition benefits whereas post-acquisition cash flow performance measures the actual benefits, if any, generated by acquisitions. Rayburn (1986) and Bowen, Burgstahler and Daley (1987) provide evidence that cash flows can incrementally explain abnormal stock returns in the USA. Bowen et al. (1987) found: that cash information is consistent with the information impounded in security prices and also has incremental explanatory power beyond that contained in accrual flows alone (p. 746). The use of an operating cash flow measure also allows the impact of the acquisition to be assessed independent of the accounting method adopted to report the acquisition or changes in accounting policies post-acquisition. Thus, the primary meas-

12 PERFORMANCE OF MALAYSIAN COMPANIES 369 ure of performance used in the current study is the ratio of operating cash flow to operating assets of the companies involved. The cash flow measure used in this study is defined as profit before tax, depreciation and interest, adjusted for changes in working capital. This definition is similar to that adopted by Bowen et al. (1986), Manson et al. (1994) and Ali and Pope (1995) but differs from that used by Healy et al. (1992 and 1997), Anand and Singh (1997) and Ghosh (1998). The latter authors make no adjustment for working capital accruals, which have been shown to be subject to manipulation by management. 16 Dechow (1994) points out that: many financial analysts regard operating cash flow as a better gauge of corporate financial performance than net income, since it is less subject to distortion from differing accounting practices (1994, p. 5). Murphy and Zimmerman (1993) regard the difference between accounting profits and cash flows, as that over which managers can exercise the most discretion. To compare performance across firms, measures of operating cash flow are scaled by the book value of assets, calculated as book value of shareholders funds and total debt less cash and marketable securities at the beginning of the relevant year. We use book values rather than market values for two reasons. First we do not possess market values for our target companies as these were private companies. Secondly one of the problems that previous authors (in particular Healy et al., 1992) have tried to avoid is to bias the denominator by prior expectations of how the takeover is likely to perform. In their studies they attempt to exclude any market reaction to the bid itself. If the market had perfect foresight then the share price would capture all of the expected gains. The use of market values in the denominator would then hide any post-acquisition change in performance. As we are attempting to capture the change in operating performance we need to exclude any share price change that reflects the expected performance change resulting from the acquisition. Healy et al. (1992) achieved this by eliminating price changes around the bid announcement date on the assumption that the market is efficient in relation to the news and that there is no long-term adjustment. On the basis of recently published research (for example Agrawal, Jaffe and

13 370 RAHMAN AND LIMMACK Mandelker, 1998; and Gregory, 1997) we cannot be confident that the price adjustment is swift. Hence even if it were possible to provide proxy values for target companies we would still have a problem in identifying the period over which value changes should be excluded from the denominator. (iii) Expected Performance Model We construct a consolidated measure of combined bidder and target pre-acquisition operating performance for each of the four years ( 4 to 1) prior to takeover (wherever available). 17 The performance of each is weighted by the book value of operating assets of each at the beginning of the relevant year. 18 The post-acquisition operating performance is calculated using the actual values reported by the combined firm, deflated by the book value of operating assets at the beginning of the relevant year. 19 We exclude from our denominator any identified revaluation of fixed assets or goodwill arising from the acquisition. 20 The control benchmark operating performance measure is then calculated for each pair of control companies in each of the four years prior to the bid, where available, and for the five years post-bid. In the pre-acquisition period, the combined control company performance measure is weighted by the relative operating asset values of the bidder and target firms at the beginning of the relevant year. The post-acquisition combined benchmark is computed by weighting the individual company performance by the relative asset value of the bidder and target firms at the end of the year prior to the acquisition. The controladjusted operating performance is obtained by subtracting the relevant benchmark measure of operating performance from the combined measure for the bidder and target companies. We report measures of median annual performance based on all observations for the pre- and post-acquisition periods. In addition summary statistics are constructed for each takeover individually by calculating the median of the control-adjusted measures ðap c preiþ over the four years prior to acquisition, where available, and also for the median value over the five years after acquisition. Tests are undertaken of the difference in median control-adjusted performance from before to after the

14 PERFORMANCE OF MALAYSIAN COMPANIES 371 bid. In addition the change in control-adjusted operating performance is estimated using the following model: 21 where: AP c posti ¼ þ AP c prei þ " i AP c posti is the median annual control-adjusted operating performance for company i for the post-acquisition years AP c prei is the median annual control-adjusted pre-acquisition operating performance median for the pro-forma combined bidder and target company. The intercept represents the change in control-adjusted performance. The slope coefficient captures any correlation in cash flow returns between pre-and post acquisition years. The advantage of using a regression analysis is that it avoids making assumptions about the relative change in performance from pre- to post-acquisition (Ghosh, 1998) RESULTS OF TESTS ON OPERATING PERFORMANCE This section provides discussion and analysis of the results for the various tests with both median and mean values reported in the tables. However, analysis focuses on median returns as the mean values obtained in this study are influenced significantly by outliers. In comparing median performance across the different sample groups, the Wilcoxon signed rank tests and Mann-Whitney tests are used. (i) Cash Flow and Asset Growth Rates The rate of change in cash flow and operating assets in both pre-and post-acquisition periods are reported in Table 3, relative to the values in the year prior to acquisition. Thus the mean and median change for period ( 1, þ2) represents the change from year 1 to the second year post-acquisition. Panel A of Table 3 shows that the rate of growth in operating cash flows of the combined bidder and target was not significantly

15 Table 3 Control-adjusted Growth in Operating Cash Flow and Operating Book Value of Assets for Combined Bidder and Targets in Malaysian Acquisitions in the Period 1988 to Period Relative to Acquisition Year ( 4, 1) ( 3, 1) ( 2, 1) ( 1, þ1) ( 1, þ2) ( 1, þ3) ( 1, þ4) ( 1, þ5) % % % % % % % % Panel A: Rate of Growth of Cash Flow Median: Combined Firm 71 b 98 b a 83 a 210 a 316 a 322 a Control 64 b b 25 b 30 b 7 Control-adjusted b 1 b 197 a 288 a 353 a Trimmean (20%) Combined Firm 119 c 128 b 43 b 101 a 126 a 330 a 657 a 642 a Control 126 b b 9 59 a 54 a 68 a 20 a Control-adjusted a 44 a 365 a 629 a 782 a % Positive Combined Firm Control Control-adjusted Panel B: Rate of Growth of Operating Assets Median: Combined Firm 15 a 18 a 5 a 95 a 116 a 188 a 271 a 330 a Control 7 11 b 3 b 9 22 a 36 a 54 a 95 a Control-adjusted a 109 a 153 a 190 a 229 a 372 RAHMAN AND LIMMACK

16 Trimmean (20%) Combined Firm 42 b 35 a a 200 a 327 a 462 a 671 a Control 29 b 18 b 8 b 19 a 39 a 57 a 97 a 151 a Control-adjusted a 120 a 190 a 231 a 331 a % Positive Combined Firm Control Control-adjusted No. of observations Notes: 1 Operating cash flow is defined as operating profit before tax and extraordinary items, adjusted for depreciation and goodwill and changes in working capital (that is, changes in stocks, trade debtors and prepayments and changes in creditors and accruals). The operating book value of assets at the beginning of the year is the book value of shareholders funds (share capital plus reserves) plus total debt, less cash and marketable securities. Before the acquisition (year 1), cash flow and operating asset values of the combined firm and their controls are weighted averages of the acquirer and target values, with the weight being the relative operating asset values of the two firms. The values of the combined firm are used in the post-acquisition period. Post-acquisition control returns are target control and bidder control values, weighted by the relative operating asset values of the two corresponding bidder and target firms at the beginning of the year prior to acquisition (year 1). Control-adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the control firm in the same industry during that period. a Significantly different from zero at the 1% level, using a two-tailed test. b Significantly different from zero at the 5% level, using a two-tailed test. c Significantly different from zero at the 10% level, using a two-tailed test. PERFORMANCE OF MALAYSIAN COMPANIES 373

17 374 RAHMAN AND LIMMACK different to that of the combined control companies in any of the years prior to the takeover. By contrast, the rate of growth was significantly greater than that of the controls in each of the five years post-acquisition. The high values for the rate of increase in cash flow in post-acquisition years is partly due to a low cash flow for some observations in the year prior to acquisition. 23 However, the significantly higher rates of growth are still observed when attention is focused on median values. As shown in Panel B, the median and trimmed mean controladjusted rates of growth in operating assets are also significantly positive in the post-acquisition period (although not in the pre-acquisition period). The period of time covered in the current study was one of a high rate of growth in the Malaysian economy. The results suggest that firms involved in takeovers were at the forefront of this growth. Of itself these preliminary results are not inconsistent with the notion advanced by some authors that takeovers are undertaken in the interests of management (Mueller, 1980). In the next section we examine whether Malaysian takeovers also create the opportunity for shareholders to benefit. (ii) Pre-acquisition Operating Performance Table 4 provides summary statistics of the median and mean operating performance of targets and bidders relative to their controls, and of bidders relative to targets in each of the four years prior to acquisition. As shown in Panel A of Table 4, the median control-adjusted operating returns for targets are positive in each of the four pre-acquisition years although they are only significantly different from zero in years 2 and 1. The use of cash flow measures of performance reduces the likelihood that what we are capturing here is accounting policybased earnings management by the directors of the target company to enhance a potential bid price although we cannot rule out non-accounting-based earnings management. To the extent that the latter behaviour has occurred our subsequent analysis will be biased against finding any evidence of improvement in post-acquisition operating performance. Allowing for the possibility of unobservable earnings management the results suggest that prior to the acquisition target companies perform

18 Table 4 Summary Statistics Panel A: Pre-acquisition Control-adjusted Operating Performance for 113 Targets and 94 Bidders 1 Targets Control-adjusted Bidders Control-adjusted Year Relative to Acquisition Median % Mean % % Positive Obs. Median % Mean % % Positive No. of Obs b b b c c 1.23 c Average for years ( 4 to 1) 3.60 a a 0.45 c Panel B: Pre-acquisition Operating Performance of Bidders Relative to Targets 1 Bidders Targets Bidders vs Targets Year Relative to Acquisition Median % Mean % Median % Mean % Median % Mean % % Positive No. of Obs Average for years ( 4 to 1) Notes: 1 Operating performance is defined as operating cash flow deflated by the book value of operating assets. Operating cash flow is defined as operating profit before tax and extraordinary items, adjusted for depreciation and goodwill and changes in working capital. Operating assets at the beginning of the year is the book value of equity plus total debt less cash and marketable securities. Control-adjusted values are computed for each firm and year as the difference between the firm value in that year and the corresponding value for the industry-matched control firm. a Significantly different from zero at the 1% level, using a two-tailed test. b Significantly different from zero at the 5% level, using a two-tailed test. c Significantly different from zero at the 10% level, using a two-tailed test. PERFORMANCE OF MALAYSIAN COMPANIES 375

19 376 RAHMAN AND LIMMACK better than non-acquired companies. Not surprisingly, given the nature of our sample of private targets, this result is contrary to that expected based on traditional theories of the disciplinary role of takeovers and also contrasts with that found in early accounting studies of target company performance. 24 Holl and Pickering (1988) and Kennedy and Limmack (1996) for the UK; and Palepu (1986), Lang, Stultz and Walkling (1991) and Mikkelson and Partch (1997) for the USA, found evidence supporting the hypothesis that targets are generally taken from the population of firms with poor pre-bid performance. The sample of acquisitions in the current study, representing takeovers of private Malaysian companies, does not however, appear to be disciplinary in nature. Panel A of Table 4 also summarises operating cash flow returns of bidders relative to their control companies from years ( 4 to 1). Again traditional economic theory suggests that takeover bids are likely to be initiated by companies from the more efficient sectors of an economy. 25 However, as shown in Table 4, bidders underperform relative to the control companies during the period prior to acquisition. The median performance for the acquiring companies is lower than that of their controls for all pre-acquisition years except year-4. The median annual control-adjusted performance is also significantly negative ( 0.88%) over the 4-year period prior to acquisition. Panel B of Table 4 summarises the results of comparison of operating performance of acquiring with target firms. Lang, Stultz and Walkling (1989) and Servaes (1991) provide evidence that well-managed bidders with high q ratio created more value by taking over poorly performing companies with low q ratio. Holl and Kyriazis (1997) also provide evidence that target companies in the UK with low Tobin s q are taken over by acquiring firms with high Tobin s q. 26 As reported in Panel B of Table 4, the median operating performance of bidders is higher than that of targets in years 4 and 3 prior to acquisition, but lower in the two years immediately prior to acquisition. In none of the years is the difference significantly different from zero. The result is inconsistent with Manne s (1965) concept of a market for corporate control in which the more capable and competent executive teams tend to replace those that are less capable and

20 PERFORMANCE OF MALAYSIAN COMPANIES 377 competent. Again, however, we emphasise that this disciplinary role is unlikely to be present in agreed bids, representing the normal characteristic of the private acquisitions included in the current sample. (iii) Operating Performance of Combined Firms As reported in Table 5, the median operating performance for the combined firms in the four years prior to acquisition ranges from 3.98% to 7.16%. In the five years subsequent to the acquisition, median operating performance of the combined firms has improved, with values ranging from 5.42% to 9.93%. The median annual post-acquisition performance for years (þ1 to þ5) for the combined firm is 7.37%, and is significantly higher than the median annual pre-acquisition performance. No significant change in operating performance of the control companies was however, reported from pre-(6.44%) to post-acquisition (4.95%). The control-adjusted median and mean performance measures are also reported in Table 5. In the pre-acquisition period, median control-adjusted operating performance ranged from 1.26% to 5.15%, although in no year was this significantly different from zero. The median (and mean) control-adjusted operating performance is positive in each of the five postacquisition years, ranging from 0.27% to 5.65% (2.75% to 11.23%). The control-adjusted median performance is significantly positive in years þ3, þ4 and þ5 while the mean return is significantly positive in four out of five post-acquisition years. In addition the (average) median control-adjusted annual performance of the combined bidder and target firms increased significantly from 0.12% in the period prior to acquisition to 2.95% in the post-acquisition period. In summary the results demonstrate that the operating performance of Malaysian companies improves following acquisition. It should be emphasised that this improvement is not simply as a result of some form of profit-averaging, through the acquisition of relatively profitable targets, as our basis of comparison is with a weighted average of target and bidder control-adjusted pre-acquisition performance.

21 Table 5 Operating Performance for 94 Combined Acquiring and Target Firms for Malaysian Acquisitions Completed in the Period Panel A Combined Firm Control Control-adjusted Year Relative to Acquisition Median % Mean % Median % Mean % Median % Mean % % Positive No. of Obs Average for years ( 4 to 1) þ c þ þ b 7.11 a þ a a þ a 7.79 b Average for years (þ1 toþ5) 7.37 a b a 6.88 b RAHMAN AND LIMMACK

22 Panel B: Change in Control-adjusted Operating Performance AP c posti ¼ 0:0375 þ 0:204 APc prei (3.3) a (2.95) a (3.5) a (2.60) b R 2 ¼ 0.09 F-statistic ¼ 8.72 a AP c posti and APc prei are the median annual adjusted operating performance in the pre- and post- acquisition period for firm i. T stats in second brackets are those adjusted for heteroscedasticity (White s correction) Notes: 1 Operating performance in the pre-acquisition period is calculated as pre-tax operating cash flow return on operating assets of target and bidder, weighted by the relative asset sizes of the two firms. Post-acquisition performance uses data for the combined firms. Pre-acquisition combined control returns are calculated using the individual target and bidder control returns, weighted by the relative operating asset values of the corresponding bidder and target firms at the beginning of the years. In the post-acquisition period the weights used to compute control company returns are the relative operating asset values of the acquirer and target firms in year 1. Control-adjusted values are computed for each firm and year as the difference between the firm value in that year and the value of the corresponding industry-matched control firm. a Significantly different from zero at the 1% probability level, using a two-tailed test. b Significantly different from zero at the 5% probability level, using a two-tailed test. c Significantly different from zero at the 10% probability level, using a two-tailed test. PERFORMANCE OF MALAYSIAN COMPANIES 379

23 380 RAHMAN AND LIMMACK As explained earlier, we also calculate the median of the four years pre-acquisition performance for each combined firm with a similar measure calculated over the five post-acquisition years. Post-acquisition median operating performance is then regressed on the pre-acquisition operating performance. 27 Panel B of Table 5 shows the results of the regression on abnormal control-adjusted cash flow returns. The slope coefficient is significantly different from zero but the value, of 0.204, indicates a lack of persistence in control-adjusted cash flow returns over time. The intercept,, is , indicating that the combined firms obtain a significant increase of 3.75% per year in post-acquisition period after controlling for the pre-acquisition performance. 28 The decision to allow the slope coefficient to be unconstrained (as above) is open to potential criticism. For example, Manson et al. (1994) and Gadad and Thomas (2000) argued that a slope coefficient close to zero reflects a highly competitive industry in which the pre-acquisition control-adjusted performance ðappre c Þ would therefore be expected to steadily revert to zero as any comparative advantage is eventually eliminated. By contrast a slope coefficient of one would reflect an uncompetitive economy in which comparative advantage is retained, while an intermediate slope coefficient reflects an imperfectly competitive industry. We therefore repeat the analysis but constrain the slope coefficient to be either zero or unity (assumed extreme values). When the slope coefficient is constrained to be equal to one, the improvement in performance is 2.75% (insignificantly different from zero). However, when the slope coefficient is constrained to be equal to zero, then the improvement in performance is a statistically significant 4%. The above results are consistent with those reported by Healy et al. (1992) and Ghosh (1998) for US acquisitions, and by Manson et al. (1994) for UK acquisitions. Healy et al. (1992) found a significant improvement in pre-tax operating performance relative to the industry average over the post-acquisition period. 29 Ghosh (1998) also found that acquiring firms show significant improvement in their post-acquisition operating performance relative to their industries average. 30 Similarly, Manson et al. (1994) reported a significant improvement in

24 PERFORMANCE OF MALAYSIAN COMPANIES 381 operating performance after acquisition for takeovers in the UK during the period 1985 to In order to test the robustness of the above results to alternative definitions of operating performance, we repeat our analysis using the definition adopted by Healy et al. (1992) i.e. excluding any adjustments for working capital accruals. We again find a significant increase in the average annual controladjusted post-acquisition performance although application of this alternative definition results in a smaller proportion of observations reporting positive post-acquisition controladjusted performance (58% versus 62%). 31 Overall, however, the conclusions are unaffected by the use of the alternative definition. In summary the results above indicate that our sample of Malaysian takeovers do lead to an improvement in operating performance even when the companies involved are not driven by disciplinary motives, as is likely to be the case in our sample. 32 (iv) Sources of Operating Performance Changes In the current section we attempt to identify the source of the improvement in operating performance. The operating cash flow return on assets can be decomposed into operating margin and sales turnover. The operating margin measures pre-tax operating cash flow per dollar of sales while the sales turnover ratio measures the dollar sales generated for each dollar invested in assets. Table 6 provides summary data on operating margin for the 94 combined target and bidder firms in the years surrounding the acquisition. The median annual operating margin increased (not statistically significant) from in the pre-acquisition period to in the post-acquisition period. The combined control companies, on the other hand, experienced a decrease in operating margin from to over the same time period. The control-adjusted median operating margin increased from in years ( 4 to 1), to in the post-acquisition period. The Wilcoxon signed rank test indicates that the increase in median annual control-adjusted postacquisition operating margin is significant at the 5% level.

25 Table 6 Operating Margin for 94 Combined Acquiring and Target Firms for Malaysian Acquisitions Completed in the Period Panel A Combined Firm Control Control-adjusted Year Relative to Acquisition Median% Mean% Median% Mean% Median% Mean% % Positive No. of Obs b b b b Average for years ( 4 to 1) þ þ þ c þ a þ a Average for years (þ1 toþ5) c a b b RAHMAN AND LIMMACK

26 Panel B: Abnormal Adjusted Post-acquisition Operating Margin (t-values in parentheses) AS c posti ¼ 0:10 þ 0:12 ASc prei R 2 ¼ 0.09 F-statistic ¼ 9.4 a (2.04) b (1.18) (2.53) b (1.02) AS c posti and ASc prei are the median annual control-adjusted operating margin in the post- and pre-acquisition period for firm i. T stats in second brackets are those adjusted for heteroscedasticity (White s correction) Notes: 1 Operating margin is defined as operating cash flow as a percentage of sales. Performance measures for the combined firm in the pre-acquisition period are weighted by the relative asset sizes of the two firms. Post-acquisition-performance used data from the combined firms. Pre-acquisition control company operating margins are calculated from individual target and bidder control returns, weighted by the relative operating asset values of the two corresponding bidder and target firms at the beginning of the relevant year. In the post-acquisition period the weights used to compute combined control company returns are the relative operating asset values of the acquirer and target firms in year 1. Control-adjusted values are computed for each firm and year as the difference between the firm measure in that year and the measure for the control firm during that period. a Significantly different from zero at the 1% probability level, using a two-tailed test. b Significantly different from zero at the 5% probability level, using a two-tailed test. c Significantly different from zero at the 10% probability level, using a two-tailed test. PERFORMANCE OF MALAYSIAN COMPANIES 383

27 384 RAHMAN AND LIMMACK We also report the results of regression of the median operating margin for each combination over the 5-year postacquisition period on the median of the 4-year pre-acquisition control-adjusted operating margin in Panel B of Table 6. The intercept in the cross-sectional regression is a significantly positive 0.10, confirming that the operating margin has improved post-acquisition. Healy et al. (1992) also found that combined firms earned a higher operating margin than their industry counterparts in the post-acquisition period. However, the latter authors were reluctant to attribute the increase to the acquisition itself as the median operating margin was also higher than that of the control companies in the pre-acquisition period. Summary statistics on sales turnover, defined as the ratio of sales to operating assets, are reported in Table 7. The median annual pre- and post-acquisition rates of turnover for the combined firms are 0.41 and 0.52, respectively. Examination of results for individual years pre-acquisition suggest that the improvement had begun in year 1 although we are unable to speculate as to whether this improvement would have continued in the absence of the takeover. The median annual sales turnover rate is relatively unchanged for the control companies, increasing slightly from 0.47 before the acquisition to 0.51 post-acquisition. Overall the median annual control-adjusted sales turnover for the combined firms is 0.05 over years ( 4 to 1) but increased significantly to 0.01 over the period (þ1toþ5). The results of regression of median control-adjusted post-acquisition sales turnover rates for each company on the pre-bid measure are summarised in Panel B of Table 7. The value of the intercept term at 0.08 is significantly different from zero, and confirms the finding of a significant improvement in the combined firms sales turnover in the post-acquisition period. The results are again similar to those reported by Healy et al. (1992) who found that the median sales turnover of the combined firms was lower than that of the industry control prior to acquisition but comparable to the industry control in the post-acquisition period. One puzzling feature of the above analysis is that the improvement in operating performance and in operating margins is most apparent in post-acquisition years three to five whereas the greatest improvement in sales turnover came in the first two