Effects of Take-over Motives and Ownership Structure on Premiums Paid : Evidence from Malaysia

This study seeks to determine the effect of motives for take-overs and ownership structure on the level of premiums paid in take-over transactions in Malaysia. Using 136 take-over transactions for the period of 1990 -1999, we found that the take-over premiums were much higher than those in developed countries. The various motives for take-overs had different impact on the level of premiums paid in different industries. The evidence also suggests that the main motive of the bidders was more towards enhancing the earnings base of the bidders rather than playing the disciplinary role for under-performed target. Family-owned firms paid a lower level of premiums compared to non-family firms unless the targets had good performance (relatively higher ROAs). This study concludes that ownership by family aligned the interests of the owners to that of the shareholders rather than resulting in the expropriation of minority shareholders.


Introduction
Generally premiums paid in take-over transactions are positively related to potential acquisition-related benefits, either from expected synergistic gains or private benefits of control.Given the high concentration of ownership especially in the hands of families, the large controlling owners are expected to align their interests to that of the shareholders (Jensen and Meckling, 1976).However, many studies highlight that the entrenched manager (Note 1) or the controlling shareholders in firms with high concentration of ownership engage in non-value-maximizing activities for private gains in East Asian countries (Claessens et al, 1999;Lemmon and Lin 2003).The phenomenon above is highlighted by Johnson et al. (2000) as "tunnelling" which reflects that controlling shareholders have strong incentive to transfer out resources from the firm for their own private benefits at the expense of minority shareholders.For instance, the controlling shareholders can buy over companies or assets owned by them at a very high premium (Cheung, 2004).
This study seeks to examine the effect of ownership structure and the motives for take-overs on the level of premiums paid in take-over transactions in Malaysia.Malaysia provides a good case of study as it was among the most active corporate control markets in the ASEAN region in the 1990s (Metwalli and Tang, 2002).It accounted for 41 percent of the total deals and 38 percent of the M&A transaction value of target firms in ASEAN from 1990ASEAN from to 2000.M&A transactions between local firms accounted for 74 percent of all the M&A transactions.
Using 136 take-over transactions for the period of 1990 -1999, we found that the take-over premiums were much higher than in developed countries.The various determinants have different impact for the level of premiums paid in different industries.Malaysian take-over market was used more for enhancing the earning base of the bidder rather than playing the disciplinary role for under-performed targets.It is found that family owned firms pay a lower level of premiums compared to non-family firms unless the targets have high ROA.High concentration of ownership (by family) tends to align the interests of the owners to that of the shareholders (Jensen and Meckling, 1976) rather than the expropriation of minority shareholders (Claessens et al, 1999).
The structure of this paper is organised as follows: Section 2 summarises the existing literature on motives for take-overs, ownership structure, and other factors that could have an effect on the level of premiums paid.Section 3 describes the methodology, and the a priori expectation of the variables used.Section 4 describes the data.Section 5 reports results from cross-sectional regressions and finally section 6 summarizes and concludes the paper.

Literature
Generally, in finance literature, there are three main motives for take-overs, namely, operational, managerial, and financial synergies (Megginson, 2004).The premiums paid represent the expected synergies to be gained from the take-overs.In order to obtain operational synergies, the business combination would require much overlap in the activities of the two firms (Sudarsanam, 1996).This synergy is expected when the assets of the target and bidder firms are used more effectively by the combined firms than by the target and bidder operating separately.As a result of the possible synergistic benefits that may accrue from related acquisitions, e.g economies of scale and scope, the premiums paid should be higher for the target.
In terms of managerial synergies, managers will vigorously seek under-performing targets in order to exercise their managerial expertise and transform the target firms into profitable ones.This disciplinary role is more prevalent in the US especially in the 1980s (Shleifer and Vishny, 1988).Managerial synergies could also be generated if targets have unique expertise and thus have strong earning capabilities that would enhance the performance of the bidding firms.Thus, the bidder is not only acquiring the target's stock but also the right to control that will enable the bidder to use the target's assets and resources to maximise the value of the bidder's firm value.
Financial synergies occur when a take-over results in lower costs of capital, lower default risks or coinsurance effects of the combined firms.For instance, the bidders would be able to get tax advantages from high gearing targets.If one of the parties, either the bidder or the target, has unused debt capacity while the other already operate at its optimal gearing level, the combination of the two firms will be able to exhaust the unused debt capacity.As such, if the bidder's motive is to take advantage of the target's gearing level, he would be willing to pay a higher premium.This is supported by Walkling and Edmister (1985) who found that there is a negative relationship between debt ratio and premiums paid.
Generally, there is little empirical evidence on the determinants of premiums paid especially outside the United States.Slusky and Caves (1991) using the U.S. data from 1986-1988, found that real and financial synergies evidently contribute less in explaining the variance of premiums than do agency factors.They contend that weakly monitored managers of acquiring firms overpay.They find a negative relation between premiums paid and substantial shareholders who are not directors, and larger managerial ownership.
In companies with high concentration of ownership such as those in the East Asian countries, the direction of the relationship between premiums paid and ownership structure is unclear.Agency theory predicts that managers may indulge in non-maximising transactions such as excessive consumption of perquisite or sub-optimal risk taking activities, such as take-overs, when they do not have a significant ownership stake in the firm (Hubbard and Palia, 1995;Lewellen, Loderer and Rosenfeld, 1985).If managers act to serve their self-interest rather than the shareholders, the tendency for them to overpay the bid premiums is greater as the take-over of companies is associated with greater perquisite and prestige, and lower risk (Amihud and Lev, 1981).At the lower managerial ownership stake, managers are willing to pay a high bid.However if managers hold higher ownership stake, the interest of the managers is more aligned to that of shareholders' (McConnell and Servaes, 1990) and will result in a negative relationship between the bid premiums and the managerial equity stake.
Nevertheless, Hubbard and Palia (1995) highlight that at higher managerial ownership stakes, managers enjoy private benefits of control such as discretion over the choice of investment projects and control over the appointment of members of the board of directors.As such, if they value this control, they are unwilling to sell their stake in the financial markets, which results in managers holding a large non-diversified financial portfolio.In order for the managers to diversify their portfolio, they are willing to pay a high premium for the target even when they own a significant equity ownership.Thus, it is obvious that the impact of large shareholders on mitigating agency problems is still vague.
However, given the background that diffused ownership is less desirable in the weak legal systems (Morck and Yeung, 2004) in emerging markets, the presence of large shareholders in mitigating the agency conflicts is more likely.Accordingly, as family ownership is usually concentrated, the motives for take-over by family ownership would mitigate the level of premiums paid.
Other determinants of premiums used in the US studies include take-over characteristics such as method of payment and size (Slusky and Caves, 1991;Loughan and Vijh, 1997), number of bidders ( Walker, 2000), and hostile or friendly take-overs (Tse and Soufani, 2001).The take-over characteristics in this study which are relevant, are the size and method of payment.This is because most of the take-over deals in Malaysia involved only single bidders and are done on a "willing buyer and willing seller" basis.Myers and Majluf (1984) contend that if the management of the bidding firm has superior inside information that its assets are undervalued, cash financed acquisition is more likely to happen.This is a positive signal sent by the bidder to the market that the bidder's existing assets are undervalued.Shawky et al (1996) argue that there should be a negative relationship between premiums paid and cash offer.However their empirical findings show otherwise.On the other hand, cash offers involve immediate capital-gains tax.(Note 2)Thus, bidders need to pay a higher premium in order to offset the capital gain tax incurred by target shareholders (Wansley, Lan, and Yang, 1983;Brown and Ryngaert, 1991).
In a study by Chang (1998) on the returns of bidders on the acquisition of privately-held targets, he found that in stock offers, bidders experience positive abnormal returns, which is in contrast to the negative abnormal returns typically found in acquiring a publicly traded target.He contends that this is due to the creation of large blockholders in the bidding firm from the target if common shares are issued to the target shareholders.These blockholders can serve as an effective monitor of managerial performance.The willingness of target shareholder to take large positions in a firm also conveys favourable information about the firm.Their findings are also supported by Fuller et al (2002) andAbdul Rahman (2002).Thus, the premiums paid would be higher for payments involving cash instead of equity.On the contrary, if the take-overs involve related party transaction, the potential for the expropriation by the controlling parties would be greater (Cheung, 2004).Thus, premiums paid would be greater.
The size of the target could be one of the factors that affects the premiums paid to the target.If the target size is bigger, the impact of the combination will be more significant.Jarrell and Poulsen (1989) and Loderer and Martin (1992) found a positive relationship between the relative target size and the cumulative abnormal returns of the bidder upon announcement of the combination.However, this is in contrast to the argument by Palia (1993) and Shawky et al (1996) that smaller targets are priced more competitively than bigger targets.This is because smaller targets will put less strain on the bidder's financial resources and will require less integration efforts.Thus, the relationship is not absolute.
The model for the determinants of premiums paid and their a priori expectations are specified in the next section.It specifically examines the impact of the synergy and ownership variables on the premiums paid, controlling target size, method of payment and Asian Financial crisis that could have had an impact on the premiums paid.

Methodology
To examine the impact of synergy and ownership variables on premiums paid, we regress the premiums paid on the proxies for the various synergies, ownership structure and control variables.The first and second base models are as follows Mathematically the models are expressed as follows: In the first model, relatedness of business (RELATE as a proxy for types of combination and operational synergies), the difference in debt/asset between the target and bidder (T_DTA as a proxy for latent debt capacity and financial synergies), pre-take-over returns on assets (AVGROA as a proxy for managerial efficiency), ownership concentration (BIDLARGE and BIDLARGESQ for non monotonic relation), method of payment (MPAY), new dominant block of ownership created in the combined firms (NEWBLOC), target size (LOTAR_TA), crisis period dummy variable (CRIYEAR), and industry sector dummy variable (DUMSEC) are used as the independent variable to explain the variations in premiums paid.The computation of premiums paid follows the measure used by Palia (1993) and Shawky, Kilb and Staas (1996) who use the ratio of the offer price divided by the book value of the target.This measure is used as most of targets in the sample were non-public listed firms.This ratio gives an indication of how many times the bidder is willing to pay for the target firm over its book value.
The second model includes the interaction effect of the identity of the dominant ownership and motives for take-over.FOWN is a dummy variable equal to one if the firm's dominant owner is an individual or family, zero otherwise.Appendix 1 shows the definition of the variables used.

A priori expectation of the determinants of premiums paid
The synergy variables namely RELATE, AVGROA, and T_DTA in Equation 1 are expected to have positive signs as the potential synergies enhance the value of the combined firm, and thus induce the bidder to pay a higher premium.The ownership concentration is expected to have a curvilinear relationship.At the lower level of managerial ownership stake, managers are willing to pay a higher bid in order to facilitate additional perquisite consumption.However if managers hold higher ownership stake, the interest of the managers is more aligned to the shareholders' interests (McConnell and Servaes, 1990) and will result in a negative relationship between the bid premiums and the managerial equity stake.The MPAY is expected to have positive sign as bidders need to pay a higher premium in order for the target shareholders to relinquish their shares.This is in contrast to equity payment whereby the privately-held target shareholders will potentially become the blockholder in the combined company.Subsequently, the coefficient for NEWBLOC is expected to have a negative sign.Conversely, if the transaction involves related party transaction, the relationship is expected to be positive as this may be a source of tunnelling.
The sign for the target size variable is expected to be negative since smaller size puts less strain on bidder's financial standing.
During the Asian Financial crisis, M&A activities dropped sharply in Malaysia creating a distorted picture.To account for this distortion, we introduced a dummy variable (CRIYEAR) with a value of 1 for the crisis years of 1997 and 1998 and 0 otherwise.The expected sign for this variable was negative.
Equation 2 shows the interaction effects of the identity of the dominant ownership (family ownership versus others) and bidders' motives for take-over on the level of premiums paid.All other explanatory variables are expected to have the same signs as in equation 1. Accordingly, as family ownership is usually concentrated and the objective is to protect the family's business, the motives for a take-over by family firm would mitigate the level of premiums paid.The interaction term of family variable and the motives should indicate which of the dominant factors contribute to the level of premiums paid by family ownership.

Data
The initial M&A announcement list from 1990-1999 was identified from the Investors Digest published by KLSE (various issues).The actual combinations of the firms were confirmed by checking through the Companies Announcement Files, (Note 3)Annual Reports and the KLSE Annual Companies Handbook.Financial data for the quoted firms were obtained from the KLSE Annual Companies Handbook while data for the non-quoted targets were obtained from the Companies Announcement Files of the bidders.The pre-take-over financial data was collected for three years prior to take-over and 4 years for the post-take-over period.Thus, the data collected spread from 1987 to 2003, covering a period of 17 years.As actual combinations of the businesses took some time to materialise after the announcement dates, the financial characteristics of the bidders were further confirmed using the corporate handbooks or annual reports.The ownership data was obtained one year prior to take-overs and the new block created was examined after the take-over year.If the dominant owner was a company, the owner of the dominant owner was traced further in order to get the ultimate owner.If the ownership chain included any non-public listed companies, the records kept by the Companies Commission of Malaysia (CCM, formally Registrar of Companies) were used.
It was found that only about 60 (466/781) percent of the targets announced were successfully taken over by the bidders.About 80 percent (376/466) of the targets were non-public listed firms while 13 percent were public listed firms and another 6 percent were foreign firms.The small number of cross-border deals could be due to risk aversion by management, as it is more difficult to get suitable targets in foreign acquisitions and they require greater integration efforts.Moreover, domestic acquisitions are expected to generate more synergies as compared to foreign acquisitions as the combined firms could share the primary activities and secondary activities of a firm (Porter, 1985).
As the majority of the targets were from non-listed companies, which were relatively smaller and closely held, only those with more than 51 percent acquisition stakes were included.This is to ensure that the take-over would result in a change in control of the targets.Although the take-over Code is meant for public listed companies, pursuant to Practice Note 3 of the Code of Take-over and Mergers (1987 and 1998), it also applies to private limited companies having shareholders' fund of more than RM5 million or where the purchase consideration is more than RM10 million.It was later increased to RM10 million in book value and purchase consideration of not less than RM20 million in 1998.Thus, for the purpose of this study, the selected target had purchase price of not less than RM5 million as too small a target would not have any significant impact on the bidders (Seth, 1990).Minority buyout or situations where the controlling parties purchase the remaining shares of the firms from the minority shareholders were excluded, as the impact of these kinds of acquisition would not be as apparent.As for the public listed firms that were relatively larger, only those with more than 20 percent acquisition stakes were considered.This is because a 20 percent purchase stake is sufficient to effect a change in control of the public listed corporation (Loh, 1996).
About 24 percent (110/466) were excluded due to complications involved in the restructuring exercise; for instance, the purchase of RZ equities by Malaysian Helicopter Service Bhd followed by the restructuring of MAS.Other conglomerates such as UEM, Berjaya Group, Uniphone Telecommunication which announced the take-overs of the targets but were followed by major series of corporate rationalisation activities were also excluded.Other exclusion criteria for the sample include those targets that did not have the profit and loss account or balance sheet before the announcements.These were usually newly incorporated companies, dormant companies, foreign targets, and targets that held concessions or licenses for operation.For instance, the acquisition of Sampling Plywood (Baramas) Sdn Bhd, which held timber concessions, by Glenealy Bhd was valued based on the estimated cash flow of the concession and thus financial statement were not applicable.
Out of the 158 available targets, 14 with negative book values and 8 with incomplete transaction information were omitted.Thus, the final sample for the targets was about 30 percent (136/466) of the confirmed targets that met the criteria to be included in the sample.

Descriptive Statistics
Take-over activities in Malaysia were concentrated more in the first half of the 1990s, with 79.5 percent in value and 78.4 percent in number of cases by 1995.The M&A activities were at its peaks in 1993 and 1995.Table 1 shows the descriptive statistics of the targets and bidders in the sample.More than 50 percent of the targets earned a positive profit.About 50 percent of the targets had asset size of more than RM51 million.On average the total liabilities and book value of the targets were about two third (98/145) and one third (46/145) of their assets respectively.The average returns on assets (ROA) was rather high at 56 percent, but the median ROA was only 7 percent.
The mean and median of the premiums paid were at 8.16 and 3.64 times book value, respectively.(Note 4) This is higher than the average paid in the US (Shawky, Kilb and Staas, 1996, 2.24x;Slusky and Caves, 1991, 1.5x;Walkling and Edmister, 1985, 1.5x;Hanouna, Sarin, and Shapiro, 2001, 1.3x).As for the largest ownerships stake, it is found that 50 percent of the largest owners had 59 percent stake and above while the average was at 62 percent.67 percent (92/136) had ownership stakes of 100 percent.As the majority of the targets were closely held, very often the largest shareholders were also the founders of the companies.As such, the successful rate of this type of take-overs was higher as control was solely in the hand of the founders and thus, most of the time they were friendly take-overs.
The largest ownership stake of the bidders on average was 32 percent.Related party transaction (where the substantial shareholders or directors of the bidding firms own equity stakes in the target) constitute about 40 percent of the sample.About 12 percent of the transactions resulted in reverse take-overs where the targets' dominant owners became the dominant owners in the bidding firms.
Table 2 shows the characteristics of various proxies for motives and premiums paid for the targets by industry.
The sample shows that in the 1990, the majority of the targets (54/136, 39 percent) were from the trading and services sector, followed by property sector (22 percent) and industrial products sector (17 percent).The mining sector, which was a major sector in the KLSE in the early days, had the fewest number of take-over activities as many mining firms had ceased operation in the 1990s.
Most of the take-overs in the sample had purchase prices ranging between RM 20 -100 million and about one third had a purchase price more than RM100 million.In terms of diversification, 54 percent of the transactions were in the same or related industries and 46 percent were in unrelated industries.The majority of the bidders in the property sector and trading and services sector diversified into related sectors.In the case of targets, all the firms in consumer products sector were acquired by bidders from the same sector.
Out of the 136 targets, 18 (13 percent) were suffering losses when acquired.The majority of these firms came from the property sector.Most of the targets were already performing well prior to acquisition.This shows that the disciplinary role of market for corporate control was not really taking place here.The take-over could have been motivated by the prospects of broadening the earnings base of the bidders as the majority of the acquisitions were in the trading and services sector which was booming in the 1990s.The service industry was the major contributor to Malaysia's gross domestic products (GDP) in the 1990s and is much larger than the manufacturing and agricultural sectors in Malaysia's economy.
Except for the plantation sector, the majority of the targets had very high debt/asset ratios.Theoretically, capital structure should differ systematically across industries; however, in the context of Malaysia, there was generally a high level of debt across most industries, perhaps driven by government's implicit or explicit encouragement for debt financing.(Note 5) This was partly due to the government's commitment to a high growth policy based on a high ratio of investment to GDP in the 1990s.Some of the lending, especially to those involved in mega projects, was based on collateral and implied government support and not project cash flows (Thillainathan, 1999).As such, high lending rates and poor cash flow returns eventually caused problem loans during the financial crisis in 1997.
In terms of method of payment, only about 27 percent of the targets were financed by cash, mainly from the property sector and the trading and services sector.The rest were non-cash deals or financed solely by equity.If the target was big, the owner of the targets would then became the block holders in the bidding firms after the business combination if it was financed by equity.This study found that in 12 percent of the cases, the owners of the target firms eventually became the dominant shareholders in the combined firm.Basically, this resulted in a reverse take-over of the bidding firm.
The last column of Table 2 provides the distribution of the premiums paid by industry.Less than 10 of the above take-over deals (or 21 percent) were transacted at a discount (premiums less than 1).These companies were mainly from the service sector.About 35 percent of the targets were acquired with a premium of between 1 to 3 times the book value of the firms.For 57 percent of the targets, the premium paid was more than 3 times the book value with the trading and services sector and the property sector recording the highest number.In terms of percentage, in the consumer products sector 100 percent of the firms were paid premiums of more than 3 times the book value.
The majority of the targets were family-owned (77 percent).State-owned firms made up about 10 percent while others owned about 13 percent (Table 3).63 percent of the targets within the family-owned group received more than 3 times the book value of the firms, which is higher than for other types of ownership.In order to test any association between the types of ownership and the level of premiums paid, chi square test and One-way Anova were used.The Phi Cramer's V statistics shows that there is an association between types of ownership and the premiums received by the targets, at the level of significance of 7 percent if the premiums were classified as high for more than four times and as low otherwise.The Anova test shows that the link is marginally significant at the level of significance of 10 percent.The post hoc test shows that the premiums received by family-owned firms are significantly different from other types of ownership.

Results
This section attempts to ascertain the factors that could have affected the level of premiums paid.Table 4 shows the results of the estimation models for premiums paid.The analysis is divided into 2 parts: the overall model which consists of 134 observations and the sub-sample analysis where the firms were categorised into industrial products sector (n=24), trading and services sector (n=53), (Note 6) property sector (n=31) and others (n=26).The purpose of the sub-sample analysis is to see whether any systematic pattern persists in explaining the level of premiums paid for different sectors.All the models estimated were adequate as F-statistics shows that they are significant at the level of at least 5 percent.The adjusted R square for the overall model is at 42.5 percent.When the analysis is broken down into different industrial sectors, it shows higher percentages of 46, 48, 58 and 54 percent for the industrial products sector, the trading and services sector, the property sector and others, respectively.When an alternative measurement for premiums paid, (that is the purchase price deflated by total asset) was used, the explanatory power of the adjusted R square is slightly lower at 40.5 percent (details not reported here).The significance of individual coefficients remains unchanged except for MPAY.
In all the models, the targets' debt to total asset ratio (T_D_TA) has a significant positive impact on the premiums paid.Although the relatedness of businesses between target and bidder (RELATE) variable is not significant in the overall model, it has a significant impact on the industrial product sector.Average ROA (AVGROA) is able to explain the level of premiums paid in the industrial products sector and the trading and services sector.
The curvilinear relationship between premiums paid and bidder's ownership concentration was not found in the overall model.However, it was present in the trading and services sector where the premium paid was initially positively related to the bidders' ownership concentration but after a threshold of 42 percent, the relationship was negative.It shows that large shareholdings mitigate agency problem as premiums paid decreases as concentration of shareholding increases.Thus, the hypothesis that there is a positive relationship between premiums paid and ownership concentration at the lower level of ownership and a negative relationship at the higher level of ownership is supported for the trading and services sector.
Reverse take-over, which results in the creation of a new dominant block in the bidders (NEWBLOC) has a positive impact on the overall model.However, in the individual sectors, only in the trading and services sector, there was a marginal impact on the premiums paid.The ownership of bidders' directors or its substantial shareholders in targets (DIRO) in the overall model is not significant.However, it was marginally significant in the industrial product sector.Thus, the argument that related party transactions lead to the tunnelling out of corporate resources by paying high premiums at the expense of the minority shareholders is only weakly supported.
The control variables, namely CRIYEAR, MPAY and LOTAR_TA have significant impact on the premiums paid.However, when we breakdown the analysis into sub-sectors, only the asset size of the targets (LOTAR_TA) shows a significant negative relationship with the premiums paid, implying that the larger the size of the target, the greater is the constraints on the bidders to pay a higher premiums.This is supported by Palia (1993) and Shawky et al (1996) who argue that smaller targets are priced more competitively than bigger targets.This is because smaller targets will put less strain on the bidder's financial resources and will require less integration efforts.
When we introduce the industry dummy variables into the overall model, the trading and services sector shows a marginally significant positive impact on the premiums paid as compared to other sectors.By using the independent t-test, we found that significantly lower premiums were paid to acquire public listed targets as compared to privately held targets.
In order to further probe into the possible factors that could affect the levels of premiums paid, the interaction effect of the identity of the dominant owners and motives for take-over was examined.By using MPAY, NEWBLOC, LOTAR_TA and CRIYEAR as control variables, (Note 7) we find that family-owned firms tend to pay a lower premium as compared to other types of ownership, consistent with the hypothesis that family ownership mitigates the level of premiums paid.The interaction effect shows that bidding firms owned by family tend to pay a higher premium if the targets have higher ROA.Thus, we can infer that the motive for take-overs in the Malaysian capital market was more for enhancing the earnings base of the bidders rather than playing the disciplinary role for under-performed targets.Family-owned firms also tend to pay a higher premium if the debt to asset ratio of the target is higher, indicating that the motive for financial synergies do exist.The take-over of highly geared target is in order to gain the tax shield.However, it is only marginally significant.The interaction effect between the motive for operational synergies and family firms was not significant.Thus, the hypothesis that the motives for synergies in take-over transactions by family-owned firms mitigate the level of premiums paid is partially supported.

Concluding remarks
This study uses synergy variables (operational synergies, managerial synergies and the financial synergies), ownership variables (concentration and identity) and other control variables (method of payment, related party transaction, new block created, Asian financial crisis period and target size) to ascertain the determinants of premiums paid.Generally, premiums paid for a take-over in Malaysian is much higher than those in the developed countries.It is found that there is no difference in the level of premiums paid across industries.However, the various determinants have different impact on the level of premiums paid in different industries.
It is found that the targets' debt to asset ratio has a positive impact on all sectors except the industrial products sector.This implies that the bidders could have been motivated by the tax shield that comes from the targets' higher debt levels.Relatedness of businesses is important in determining the level of premiums paid especially in the industrial product sector.If the bidder attempts to diversify into other business, the levels of premiums paid would be higher.This might be due to the fact that as the Malaysian economy was moving towards a more service oriented economy, the growth opportunities in these sectors were higher; as such, a higher premium was warranted.Acquired related business and enjoyed operational synergy is not a major motivation for the bidders in the industrial sector.
Targets' size as measured by its total assets shows a significant negative impact on the premiums paid implying that the smaller the size, the less strain it will put on the bidders financial resources and requires less integration effort.This is supported by Palia (1993) and Shawky et al (1996) that smaller targets are priced more competitively than bigger targets.On the other hand, the new dominant block created in the combined firm as a result of the take-over has a positive impact on the level of premiums paid, especially in the trading and services sector.Given that the trading and services sector in fact is the growth sector in the economy in recent times, it would have attracted reverse take-overs that command higher premiums.
The curvilinear relationship between premiums paid and bidder's ownership concentration was not found in the overall model.However, it was present in the trading and services sector where premiums paid were positively related to bidders' ownership concentration up to a threshold of 42 percent, after which the relationship was negative.It shows that large shareholdings mitigate the agency problem as premiums paid decreased as concentration of shareholding increased.The argument that related party transactions result in tunnelling out corporate resources by paying a high premium was weakly supported only in the industrial product sector.
It is found that family ownership tends to pay a lower premium as compared to other type of ownership, consistent with the hypothesis that family ownership mitigates the level of premiums paid.Family-owned bidding firms tend to pay a higher premium if the targets have higher ROA.Thus, we can infer that the main motive of the bidders in the Malaysian take-over market in the 1990s was more towards enhancing the earnings base of the bidders rather than playing the disciplinary role for under-performed target.Besides, high concentration of ownership (by family) tends to align the interests of the owners to that of the shareholders (Jensen and Meckling, 1976) rather than resulting in the expropriation of minority shareholders (Claessens et al, 1999).

Notes
Note 1. Morck at al 1988 define an entrenched manager as a 'Manager who control a substantial fraction of the firm's equity may have enough voting power or influence more generally to guarantee his employment with the firm at an attractive salary' Note 2. Capital gain tax is not applied in Malaysia.
Note 3. It contains documents related to companies' announcements such as Circular to Shareholders in relation to take-overs, etc.
Note 4. Note that premiums compiled exclude those with negative values.
Note 5.About 30 percent of the market shares were held by government owned or government controlled banks.Note 6.One outlier Texchem Resource was excluded.
Note 7. AVGROA was dropped from the model as it is found to have high multicollinearity.* Significantly different from zero at the 10 percent level, using a two-tailed test.
** Significantly different from zero at the 5 percent level, using a two-tailed test.
*** Significantly different from zero at the 1 percent level, using a two-tailed test.

Table 1 .
Determinants of premiums paid with family ownership as moderating variablesSee Appendix 1 for the definition of the variables.