DO SHAREHOLDERS CARE ABOUT CORPORATE INVESTMENT RETURNS? * OLLI HALTIA MIKKO LEPPÄMÄKI

Transcription

1 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki Fiish Ecoomic Papers Volume 13 Number 1 Sprig 2 DO SHAREHOLDERS CARE ABOUT CORPORATE INVESTMENT RETURNS? * OLLI HALTIA Europea Ivestmet Bak, 1, boulevard Korad Adeauer, L-295 Luxembourg ad MIKKO LEPPÄMÄKI Research Uit o Ecoomic Structures ad Growth, Departmet of Ecoomics, P.O. Box 54, FIN-14 Uiversity of Helsiki, Filad This paper cosiders the apparet cotradictio betwee the results of Artto (1997), who claims that shareholders i the paper idustry have gaied reasoable returs, ad Pohjola (1996), who argues that a large shareholder value has bee lost (the Artto Pohjola Paradox). We show uder fairly geeral coditios that the shareholders may ejoy reasoable retur while the maagers are simultaeously destroyig value of the firm by allocatig fuds to bad ivestmets. I this case the firm s shareholders suffer a opportuity loss equal to the value that could have bee created if the firm had paid the fuds out to them ad they had ivested the fuds i equivaletly risky projects. The paradox occurs if the growth rate of a firm s market value of equity is high eough to guaratee a oegative retur for the shareholders but too low i the sese that the shareholders would have eared more had the fuds ivested at the retur (at least) equal to the opportuity cost of capital. Our results support Jese s (1986) argumet of the icetives of corporate maagers to ivest iefficietly, sice here the shareholders do ot ecessarily challege the maagemet due to the fact that they may be perfectly happy ad satisfied i fiacial terms. (JEL: D24, G31, L73) 1. Itroductio * We thak Seppo Hokapohja, Michael C. Jese, Erkki Koskela, Juha-Pekka Niiimäki, the editor ad two aoymous referees of this joural for their commets. The views expressed are the sole resposibility of the authors, ad do ot ecessarily reflect the view of the Europea Ivestmet Bak (EIB). Do shareholders care about corporate ivestmet returs? Is it possible that the shareholders are ejoyig a reasoable retur while the firm s maagers are simultaeously destroyig the value of the firm by makig bad ivestmets? It will be show i the paper that this situatio, although apparetly cotradictory, is ideed possible. I this case the firm s shareholders suffer a opportuity loss equal to the value that could have bee created if the firm had paid the fuds available for ivestmet withi the firm out to the shareholders ad they 19

2 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki 2 had ivested the fuds i equivaletly risky projects. This situatio occurs especially if the shareholders judge the performace of the firm s maagers oly based o the shareholders retur whilst the losses due to the opportuity cost of capital are ot trasparet or observable to the shareholders. The result implies that the shareholders should ot judge the performace of the maagers oly based o the shareholders retur but also based o the productivity of capital, which is directly cotrollable by the maagers. I other words, the shareholders should esure that the maagers ivest at the retur at least equal to the opportuity cost of capital. It is well kow that the log-term ecoomic growth depeds crucially o the productivity of tagible ad itagible capital. A typical feature for a developmet path of ay ecoomy has bee a gradual move from the strategy of extesive growth, relyig o expadig use of atural ad other resources, towards the strategy of itesive growth, based o cotiuous improvemet of efficiecy i the use of existig productio factors. Yet, as idicated by the recet study of McKisey Global Istitute (1996), there are huge differeces i capital productivity eve amog the most developed atios across various idustries. I additio, capital productivity may vary drastically eve betwee the firms i the same idustry. Jese (1993) argues that low productivity durig the last two decades may have bee caused by a failure of corporate iteral cotrol systems. Corporate cotrol has bee ieffective i dealig with a slow dow i the markets, excess capacity ad exit. It should be metioed that our results further uderlie Jese s origial cocer of the importace of the iteral corporate cotrol systems. While Jese s (1986) propositio is based o the icetives of maagers to ivest iefficietly, our fidigs further support this argumet. It will be show i the paper that the shareholders do ot ecessarily challege the maagemet sice they may be perfectly happy ad satisfied i fiacial terms. 1 The productivity ad performace of the Fiish idustry has recetly bee examied by Pohjola (1996). Applyig the method established by Jese (1993), Pohjola fids that i may idustrial sectors, especially i the paper idustry, the productivity of ivestmets i has bee far too low from the poit of view of the shareholder value geerated. Based o Pohjola s fidigs oe could coclude that the shareholders would have lost up to FIM 45 b (some EUR 7.5 b) over the period i terms of opportuity cost. I a apparet cotradictio with Pohjola s fidigs, Artto (1997) shows that the shareholders i the paper idustry should have ejoyed reasoable returs (aroud 15% IRR) over the period Cosequetly, we have two views, oe of which claims that a large shareholder value has bee lost whereas the other cocludes that shareholders have gaied reasoable returs. This apparet cotradictio will be called the Artto Pohjola paradox. 2 The remaider of this paper is orgaised as follows. I sectio 2 the opportuity costs of capital are modelled i terms of cotiuous time based o the so-called Jese s method. Sectio 2 also derives the shareholders retur ad relates that to the capital productivity measures. I sectio 3 we derive the sufficiet coditio for the Artto Pohjola Paradox. Sectio 4 presets further aalysis o the ecessary coditios for the paradox. Sectio 5 cocludes the paper. 2. The opportuity cost of capital ad the shareholders retur As argued by Jese (1986), shareholders ad corporate maagers may have coflictig iterests. I particular, the maagers may have icetives to cause their firm to grow beyod the optimal size by realisig ivestmets, which do ot yield positive et preset values whe 1 See also the recet iterestig cotributio by Kaiaie (2) who cosiders corporate maagers iefficiet ivestmet behaviour uder separatio of owership ad cotrol. 2 Give the importace of the paper idustry i the Fiish ecoomy (the paper idustry still accouts for about 25% of total exports), these cotroversial results have ot oly caused a academic debate but also bee a source of public cocer. See, for example, the Helsigi Saomat 19 March 1998.

3 discouted at the relevat cost of capital. I other words, the shareholders would be better off if the maagers would disgorge the cash to shareholders rather tha ivest it iefficietly. Thus a coflict may arise betwee the divideds to be paid out versus the fuds to be ivested withi the firm. Followig this lie of thought, we employ Jese s (1993) capital productivity measures ad relate them to the explicit fiacial returs ejoyed by shareholders. A firm ca be cosidered as a value creatig process for its shareholders ad for the bakig sector (fiacig the corporate debt), i.e. as a process creatig value o the firm s capital structure. I geeral, a compay geerates shareholder value through two ways: (1) by a stream of divideds; ad (2) by icreasig the market value of the firm. I additio, a firm creates value for the bakig idustry i terms of payig the debt. Therefore, the total value V T created by the firm up to date ca be expressed as follows: 3 (1) V T =V + δ t e ρ( t) dt + β t e r( t) dt, where V = value of the firm at the future horizo date (> ); δ t = paymets to shareholders i the form divideds ad et share repurchases (> ); β t = debt repaymets icludig the pricipal ad iterest (> ); ρ = the cost of equity; r = the riskless iterest rate (riskless debt assumed). Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki The value of the firm at ay momet of time equals the sum of the market values of equity ad debt, i.e. V j = S j + B j, j =,,. What is ot paid out as a divided or a repaymet of debt from the corporate cash flow, is ivested withi the firm. To simplify our aalysis we assume o taxes. 4 The ultimate aim of a ivestmet (κ t ) ca be take as icreasig the value of the firm, ad thus cotributig towards a higher shareholder value. Cosider ext the firm that follows a alterative ivestmet strategy of puttig the fuds κ t available for ivestmet (et of depreciatio) to marketable securities (istead of ivestig i covetioal R&D ad machiery ad equipmet), yieldig a retur equal to their cost of capital, i, at the same risk level. Thus, i stads for opportuity retur at the same risk level. Deote the value created by this alterative strategy as: (2) V T =V + δ t e ρ( t) dt + β t e r( t) dt + κ t e i( t) dt. Based o assumed developmet of the equity value of the firm uder the alterative ivestmet strategy, a measure for capital productivity ca be preseted as follows. 5 Measure J2. Uder the alterative strategy, assume that the ivestmet equal to depreciatio suffices to maitai the itermediate cash flows (C t = δ t + β t + κ t ) but at the ed of the period the firm has equity value of zero. Hece, V = + B. (3) J2 = V T V T =S κ t e i( t) dt. I above, J2 represets the differece betwee the equity value at the ed of the period (cotributed to by ivestmets made withi the firm) ad the future value of a alterative ivestmet (e.g. marketable securities). Were J2 egative, the firm s ivestmet strategy would 3 See Jese (1993), which is based o discrete time. To simplify our aalysis, we use a cotiuous time model. 4 It is well kow that with a tax regime which allows debt capital iterest to be set off agaist tax liability, a compay would be able to icrease its expected aual et after-tax cash flow by gearig up. Thus, taxatio would imply a icetive to chage the firm s capital structure towards higher debt. The aalysis to be preseted i the followig is based o compariso of two alterative ivestmet strategies i a purely descriptive cotext. The effects of taxatio would be the same for both ivestmet scearios ad thus taxatio should ot ifluece the compariso outcome. A differet but related questio ot addressed here is whether a tax regime per se chages the allocatio of ivestmet fuds. 5 Jese (1993) preseted four measures. Here oly oe measure will be reviewed but the aalysis for the remaiig measures should be straightforward followig the argumet to be preseted i the followig. 21

4 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki ot have bee optimal i terms of creatig value. A larger value could have bee created if the 3. The Artto Pohjola paradox firm had ivested i alterative ivestmet opportuities, e.g. marketable securities with the opportuity retur i (at the same risk level). 6 Whe lookig at Jese s capital productivity measures, it should be oted that the retur o corporate ivestmets is usually ot observable to the shareholders, i.e. we have the case of asymmetric iformatio. The shareholders evaluate the explicit fiacial returs, i.e. the share purchase costs vis-à-vis received divideds ad capital gais i terms of icreased share value. The total value for the shareholders measured at date ca be writte as follows 7 : (4) R = S S e ρ + δ t e ρ( t) dt, where ρ is the cost of equity. Does the shareholder s iterest automatically guaratee that the firm s ivestmet strategy yields a retur o the ivestmet equal to, or higher tha, the opportuity cost of capital? I other words, does such a mechaism exist that corporate maager s bad ivestmet decisios are ecessarily reflected i the shareholder returs, leadig possibly further to the maagemet reorgaisatio by the owers? The aswer to this questio seems to be egative. The followig discussio will elaborate that a shareholder may be perfectly satisfied eve if the corporate ivestmet strategy icurs a sigificat loss i opportuity cost terms For example, aother measure preseted i Jese (1993), closely related to J2 measure examied i the preset paper, is defied as follows: Assume the ivestmet equal to depreciatio is sufficiet to maitai the iitial equity value of the firm, but due to the allocatio of (et) ivestmet i the marketable securities the equity value does ot icrease. Hece, V =S +B. J1 = V T V T = S S κ t ei( t) dt. Thus, J1 effectively compares the chage of the equity value of the firm to the future value of ivestmet i marketable securities. 7 Artto (1997) used preset value terms; here we apply future values. If R=, the ρ is the iteral rate of retur (IRR) o share purchases. If R >, the the shareholder ejoys higher retur tha IRR, i.e. IRR> ρ (the case of positive excess returs). I the followig, we will demostrate that the shareholders may ejoy a retur equal to, or eve higher tha, their cost of equity (ρ) while they simultaeously suffer a loss i opportuity cost terms due to iefficiet ivestmets by corporate maagers. I other words, the Artto Pohjola paradox appears whe the shareholders retur could have bee icreased had the ivestmets withi a firm yielded a retur equal to the oe available i alterative ivestmet opportuities. For the sake of simplicity ad without loss of geerality i the rest of the paper, we assume costat divideds, δ t = δ ad ivestmets, κ t = κ. Cosider ow the relatioship betwee the shareholders retur as defied i equatio (4) ad Jese s capital productivity measure, equatio (3). The value of ed-period equity S is critical for the ultimate appearace of the Artto Pohjola paradox. To show this, we assume that S = S e γ, i.e. that the iitial equity value grows to the ed-period value at the costat rate γ. Recall that J2 = S e γ κe i( t) dt. After itegratig, we get J2 = S e γ (κ/i)(e i 1). The first term is the ed-period value of equity, ad the secod term is the value of ivestmets i marketable securities. The shareholder s total retur is R = S e γ S e ρ + δe ρ( t) dt, ad after itegratig we get R = S (e γ e ρ ) + (δ/ρ) (e ρ 1). The first term is the capital gai, ad the secod term is the received divideds. Now we ca fid γ J ad γ R such that J2 = ad R =, respectively. The sufficiet coditio for the Artto Pohjola paradox is R ad J2 <, which is valid whe γ R γ < γ J. After some calculatio we get γ R =l[(δ/ρs )(1 e ρ ) +e ρ ]/ ad γ J =l[(κ/is )(e i 1)]/. Now we ca write the sufficiet coditio as follows: Result 1: The Artto Pohjola paradox appears if the growth rate γ of a firm s market value of equity is such that: ρ + l[(δ δe ρ )/(ρs e ρ ) + 1]/ γ< i + l[(κe i κ)/(is e i )]/.

5 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki R R < J2 Ivestor Heave Black Hole J2 < Artto Pohjola Paradox Maagemet Disaster Figure 1: Jese s Productivity Measure ad Shareholders Retur. Thus the Artto Pohjola paradox appears whe γ is high eough to guaratee a o-egative retur (R ) for the shareholders but is too low i a sese that a larger shareholder value could have bee created had the fuds bee paid out to the shareholders ad ivested further i equivaletly risky projects. While Result 1 sets forth the geeral sufficiet coditio for the Artto Pohjola paradox, a immediate ad iterestig further questio is, i which kid of idustries the paradox is likely. Aswerig the questio properly would require empirical aalysis of characteristics of various idustries, which is beyod the scope of this paper. Some further ecoomic ituitio ca, however, be gaied simply by ispectig the coditios that must hold to satisfy Result 1. Notice that the left-had side of the iequality is smaller or equal to ρ whe the term iside square brackets is strictly greater tha zero but smaller tha oe. I order to demostrate the relatioship betwee the received divideds ad the capital gai assume for a momet that δ =. The o-egative shareholder retur is possible eve with zero divideds, but the the required γ has to be higher, i fact i this case it equals to ρ. Cosider ow the right-had side of the iequality, ad assume ρ = i. The strict iequality thus holds oly whe (κe i κ)>is e i, sice oly the i γ<i+l[(κe i κ)/(is e i )]/. Assume ow that divideds are positive, the we kow that the required capital gai ad thus γ, which guaratee the o-egative shareholder retur, decrease. Therefore, lower returs o marketable securities (i) ad/or smaller ivestmets i marketable securities (κ) are eeded for the appearace of the Artto Pohjola paradox. I geeral, we ca coclude that for a give growth rate of a firm s market value of equity the appearace of the Artto Pohjola paradox is more likely for firms, which are itesive i ivestmet compared to the divideds paid out. The relatioship betwee Jese s productivity measure ad the shareholders fiacial retur are further illustrated i Figure 1. For the sake of illustratio, suppose that the shareholder IRR ca be smaller tha the cost of equity ρ, i.e. it is possible that R < (egative excess returs). The Artto Pohjola paradox appears i the low-left quadrat of the table, where the shareholders retur R is sufficiet (positive or equal to zero) but Jese s productivity measure is egative. Were the sufficiet shareholder value R associated with high positive productivity figures, we would have the situatio of ivestor heave (the top-left quadrat). Third, low returs to the shareholder (R < ), are likely to trigger a compay restructurig ( maagemet disaster ). Fially, the same sceario is likely for the top-right quadrat ( black hole ), where the firm absorbs equity while it does ot ivest. 4. Aalysis of the ecessary coditios I order to gai further uderstadig o the ecessary coditios for the paradox, we 23

6 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki preset some additioal results. I particular, a cosideratio of the differece of R-J2 appears to geerate some useful further results from a empirical viewpoit. Notice that positive R-J2 is a ecessary but ot a sufficiet coditio for the Artto Pohjola paradox (R ): the positive differece R-J2 idicates that the fiacial retur to the shareholder exceeds the capital productivity measure (furthermore with J2 <, the shareholder suffers a opportuity loss while beig satisfied i fiacial terms). The aalysis below idicates that the sufficiet coditios for the Artto Pohjola paradox are rather geeral. Note also that the magitude of the differece R-J2 per se is ot the mai subject of our iterest but the focus is istead o the sig of the differece. So, cosider the differece: (5) R-J2 = S S e ρ + δe ρ( t) dt 24 S + κe i( t) dt = δe ρ( t) dt + κe i( t) dt S e ρ. From above we get the followig result: Result 2: (R-J2) is idepedet o the ed-period equity value S ( > ). I order to simplify our aalysis we aalyse the case where ρ = i, ad the the differece becomes: (6) R-J2 = δe i( t) dt + κe i( t) dt S e i. Two simple calculatios are eeded to derive the basic descriptio for R-J2. First, the fuctio is upward slopig with respect to time if its first derivative is positive. It appears that: Result 3: (R-J2)/ >, if δ + κ >is i.e. if the sum of aual divideds ad ivestmet is greater tha the iterest o the iitial equity value. The validity of this coditio, of course, is a empirical matter. However, it should be oted that the coditio is ormally true if the compay yields retur o equity larger tha the iterest rate. It therefore is quite likely that the shareholder s retur grows faster tha does the productivity of capital J2. Secod, we shall examie whether the differece R-J2 will become positive over time: Result 4: (R-J2) > if δ + κ >is ad is greater tha * = [l(δ + κ) l(δ + κ is )]/i. Hece, (R-J2) will become positive with a sufficietly log time frame (). With regard to the above results, Figure 2 illustrates a possible outcome for the differece R-J2. 8 I brief, the shape of the curve R-J2 is determied by the fact that while the shareholders opportuity cost icreases over time oly at the rate of cost of equity (ρ), the firm s opportuity cost of capital grows faster, at the rate of iterest (i) plus the aual capital ivestmet (κ). Sice the differece R-J2 is idepedet of the developmet of the equity value S but the actual values of R ad J2 are ot each value for S will geerate differet curves for R ad J2. Therefore, with varyig S, a ifiite umber of R ad J2 curves ca be geerated, which all, however, will deliver the oly oe ad the same differece curve R-J2. Figure 3 below depicts oe possible pair of R ad J2, which geerates the same differece curve R-J2 as i figure 2. I the figure, R is a icreasig fuctio of whereas J2 is a decreasig fuctio. The fact that R is icreasig idicates that the cost of equity (ρ) is smaller tha the shareholder IRR ideed, the IRR here is 13%. The decreasig J2 curve idicates that the growth rate of the equity value (γ) is ot large eough to compesate the opportuity cost of capital expediture, determied by the available retur whe ivestig i alterative opportuities, e.g. marketable securities (a iterest rate (i) as well as the cost of equity (ρ) of 1% was assumed i the calculatios). 8 The figures are geerated based o illustrative data assumig κ =.25 ad δ =.25 for periods 1 to be paid at the ed of each period. Equity value S of 1.4 is assumed throughout the period ad it is assumed thereafter to grow at 11% per year amoutig to 4 at the ed of the period 1. Ivestor IRR is 13% over periods 1. The cost of equity, ρ, ad a iterest rate, i, are both 1%.

7 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki Figure 2: The Differece betwee Shareholder Retur (R) ad Capital Productivity (J2). Figure 3: Shareholders Retur (R), Capital Productivity (J2) ad the Differece (R-J2). I figure 3, J2 reaches zero at the seveth period ad thereafter J2 is egative whereas R remais positive ad growig. Therefore, with greater tha seve, we have the Artto Pohjola paradox with moderate shareholder retur but too low equity value to compesate the lost retur of ivestig i alterative opportuities tha the firm itself. I geeral, a decrease i the 25

8 Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki ed period equity value S would shift R ad J2 dowwards ad a icrease i the equity value would result i a shift upwards. However, the curve R-J2 is ot affected by S, as was show above. 5. Cocludig remarks It has bee show that, uder quite geeral coditios, the shareholders fiacial retur R grows faster tha Jese s capital productivity (J2) with the legtheig of the time frame. I additio, with sufficietly log time frames R always becomes larger tha J2. The Artto Pohjola paradox appears if the firm s equity value icreases at a relatively low rate. Here the opportuity cost of accumulatig ivestmet becomes large, resultig i egative capital productivity, measured i terms of J2. Simultaeously, the shareholders fiacial retur R may still remai positive idicatig reasoably profitable ivestmet i IRR terms (or i future value terms). It is therefore possible that a firm goes o ivestig i low yieldig assets while the shareholder still ejoys moderate returs without realisig the opportuity losses arisig from the ivestmets. It ca be cocluded that the Artto Pohjola paradox is ot a real paradox. It should be emphasised that our results further uderlie Jese s origial cocer of the importace of the iteral corporate cotrol systems. As argued by Jese (1986), shareholders ad corporate maagers may have coflictig iterests. I particular, the maagers may have icetives to cause their firm to grow beyod the optimal size by realisig ivestmets, which do ot yield positive et preset values whe discouted at the relevat cost of capital. While Jese s propositio is based o the icetives of maagers, our fidigs further support this argumet. The shareholders do ot ecessarily challege the maagemet sice they are perfectly happy ad satisfied i fiacial terms. It should also be oted that Jese s productivity measures may idicate the same level of productivity for a highly ivestmet-itesive compay ejoyig rapid icrease i its market 26 value of equity ad, o the other had, for a compay with a rather stagat equity value developmet ad a low scale of ivestmet activity. Eve if the productivity measures are ot directly observable per se for a ivestor, she typically gets worried if the share price of a capital-itesive firm performs poorly to some exchage idex (i.e. average performace). I geeral there are two possible reasos for the Artto Pohjola paradox ad for a low capital productivity rak à la Jese. First, the ivestors may lack sufficiet iformatio i order to make a correct judgemet o the expected returs of a firm s ivestmet activity. I fact, compaies do ofte have a icetive to coceal iformatio o their ivestmets, especially o R&D. 9 Therefore, the solutio for improvemet is i this case trivial: iform if you ca. The secod possible explaatio for a low capital productivity is that the firm s ivestmet does ot, ideed, yield sufficiet returs (from a shareholder s viewpoit). The most straightforward solutio for improvemet is here to ivest less ad to pay more divideds. The scaleddow ivestmet activity would immediately result i icreased productivity (J2) whereas the icreased divided stream would simultaeously lift up the shareholder retur (R). Also, if the firm s commitmet to holdig back its ivestmets is cosidered covicig i the log term, the icreased divided stream may become capitalised i the market value of equity, liftig the productivity measures ad the shareholder retur further upwards. Alteratively, the firm might ivest i buyig back its ow shares, hopig that it would result i a price icrease. The icreased capital gais would further improve the productivity performace as well as the shareholder retur. Furthermore, the shareholder retur could i this case still be improved by the fact that the divided pool available per share held by the ivestors would grow. The result would thus be higher J2 ad R, ad 9 For example, Raisio Yhtymä Ltd did ot perform successfully i Pohjola s (1996) study sice the research period did ot capture the lauchig of a ew product Beecol, after which the compay s share price soared.

9 a shift towards the ivestor heave as idicated i figure 1. The mai message stemmig from our aalysis is that the simultaeous appearace of shareholders reasoable retur, o the oe had, ad low capital productivity, o the other had, is ot ecessarily paradoxical. This situatio, although apparetly cotradictory, is possible whe the growth rate of equity is withi certai boudaries. I the log ru, oly firms, which are able to ivest with returs higher tha the opportuity cost of capital, are successful. Fiish Ecoomic Papers 1/2 O. Haltia ad M. Leppämäki Refereces Artto, E.W. (1997). Performace ad Iteratioal Competitiveess of Listed Paper Idustry Groups Filad vs Swede, Caada ad USA. Helsiki School of Ecoomics ad Busiess Admiistratio, Workig Papers W-182. Helsigi Saomat (Jyrki Iivoe), (1998). Paperiyhtiöide ivestoitie tuotosta sytyi kiivas oppiriita talousprofessorie laskelmissa o kymmeie miljardie markkoje eroja. Helsigi Saomat, Jese, M.C. (1986). Agecy Costs of Free Cash Flow, Corporate Fiace, ad Takeovers. America Ecoomic Review 76, (1993). The Moder Idustrial Revolutio, Exit ad the Failure of Iteral Cotrol Systems. The Joural of Fiace XLVIII, Kaiaie, V. (2). Empire Buildig by Corporate Maagers. The Corporatio as a Savigs Istrumet. Joural of Ecoomic Dyamics ad Cotrol 24, McKisey Global Istitute (1996). Capital Productivity. Washigto D.C.: McKisey Global Istitute. Pohjola, M. (1996). Tehoto pääoma. Helsiki: WSOY. 27