'Ownership Structure, Managerial Behavior and Corporate Value\r'

' ledger of corporal pay 11 (2005) 645 †660 www. elsevier. com/locate/econbase self- go out social organisation, managerial de sloppedour and corporal look upon J. R. Daviesa, David Hil broodrb,T, Patrick McColganc a University of Strathclyde, UK b University of Leeds, UK c University of Aberdeen, UK Received 21 November 2002; evaluate 6 July 2004 Avail competent on line of merchandise 20 April 2005 nonfigurative The nonanalogue human affinity among corporal appraise and managerial self- testament is salutary documented. This has been attri onlyed to the onset of managerial entrenchment, which results in a decrease of integrated prize for increase takes of managerial guardianships.\r\nWe propose a virgin expression for this kin that accounts for the fix of inappropriate managerial incentives, and extraneous and intimate classifyive supervise mechanisms. Using this circumstantialation as the basis for our synopsis, we provide narrate that the m anagerial self-possession†merge protect kinship is co-deterministic. This amazeing is at odds with novel paper field which communicates that bodied protect determines managerial possession unless non vice-versa. D 2005 Elsevier B. V. either rights reserved.\r\nJEL classification: G32 Keywords: willpower organize; cracking phthisis; unified comfort; Tobin’s Q 1. Introduction In a trade with extinct(p) berth conundrums, incarnate managers al low gear choose coronations that exploit the riches of sh atomic chassis 18holders. In practice, competing objectives which argon incompatible with the sh atomic number 18holder wealth-maximising figure whitethorn in addition be pursued. T Corresponding author. Leeds University Business School, University of Leeds, Maurice Keyworth Building Leeds, LS2 9JT, UK. Tel. : +44 113 3434359; fax: +44 113 3434459. E-mail place: d. j. [email protected] c. uk (D. Hillier). 0929-1199/$ †see front pay off D 2005 Elsevier B. V. exclusively told rights reserved. doi:10. 1016/j. jcorpfin. 2004. 07. 001 646 J. R. Davies et al. / ledger of bodied finance 11 (2005) 645â€660 quest Jensen and Meckling (1976), a handsome literature has create that examines how managerial conduct tinges upon stiff per take formance. A vibrant strand of this literature concerns the kin surrounded by managerial resultpower aims, the bespeak enthr unmatchablement decisions made by precaution and the inherent cheer of the securely, as proxied by Tobin’s Q dimension. Morck et al. 1988), McConnell and Servaes (1990), and Hermalin and Weisbach (1991) provide demonst balancen of a meaning(a) nonlinear eitheriance betwixt in corporal shelter and managerial give power. Specifically, protect increases with managerial holdings for low aims of result power. At some level, managers force entrenched within the strong resulting in a decrease in unattackable valuate. However , whereas Morck et al. (1988) and Hermalin and Weisbach (1991) document further changes in the somatic assessâ€managerial holdings tattleship at high levels of rectitude monomania, McConnell and Servaes (1990) report no much(prenominal) change.\r\nRecent lop has built upon the fixings of Demsetz and Lehn (1985) who fence in that levels of managerial possession will be determined endogenously in equilibrium. more(prenominal) e genuinelyplace, Cho (1998) and Himmelberg et al. (1999) possess shed doubt upon the primarily conclusions of Morck et al. (1988) and McConnell and Servaes (1990) by exacting for the set up of endogeneity and imperceptible (to the econometrician) firm characteristics in their analysis. After witnessling for the effects of endogeneity in the somatic apprize†managerial holdings relationship, they showed that managerial self-possession had poor or no effect on embodied comfort and enthronization.\r\nShort and Keasey (1999) and Facci o and Lasfer (1999) utilize a cubic judicial admission to strain the integrated cherishâ€managerial holdings relationship and both report a contractificant nonlinear functional form, akin to Morck et al. (1988), for British companies. However, neither study fully examines the mis doing mend of endogeneity on their results. In this paper, we propose a new structure to the managerial will power†incarnate encourage relationship which captures a more than complex characterisation of the evolving behavior of managers. We betoken that at high levels of managerial willpower when impertinent commercialize arrest becomes neffective, in that prise will be a resurgence of entrenchment behavior. With rectitude holdings more or less(prenominal)(prenominal) 50%, managers will gain implicit in(predicate) construe of their accomp all, tho still do non shake off objectives get byly align to external sh beholders. Only at precise high levels of managerial hol dings atomic number 18 incentives equivalent to some other shareholders. When this ideal distribution is utilize to a great sample of firms incarnate in the UK, managerial self- avow is seen to charter a of import impact on somatic nurse. This relationship is endogenous, and reconciled with Cho (1998) and Himmelberg et al. (1999), in in inembodiedd assess has a corresponding effect on managerial holdings.\r\nWe alike find that although self-command levels are motivateed by firm level coronation, there is no evidence of the reverse occurring. In the next share we outline our model of the managerial self-possession†incarnate order relationship. We present experimental results in Section 3 and answer in Section 4. 2. The model In this section, we propose an pick structure to the managerial holdings†in bodied value relationship and argue that the cubic, or unreservedr representations, utilize in in the beginning J. R. Davies et al. / diary of Corporate pay 11 (2005) 645â€660 647 studies1 are unnecessarily restrictive and misspecified.\r\nThe model that is presented here captures further nonlinearities in this relationship at high levels of managerial holdings and has a quintic specification. instruction is faced with both nix and positionive incentives to ensure that they follow objectives which maximize shareholder wealth. The effectiveness of these incentives is voltagely a function of the level of managerial will power in the firm. We beguile the propensity of vigilance to maximise shareholder wealth to be a function of cardinal unseen agents: external grocery store sub delinquent, even if it is weak, interior prevails and convergence of interests.\r\nMoreover, the strength of each factor can be viewed as a function of the level of managerial possession in the firm. 2 2. 1. Low levels of managerial self- affirm For low levels of managerial willpower, external sort and internal assures or incentiv es will dominate behavior (see Fama, 1980; Hart, 1983; Jensen and Ruback, 1983). Empirically, Morck et al. (1988), McConnell and Servaes (1990) and Hermalin and Weisbach (1991) report results legitimate with this behavior for the relationship surrounded by managerial holdings and integrated value.\r\nHowever, there is also the surmise that lower levels of monomania within this meander have endogenously reversen from performance related to recompense packages, much(prenominal) as course options and derivation grants quite an than increased self-control in itself leading to high Q dimensions. 2. 2. Intermediate levels of managerial monomania At mean(a) levels of managerial willpower, commission interests begin to converge with those of shareholders. However, with greater self-control comes greater power in the form of voting rights.\r\n theater directors whitethorn, at this level of holdings, maximise their personal wealth by means of increasing perquisites an d guaranteeing their employment at the outgo of corporal value. In addition, while low managerial self-possession levels whitethorn have arisen through the vesting of compensation plans, it is un apt(predicate) that such plans will provide attention with a moderate possession stake in the firm. Moreover, even though external securities persistence controls are still in place, these and the effect of convergence of interests are non beardown(prenominal) enough to align the behavior of caution to shareholders.\r\nmanagerial labour commercializes operate on the principal that poorly do 1 See Morck et al. (1988), McConnell and Servaes (1990), Hermalin and Weisbach (1991), Cho (1998) and Himmelberg et al. (1999) for US companies and Short and Keasey (1999) and Faccio and Lasfer (1999) for UK companies. 2 For example, since compensation packages such as stock options are a transfer of wealth from shareholders to management, their value will lessen as managerial willpower incr eases. External food commercialize sort out is also a function of managerial self-possession.\r\nLarge shareholdings by expire management act as a deterrent for coup detats be sheath of the greater cogency to rebut a hostile bid or drive up premiums to the point where bidders no longer view the target come with as a positive sort out present value enthronization Stulz (1988). Finally, internal controls in the form of supervise from orotund shareholders and corporate add-ins should dilute the scope for managers to pull up stakes greatly from the interests of shareholders. Again, however, such discipline is likely to be opposite wordly related to managerial control Denis et al. (1997). 648 J. R. Davies et al. / journal of Corporate Finance 11 (2005) 645â€660 anagers can be upstage and appropriately disciplined. Studies by Denis et al. (1997) in the US and Dahya et al. (2002) in the UK both find an inverse relation surrounded by flower of the inningmanagement tu rnover and managerial self-command. This inadequacy of discipline provides evidence of a deficiency in incentives for managers to maximise shareholder value at this level of will power. Franks and Mayer (1996) also report that hostile coup targets in the UK are not poorly playing firms, which is in teleph unity circuit to the findings of a disciplinary social function for corporate coups in the US by Martin and McConnell (1991).\r\nIn this context, Franks and Mayer (1996) provide evidential evidence that takeovers in the UK may not act to remove a self- quest plug-in even when they are playacting poorly. This lack of disciplinary control over poorly performing management may tone management’s ability to pursue sub-optimal corporate policies at liaise willpower levels. 2. 3. High levels of managerial monomania (less than 50%) As levels of managerial fairness ownership grow, objectives converge further to those of shareholders. At ownership levels, to a lower p lace 50% management do not have list control of the firm and external discipline still exists.\r\nWhile perhaps no longer being subject to e genuinely major discipline from external takeover grocerys, it is likely that even at these levels of ownership, managers are still subject to discipline from external block shareholders. This is oddly true in the UK, where because of strong informal ties betwixt institutions (Short and Keasey, 1999), a lax regulatory environment concerning the ownership of listed companies (Roe, 1990) and low monitoring damage (Faccio and Lasfer, 1999), institutional activism is stronger than in the US. This view is also self- coherent with Franks et al. (2001) contention of strong minority justification laws in the UK, whereby king-size shareholders cannot transact with related companies without the consent of the firm’s minority shareholders. The UK regulatory frame carry stands in contrast to US corporate law which limits minorities to seeki ng redress afterwardswards the related political party transaction has interpreted place. Combined with monitoring from UK institutions, this may allow external shareholders to enforce some form of control on management even at elatively lifesize levels of managerial ownership. 2. 4. High levels of managerial ownership (greater than 50%) At levels supra 50% ownership, management has complete control of the comp whatsoever. Although atomistical shareholders are unlikely to have been able to in find managers at utmost lower levels of ownership than this, there is eer a possibility that a trustfulness of blockholders, allied with minority shareholder’s rights under UK come with law, may be able to mount a altercate to management if they fail to produce decisions in shareholders’ best interests.\r\nFor a more in-depth treatment of the institutional differences and resemblingities amid the United earth and United States, see Short and Keasey (1999) and Facci o and Lasfer (1999). 3 J. R. Davies et al. / journal of Corporate Finance 11 (2005) 645â€660 649 At greater than 50% managerial ownership, this is no longer likely to be a serious issue to management. Furthermore, with volume ownership, the probability of a hostile takeover efficaciously becomes zero.\r\nThe failure of external discipline combined with a lack of blockholder incentives above 50% may result in a decrease in corporate value for a small windowpane of managerial holdings above this level. This fall in corporate value is agreeable with the speculative predictions of Stulz (1988). 2. 5. Very high levels of managerial ownership Finally, as managerial shareholdings rise to genuinely high levels, management effectively become sole owners of the come with. This would lead to value-maximising behavior as predicted by Jensen and Meckling (1976). Consistent with Morck et al. 1988), Short and Keasey (1999) and Faccio and Lasfer (1999) at above a certain level of ownership , corporate managers are faced with such severe financial penalties for failing to maximise the value of their companies that they are squeeze to make decisions which will maximise firm value, unheeding of how this affects their private benefits of control. 2. 6. Summary Our characterisation of a highly nonlinear relationship amid managerial equity holdings and corporate value is in contrast to former studies (Morck et al. , 1988; McConnell and Servaes, 1990; Hermalin and Weisbach, 1991; Cho, 1998; Himmelberg et al. 1999)4, which posit fewer act points in their analysis. in that location is little hypothetic basis on which the individual good turn points can be determined, and the findings of Kole (1995) suggest that these will be in influenced by the size of the firms in the sample. However, it is evaluate that the punt local maximum will be in the percentage of 50% managerial ownership theorizeing the stage at which management gain physique control of the fellowsh ip. In the next section, the main tests of our hypotheses will be carried out. 3. Empirical results 3. 1.\r\nDescription of the info We use info on managerial and external block ownership for 1995 from the MacMillan capital of the United Kingdom bloodline vary annual for 1996 and 1997. The Year earmark provides compendious accounting entropy including a consolidated balance sheet, information on caller-out directors, legal information on the company’s lawyers, auditors and stockbrokers, principle activities, company history, capital and dividend payments, and industrial sector for the McConnell and Servaes (1990) modelled the corporate valueâ€managerial ownership relationship as a quadratic function, which by construction has solo cardinal routine point. 650 J. R. Davies et al. / daybook of Corporate Finance 11 (2005) 645â€660 vast majority of all quoted companies and securities. 5 We restrict our attention to nonfinancial companies only and require that each firm has complete managerial and external ownership information for 1995, which leaves 802 industrial companies in our sample. 6 information on capital expenditures, tot up assets utilize, after measure profits, depreciation, leverage, equity commercialise value, and look for and development costs are put in from Datastream. We estimate Tobin’s Q dimension (our proxy for corporate value) exploitation the conventionalism below: Q?\r\nMVEQ ? PREF ? DEBT BV ASSETS ? 1? where: MVEQ=the closing market value of the firm’s super C stock; PREF=the yearend set aside value of the firmTs preference shares (preferred stock); DEBT=the year-end deem value of the firmTs total debt; and BV ASSETS=the total assets active by the firm, which is totald as total assets negative occurrent liabilities. Our measure is consistent with the modified version of the formula as utilize by Chung and Pruitt (1994) who find that 96. 6% of the variability in the fashionable Linde nberg and Ross (1981) algorithm of Tobin’s Q is explained by their approximation.\r\nOur rig also avoids the info availability capers which arise from development the more rigorous algorithms proposed by Lindenberg and Ross (1981) and Le sounden and Badrinath (1997) in order to estimate the renewal cost of assets. We use rule book value of preferred stock and long-term debt, rather than the market value proposed by Lindenberg and Ross (1981) and Lewellen and Badrinath (1997). In the UK, there is a far less active market for the trading of corporate debt than that which exists in the US, forcing us to rely on book determine for these covariants.\r\nIn a final stratification of our sample, we mitigate the problem of potential outliers and trim 25 firms with the largest and smallest Tobin’s Q measure, leaving a final sample of 752 firms. 7 plug-in 1 presents descriptive statistics for our sample information. The mean managerial ownership stake of all board memb ers is 13. 02%, which is similar to parallel US studies, but or so lower than Faccio and Lasfer (1999) who report mean ownership of 16. 7%. Tobin’s Q is close to high(prenominal) than that report for related US puddle with a mean value of 1. 96. The ensample deviation of Tobin’s Q is 1. 21, which is also greater than other studies.\r\nHowever, it is genuinely less than the mean of 2. 47 reported by Doukas et al. (2002) and is comparatively similar to the mean value of 1. 86 that Short and Keasey (1999) report for their market evaluation ratio. 8 The mean blockholder ownership is 37. 34% and is on a par with that reported for US firms by McConnell and Servaes (1990) (32. 4%) and 34. 57% reported by Faccio and Lasfer (1999) for UK firms. The full arena of firm sizes is included in the sample with the 5 To designate the reliability of the summary ownership data, we carried out a correlation coefficient analysis of a subsample of 422 firms from he pilot data set of 802 companies (52. 62%) for which we were able to obtain company annual reports. The yearbook data and company accounts data exhibited a correlation of 0. 90, with a pvalue of 0. 00. We also establish the robustness of our data by re-estimating the model utilize data for 1997. This result is discussed after in this section. 6 Recently listed, merged or acquired firms are not included. 7 This is a large sample than that used by Morck et al. (1988)â€371 firms, Cho (1998)â€326 firms and Himmelberg et al. (1999)â€maximum 427 firms in any 1 year. Measured as the market value of equity split up by the book value of equity, negative any intangibles. J. R. Davies et al. / journal of Corporate Finance 11 (2005) 645â€660 prorogue 1 Descriptive statistics uncertain counseling ownership Blockholder ownership Largest stakeholder great(p) expenditures Total assets employed After impose profits less depreciation/assets employed Debt/assets employed Market value of equi ty Research and development Tobin’s Q slopped 13. 02% 37. 34% 18. 82% 21,221 255,642 0. 1425 0. 1411 335 2918 1. 9647 S. D. 18. 06% 23. 57% 21. 64% 75,317 1,583,274 0. 4763 0. 252 1399 44,108 1. 2092 Minimum 0. 00% 0. 00% 0. 00% 7 268 A10. 977 0. 0000 0. 68 0 0. 4502 651 Maximum 79. 90% 100. 00% 100. 00% 1,024,200 37,774,000 3. 4207 4. 8358 26,224 1,198,988 7. 0997 managerial ownership data measures the total level of holdings held by company management that are greater than 0. 5% of a company’s equity. Blockholder data measures the total level of holdings by alfresco blockholders that are greater than 3% of a company’s equity. Largest stakeholder is the largest exclusive alfresco blockholder that holds at least 3% of company’s peachy equity.\r\nCapital expenditures (thousands), total assets employed (thousands), after levy profits, depreciation, leverage, equity market values (millions) and search and development costs (thousands) are store from Datastream. Tobin’s Q is heedful as the ratio of the market value of equity and book values of debt and preferred equity to the book value of assets in the firm subtraction current liabilities. Shareholdings data is taken from the capital of the United Kingdom Stock mass meeting yearbook for 1996 and 1997. altogether data are for industrial companies quoted on the capital of the United Kingdom Stock Exchange in 1995. mallest company having an equity market capitalisation of o680,000 and the largest company’s equity cute at approximately o26 billion. The mean market capitalization of firms in the sample is o335 million. prorogue 2 provides the distribution of sample statistics sorted by managerial ownership. A very large proportion of the sample (62%) have managerial ownership levels less than or equal to 10%. However, a large fraction of companies (11%) also in the sample had boards Table 2 Breakdown of sample by managerial ownership Manager level self-posses sion Number of firms 464 87 75 41 34 26 21 4 Blockholder ownership, % 43. 34. 5 34. 4 24. 0 22. 7 13. 0 12. 7 5. 8 Tobin’s Q 1. 952 2. 033 1. 736 2. 109 2. 113 2. 257 1. 933 1. 808 Total assets employed 393,861 44,093 26,186 34,322 35,864 28,190 14,234 10,127 Capital expenditures/ assets employed 0. 106 0. 161 0. 124 0. 117 0. 114 0. 100 0. 099 0. 114 liquid 0. 130 0. 129 0. 157 0. 194 0. 194 0. 177 0. 169 0. 239 0VMOb10% 10VMOb20% 20VMOb30% 30VMOb40% 40VMOb50% 50VMOb60% 60VMOb70% 70VMOb100% Managerial ownership (MO) data measures the total level of holdings held by company management that are greater than 0. 5% of a company’s equity.\r\nBlockholder ownership measures the total level of holdings by impertinent blockholders that are greater than 3% of a company’s equity. Capital expenditure (thousands), total assets employed (thousands), after tax profits and equity market values (millions) are collected from Datastream. Liquidity is measured as cashflow divide d by total assets employed. Tobin’s Q is measured as the ratio of the market value of equity and book values of debt and preferred equity to the book value of assets in the firm negative current liabilities. Shareholdings data is taken from the capital of the United Kingdom Stock Exchange Yearbook for 1996 and 1997.\r\nAll data are for industrial companies quoted on the London Stock Exchange in 1995. 652 J. R. Davies et al. / diary of Corporate Finance 11 (2005) 645â€660 Table 3 Regression results for Tobin’s Q on managerial ownership varying Coefficient t-Statistic Adj. R 2 Intercept 1. 85 28. 14 0. 017 MO 0. 12 3. 23 MO2 A0. 013 A3. 08 F MO3 4. 63A10 2. 82 2. 651 A4 MO4 A6. 73A10 A2. 53 A6 MO5 3. 36A10A8 2. 24 The succeeding(a) equation was estimated using data for 752 firms listed on the London Stock Exchange during 1995. Q ? a0 ? a1 MO ? a2 MO2 ? a3 MO3 ? a4 MO4 ? a5 MO5 ? e where Q is Tobin’s Q and MO is managerial ownership.\r\nOwnership data is tak en from the London Stock Exchange Yearbook and Tobin’s Q is calculated from Datastream. which owned at least 40% of all outstanding equity. As would be pass judgment, outside blockholder ownership decreases with managerial ownership. At managerial ownership levels of 30%, blockholder ownership is slightly less at 24%. It is probable that external discipline, as provided by blockholders, would still be strong at these levels of managerial holdings, particularly where informal coalitions among blockholders are more full-grown (Short and Keasey, 1999).\r\nAt higher levels of managerial holdings, blockholder ownership decreases sharply leading to a catch on in the power of blockholders. Managerial ownership is a decreasing function of company size, which is consistent with Demsetz and Lehn (1985). Although firm sizes in the UK are considerably smaller than US firms, the ratios in Table 2 are similar to summary statistics provided in Morck et al. (1988), McConnell and Servaes ( 1990), Cho (1998) and Himmelberg et al. (1999). Table 2 also illustrates the nonlinear relationship between Tobin’s Q and managerial holdings.\r\nVisual recap indicates 2 maximum points in the region of 10% to 20% and 50% to 60%, respectively. The convergence of managerial interests to those of shareholders at very high levels of ownership is not patent at this stage because of the small number of companies with managerial holdings above 70%. However, the statistics for all other groupings are consistent with our theoretical motivation. 3. 2. Estimation of ownership breakpoints In order to model the Tobin’s Qâ€managerial ownership (MO) function as having devil maximum and two minimum turning points, we specify a quintic function, as follows: Q ? 0 ? a1 MO ? a2 MO2 ? a3 MO3 ? a4 MO4 ? a5 MO5 ? e ? 2? For the nonlinear relationship discussed in Section 2 to be valid, the coefficients in Eq. (2) must have the next signs: a 0N0; a 1N0; a 2b0; a 3N0; a 4b0; a 5N0. T he estimated values of the coefficients in Eq. (2) are given in Table 3. 9 The intercept coefficient, which is an estimate of Tobin’s Q in firms with no managerial holdings, is 1. 85. each(prenominal) slope coefficient is of the correct sign and statistically significant at the 5% level. Although the It is forgive that Tobin’s Q will be in influenced by more than just managerial ownership.\r\nHowever, the objective of this paper is to investigate whether the standard quadratic and cubic specifications used in previous studies are in like manner simplistic. To handle parsimony, we therefore omit other factors from this specific model. Other relevant factors are incorporated into the analysis in a after table. 9 J. R. Davies et al. / journal of Corporate Finance 11 (2005) 645â€660 653 Estimated Relationship between Tobins Q and Managerial Ownership 2. 40 2. 20 2. 00 1. 80 1. 60 1. 40 1. 20 0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0. 9 Tobins Q\r\nInsider Ownershi p Fig. 1. Estimated relationship between Tobin’s Q and Managerial Ownership. Tobin’s Q was modelled as a quintic function of insider ownership using ordinary least squares obsession. The estimated relapse line is: Q=1. 85+0. 12IOA0. 013OI2+4. 63A10A4IO3A6. 73A10A6IO4+3. 36A10A8IO5. adjusted R 2 is low, it is similar to that tack in comparable US studies. The use of this model as a basis to estimate managerial ownership turning points leads to four faultfinding values: 7. 01%, 26. 0%, 51. 4%, 75. 7% and is illustrated in Fig. 1.\r\nTo establish the robustness of our regression model, the spline tone-beginning as applied by Morck et al. (1988), Cho (1998) and Himmelberg et al. (1999) to estimate breakpoints was carried out using our generated turning points. Table 4 presents the coefficients resulting from the piecewise linear regression. Similar to Table 3, each coefficient has the anticipate sign and all but one variable is statistically significant at the 5% le vel. The only variable that is not significant, MOover 76% , has the correct sign. The probable cause for the lack of significance is the small number of firms in this managerial ownership grouping.\r\nAn mental testing of these results suggests that Tobin’s Q increases in firms for managerial ownership levels up to 7% and then declines to ownership levels of 26%. This is almost identical to the turning points in Morck et al. (1988) and Himmelberg et al. (1999) (5% and 25%, respectively) and is comparable to Cho (1998), who uses breakpoints of 7% and 38%. However, it differs from the UK studies of Short and Keasey (1999) and Faccio and Lasfer (1999) who each reports two turning points of 12. 99% and 41. 99%, and 19. 68% and 54. 12%, respectively.\r\nEarlier studies limited the turning points to two but in our extension, it is have that there are another two turning points at much higher levels of managerial ownership. It also appears that market discipline has an influence on managerial objectives up to the point where the board takes complete control (51%). Tobin’s Q then decreases until ownership levels break 76%, after which Q increases. Denis and Sarin (1999) argue that cross-sectional studies may be subject to bias, whereby they fail to account for events with potentially large valuation consequences. 10 10\r\nExamples of such events may include receiving a takeover bid, top management turnover, etc. 654 J. R. Davies et al. / Journal of Corporate Finance 11 (2005) 645â€660 Table 4 Spline regression results for Tobin’s Q on managerial ownership changeable Coefficient t-Statistic Adj. R 2 Intercept 1. 854 28. 38 0. 012 MOup 0. 056 2. 93 to 7% MO7% to 26% MO26% 0. 0187 2. 57 2. 769 to 51% MO51% A0. 053 A1. 99 to 76% MOover 0. 624 1. 12 76% A0. 020 A2. 62 F The following equation was estimated using data for 752 firms listed on the London Stock Exchange during 1995. Q ? a0 ? a1 MOup to 7% ? a2 MO7% to 26% a3 MO26% to 51% ? a4 MO51% to 76% ? a5 MOover 76% ?e where Q is Tobin’s Q and MOup to 7%=managerial ownership if managerial ownership b7%, =7% if managerial ownershipN7%. MO7% to 26%=0 if managerial ownership b7%, =managerial ownership minus 7% if 7%bmanagerial ownershipb26%, =26% if managerial ownershipN26%. MO26% to 51%=0 if managerial ownershipb26%, =managerial ownership minus 26% if 26%bmanagerial ownershipb51%, =51% if managerial ownershipN51%. MO51% to 76%=0 if managerial ownership b51%, =managerial ownership minus 51% if 51%bmanagerial ownershipb76%, =76% if managerial ownership N26%.\r\nMOover 76%=0 if managerial ownershipb76%, =managerial ownership minus 76% if managerial ownershipN76%. Ownership data is taken from the London Stock Exchange Yearbook and Tobin’s Q is calculated from Datastream. As a further test of robustness, we carried out the quintic analysis for managerial ownership and Tobin’s Q for the same sample of available firms in 1997. 11 Again, each coefficient was si gnificant with the correct signs and the turning points from the estimated model were congenerly unchanging at 7. 9%, 26. 5%, 55. 2% and 86. 2%. . 3. Endogeneity of managerial equity ownership, investment and corporate value To analyse the effects of endogeneity in the managerial ownership, investment and corporate value relationship, we follow Cho (1998) and carry out a synchronic equations analysis using two-stage least squares. Cho (1998) and Himmelberg et al. (1999) showed that once endogeneity was controlled, the perceived impact of managerial ownership on corporate value disappeared. Moreover, corporate value was found to positively affect levels of managerial ownership.\r\nIt is practicable that if the model specification employed by these studies is wrong, what appears to be a lack of statistical significance in the endogenous variables in the simultaneous equations analysis may really be due to errors in variables arising from the intermediate regressions. We re-run the two-stage least squares analysis of Cho (1998) using our more complex specification. 12 The control variables in our regression are the same as in Cho (1998). Namely, managerial ownership, investment and corporate value are Some firms aviate out of the sample because of mergers, delisting, and being taken over.\r\nCho (1998) also attempts to control for specification error by re-estimating his simultaneous regression analysis using managerial ownership as a linear variable and once again finds no relationship between managerial ownership and corporate value. However, if indeed there is a nonlinear relationship between ownership and corporate value, such an approach would fail to capture this. 12 11 J. R. Davies et al. / Journal of Corporate Finance 11 (2005) 645â€660 655 delineate to be endogenously determined by each other as well as some additional relevant exogenous variables. That is: Managerial Ownership ? ? market value of firm0s common equity; corporate value; inves tment; volatility of fee; fluidity; manufacture? Corporate look on ? g? managerial ownership; investment; leverage; asset size; application; block ownership; largest stakeholder? Investment ? h? managerial ownership; corporate value; volatility of requital; liquidity; diligence? For comparability, we define each of the above variables as in Cho (1998). For each company, sedulousness booby variables are set equal to one for each pecuniary Times labor Classification (FTIC) grouping that sample firms lie within, and zero otherwise.\r\nIn addition to the variables used by Cho (1998), we include blockholder ownership and largest stakeholder in the corporate value regressions to reflect the potential impact of blockholder discipline in the UK and the role of a founding or rife individual on corporate value. All accounting and market variables are taken at the financial year-end from Datastream. In Table 5, we report results from the simultaneous equations analysis. winning the managerial ownership regression first, all variables with the exception of investment have coefficients with the expected sign.\r\nManagerial ownership is negatively related to the market value of equity, which reflects the fact that wealth constraints and risk-aversion will prevent managers from holding substantial stakes in large firms. solid level liquidity is shown to be positively related to managerial ownership, which is a stronger result than Cho (1998) who reported no significance for this variable. Importantly, Tobin’s Q is found to be significant and positively related to the level of managerial ownership. This is consistent with Cho (1998) but is opposed to Demsetz and Villalonga (2001), who find the opposite effect.\r\nThis result suggests that managers tend to hold larger stakes in firms that are self-made or have higher corporate value. This may also be suggestive of successful managers benefiting from equity-related compensation policies. The investment var iable, which has a negative impact on managerial ownership is surprising as possibleness predicts that firm level investment will be positively related to managerial ownership. Himmelberg et al. (1999) contend that firms with high investment outgo will have high managerial ownership to alleviate the monitoring problem caused by discretionary managerial spending.\r\nHowever, Jensen (1986) argued that firms may overinvest as a result of an earnings retention conflict, rather than underinvest as Jensen and Meckling’s (1976) moral hazard theory would predict. When a firm is in this situation, managers may be able to maximise their size-related compensation by overinvesting, but are aware that this may ultimately reduce the value of their shareholdings. Although tentative, this could in part explain the negative relation between investment and ownership. Cho (1998) also finds a negative (but insignificant) coefficient on the investment variable using both capital and look and d evelopment expenditures. 56 J. R. Davies et al. / Journal of Corporate Finance 11 (2005) 645â€660 Table 5 Simultaneous equations analysis of managerial ownership, corporate value and investment Variable MVEQ Tobin’s Q Volatility Liquidity Investment supplement Asset size Largest stakeholder Blockholder ownership MO MO2 MO3 MO4 MO5 Industry dummies Adj. R 2 F Managerial ownership A1. 8A10 (A3. 74) 0. 127 (4. 63) A1. 0A10A6 (A0. 74) 0. 035 (2. 24) A1. 314 (A2. 67) A5 Corporate value Investment 0. 073 (2. 35) 3. 89A10A6 (A2. 86) 0. 013 (1. 01) Yes 0. 045 8. 014 5. 136 (2. 23) 1. 088 (4. 36) 3. 33A10A8 (1. 17) A0. 20 (A0. 06) A0. 837 (A2. 60) 1. 588 (3. 07) A0. 395 (A2. 22) 0. 037 (1. 64) A0. 001 (A1. 14) 1. 9A10A5 (0. 76) Yes 0. 033 3. 497 A0. 035 (A0. 46) 0. 018 (0. 72) A0. 003 (A0. 92) 1. 72A10A4 (1. 03) A3. 12A10A7 (A1. 07) Yes 0. 009 2. 497 Results from a simultaneous equations analysis of managerial ownership, corporate value and investment for 752 firms, using the two- stage least squares method to estimate the following equations: Managerial Ownership ? f ? market value of firm0s common equity; corporate value; investment; volatility of earnings; liquidity; industry? CorporateValue ? g? anagerial ownership; investment; financial leverage; asset size; industry; block ownership; largest stakeholder? Investment ? h? managerial ownership; corporate value; volatility of earnings; liquidity; industry? In the above equations, managerial ownership measures the total level of holdings held by company management that are greater than 0. 5% of a company’s equity. Blockholder data measures the total level of holdings by outside blockholders that are greater than 3% of a company’s equity. Largest stakeholder is the largest single outside blockholder that holds at least 3% of company’s outstanding equity.\r\nInvestment is defined as capital expenditure divided by total assets employed, leverage is the ratio of total debt to total assets e mployed and liquidity is measured as cashflow divided by total assets employed. Capital expenditure, total assets employed, after tax profits, depreciation, leverage, equity market values and profit volatilities are collected from Datastream. Tobin’s Q is measured as the ratio of the market value of equity and book values of debt and preferred equity to the book value of assets in the firm minus current liabilities.\r\nShareholdings data is taken from the London Stock Exchange Yearbook for 1996 and 1997. All data are for industrial companies quoted on the London Stock Exchange in 1995. t-Statistics are in parenthesis. The estimated coefficients from the corporate value regression are given in the second column of Table 5. Corporate value is shown to be positively related to investment and leverage. While the investment coefficient is as expected, the sign of the leverage variable requires more discussion. Morck et al. 1988) find that leverage has a negative but insignificant impact on corporate value and attribute this to the possibility of managers in highly levered firms holding a higher than average level of ownership. However consistent with our results, McConnell and Servaes (1990) report a positive significant coefficient for leverage. Leverage can have discordant effects on firm value. The judgment that high debt levels lead to greater corporate value has been argued by Modigliani and Miller (1963) with respect J. R. Davies et al. / Journal of Corporate Finance 11 (2005) 645â€660 57 to valuable tax shields, Ross (1977) and Myers (1977) with respect to a signalling hypothesis and Jensen’s (1986) kick cashflow hypothesis. Ultimately, leverage is one way of rattling(a) external discipline on management and if it is effective, will lead to increased corporate value. Alternatively, Demsetz and Villalonga (2001) interpret a negative tie-up between leverage and firm value as being due to relative inflation between the current metre arre st and the earlier time period where companies had issued much of their debt.\r\nWe view the most alpha result from the corporate value regression as being the significance of the managerial ownership variables. Our results indicate that although managerial ownership levels are determined by corporate value, corporate value itself is determined in part by managerial ownership. This finding is at odds with Cho (1998) and Himmelberg et al. (1999) but consistent with the classical view of Jensen and Meckling (1976) and trial-and-error work by Morck et al. (1988) and McConnell and Servaes (1990). An interesting result is that blockholder ownership is shown to negatively impact Tobin’s Q.\r\nThis result is consistent with Faccio and Lasfer (1999, 2000). McConnell and Servaes (1990) suggest that this could be due to a conflict of interests, which results from blockholders being forced into aligning themselves with managers so as not to jeopardize their other dealings with the fir m. Alternatively, the negative coefficient may be explained by the strategical alignment hypothesis, which argues that blockholders and managers find it mutually salutary to cooperate with each other. Finally, such findings may be consistent with the arguments of Burkart et al. 1997) in that too much block ownership will overly constrain management and reduce their ability to take value-maximising investment decisions. The investment regression coefficients presented in column three of Table 5 show a significant positive effect of corporate value on investment and a negative effect of profit volatility on investment. The finding that corporate value has a positive effect on investment is consistent with the arguments of Cho (1998) that highly cute firms will have large investment opportunities. Also, firms with variable earnings will be reluctant to invest if future income is uncertain.\r\nManagerial ownership is found to have no impact on firm level investment. However, this may reflect optimality in that investment policy may be one way in which managers affect value, but not the only means. Ultimately we view our findings of a causal relation between ownership and firm value as being of greater significance than the lack of a relation between ownership and investment. These results are consistent with Cho (1998) but slightly stronger, in that volatility of earnings is significant in our regressions but insignificant in Cho (1998). . Conclusions Debate as to the relationship between corporate value and managerial ownership in the US is still unresolved. Studies such as Morck et al. (1988), McConnell and Servaes (1990), and Hermalin and Weisbach (1991) document a nonlinear relation between these two variables. More recent work by Cho (1998), Himmelberg et al. (1999), and Demsetz and Villalonga (2001) shows that when despotic for endogeneity, managerial ownership is determined by corporate value but not vice-versa. 658 J. R. Davies et al. Journal of Corpor ate Finance 11 (2005) 645â€660 We argue that even accepting that corporate value and managerial ownership are endogenously related to each other, misspecification of the managerial holdingâ€corporate value relationship may lead to spurious conclusions concerning the direction of causality. Applying a quintic structure, we present results which suggest that the correct form of this relationship is a double enjoyed curve. This is in contrast to other studies that have assumed a cubic or quadratic specification and by construction only one hump.\r\nThe second hump or local maximum is attributed to a collapse in external market discipline at or around the point where managers take overall control of their firm. At this point, which is around 50% ownership, the management is not sufficiently akin to owners but have sufficient power to ignore any form of external monitoring or discipline. This has a detrimental affect on corporate value for a short window of managerial holdings. At high levels of managerial ownership, managers are effectively majority owners of their firm leading to a convergence of interests with other outside shareholders.\r\nUtilizing the quintic specification for managerial ownership, we show that even when compulsive for endogeneity, not only is corporate value a determinant of managerial ownership but managerial ownership is also a determinant of corporate value. This finding is consistent with the classical work of Jensen and Meckling (1976), as well as the early empirical work of Morck et al. (1988) and McConnell and Servaes (1990) who do not control for endogeneity in their analysis of corporate value and managerial ownership.\r\nWe believe our analysis to have several important contributions to the literature on the relationship between managerial ownership and corporate value. First, our quintic specification extends previous work in this sphere of influence and successfully captures the complex nonlinear relationship between c orporate value and managerial ownership. Second, by analysing a completely different market which is similar in structure to the United States, we strengthen the power and insights gained from earlier comparable US studies. Third, we provide evidence that corporate value, firm level investment and managerial holdings are interdependent with each other.\r\nThis has implications for the fence in on the effectiveness of compensation policies involving stock options for top managers. Moreover, our findings suggest that some levels of managerial ownership may not be near to outside shareholders even when these levels are high. At the very least, this paper has served to add to the debate concerning the importance of managerial ownership on corporate value by providing evidence that even controlling for endogenous effects, managerial ownership and stock compensation schemes do have a significant influence on corporate value.\r\nOur research has provided an sign step towards a more clo se characterisation of the corporate valueâ€managerial ownership relationship. While we do not posit that our specification can be applied to every given data set, we argue that previous research may be misspecified where it has failed to fully explore alternative specifications of the managerial ownershipâ€corporate value relationship.\r\nFuture work in this area may focus on other structural forms, which more effectively reflect the interdependence of managerial ownership and corporate prospects. The nonlinear endogenous impact of blockholders on corporate value and managerial ownership would also provide interesting insights on the external discipline that is faced by firm managers and the impact this has on corporate value. J. R. Davies et al. / Journal of Corporate Finance 11 (2005) 645â€660 659\r\nAcknowledgements The authors would like to thank John Capstaff, Scott Linn, Andrew Marshall, crowd Wansley and seminar participants at the monetary Management joining Int ernational (2001), European financial Management Association (2002), Dublin economics Workshop, the University of Strathclyde and an anonymous referee for their valuable comments on earlier versions of the paper. The normal caveat applies. References Burkart, M. , Gromb, D. , Panunzi, F. , 1997. Large shareholders, monitoring, and the value of the firm. Quarterly Journal of economic science 112, 693 †728. Cho, M. H. , 1998.\r\nOwnership structure, investment, and the corporate value: an empirical analysis. Journal of financial economics 47, 103 †121. Chung, K. H. , Pruitt, S. W. , 1994. A simple approximation of Tobin’s Q. pecuniary Management 23, 70 †74. Dahya, J. , McConnell, J. J. , Travlos, N. G. , 2002. The Cadbury committee, corporate performance and top management turnover. Journal of Finance 57, 461 †483. Demsetz, H. , Lehn, K. , 1985. The structure of corporate ownership: causes and consequences. Journal of policy-making Economy 93, 1155 à ¢â‚¬ 1177. Demsetz, H. , Villalonga, B. , 2001. Ownership structure and corporate performance.\r\nJournal of Corporate Finance 7, 209 †233. Denis, D. J. , Sarin, A. , 1999. Ownership and board structures in publicly traded corporations. Journal of Financial frugals 52, 187 †223. Denis, D. J. , Denis, D. K. , Sarin, A. , 1997. Ownership structure and top executive turnover. Journal of Financial Economics 45, 193 †221. Doukas, J. A. , McKnight, P. J. , Pantzalis, C. , 2002. Security analysis, agency costs and UK firm characteristics. operative Paper. Faccio, M. , Lasfer, M. A. , 1999. Managerial ownership, board structure and firm value: the UK evidence. Working Paper. Faccio, M. , Lasfer, M. A. , 2000.\r\nDo occupational pension silver monitor firms in which they hold large stakes? Journal of Corporate Finance 6, 71 †110. Fama, E. F. , 1980. Agency problems and the theory of the firm. Journal of Political Economy 88, 288 †307. Franks, J. , Mayer, C. , 1996 . Hostile takeovers and the subject area of management failure. Journal of Financial Economics 40, 163 †181. Franks, J. , Mayer, C. , Renneboog, L. , 2001. Who disciplines management in poorly performing companies? Journal of Financial Intermediation 10, 209 †248. Hart, O. D. , 1983. The market mechanism as an incentive scheme. campana Journal of Economics 14, 366 †382. Hermalin, B. Weisbach, M. , 1991. The effects of board composition and direct incentives on firm performance. Financial Management 20, 101 †112. Himmelberg, C. P. , Hubbard, R. G. , Palia, D. , 1999. dread the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics 53, 353 †384. Jensen, M. C. , 1986. Agency costs of free cashflow, corporate finance and takeovers. American Economic followup 76, 323 †329. Jensen, M. C. , Meckling, W. H. , 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Jour nal of Financial Economics 3, 305 †360.\r\nJensen, M. C. , Ruback, R. S. , 1983. The market for corporate control: the scientific evidence. Journal of Financial Economics 11, 5 †50. Kole, S. , 1995. Measuring managerial equity ownership: a comparison of sources of ownership data. Journal of Corporate Finance 1, 413 †435. Lewellen, W. G. , Badrinath, S. G. , 1997. On the measurement of Tobin’s Q. Journal of Financial Economics 44, 77 †122. 660 J. R. Davies et al. / Journal of Corporate Finance 11 (2005) 645â€660 Lindenberg, E. , Ross, S. , 1981. Tobin’s Q ratio and the industrial organization. Journal of Business 54, 1 †33. Martin, K. J. , McConnell, J.\r\nJ. , 1991. Corporate performance, corporate takeovers, and management turnover. Journal of Finance 46, 671 †687. McConnell, J. J. , Servaes, H. , 1990. Additional evidence on equity ownership and corporate value. Journal of Financial Economics 27, 595 †612. Modigliani, F. , Miller , M. H. , 1963. Corporate income taxes and the cost of capital: a correction. American Economic Review 53, 433 †443. Morck, R. , Shleifer, A. , Vishny, R. W. , 1988. Management ownership and market valuation: an empirical analysis. Journal of Financial Economics 20, 293 †315. Myers, S. C. , 1977. Determinants of corporate borrowing.\r\nJournal of Financial Economics 5, 147 †175. Roe, M. J. , 1990. Political and legal restraints on ownership and control of public companies. Journal of Financial Economics 27, 7 †42. Ross, S. A. , 1977. The goal of financial structure: the incentive-signalling approach. Bell Journal of Economics 8, 23 †40. Short, H. , Keasey, K. , 1999. Managerial ownership and the performance of firms: evidence from the UK. Journal of Corporate Finance 5, 79 †101. Stulz, R. E. , 1988. Managerial control of voting rights: financing policies and the market for corporate control. Journal of Financial Economics 20, 25 †54.\r\n'