Bertrand&Mullainathan.2001.Are CEOs Rewarded for Luck

7/31/2019 Bertrand&Mullainathan.2001.Are CEOs Rewarded for Luck

1/33

ARE CEOS REWARDED FOR LUCK? THE ONESWITHOUT PRINCIPALS ARE*

MARIANNE BERTRAND AND SENDHIL MULLAINATHANThe contracting view of CEO pay assumes that pay is used hy shareholders tosolve an agency problem. Simple models of the contracting view predict that payshould not be tied to luck, where luck is defined as ohservahle shocks to perfor-mance heyond the CEO's control. Using several measures of luck, we find thatCEO pay in fact responds as much to a lucky dollar as to a general dollar. Askimming model, where the CEO has captured the pay-setting process, is consis-tent with this fact. Because some complications to the contracting view could alsogenerate pay for luck, we test for skimming directly by examining the effect of

governance. Consistent with skimming, we find that better governed firms paytheir CEO less for luck.

I. INTRODUCTIONCEO pay is usually viewed through the lens of principalagent models. Under this contracting view, pay is used to reducethe moral hazard problem that arises hecause CEOs often own

very little of the firms they control. Shareholders (perhaps actingthrough the board or the compensation committee) optimallydesign the pay package in order to increase the CEO's incentive tomaximize firm value. ^ Simple models of the contracting viewgenerate one important prediction. Shareholders will not rewardCEOs for obseruable luck. By luck, we mean changes in firmperformance that are beyond the CEO's control. Tying pay to

* The results in this paper were previously circulated as part of a largerworking paper entitled "Do CEOs Set Their Own Pay? The Ones Without Princi-pals Do." We are extremely grateful to Daron Acemoglu, Rajesh Aggarwal, GeorgeBaker, Patrick Bolton, Peter Diamond, Rohert Gibbons, Denis Gromb, Brian Hall,Bengt Holmstrom, Caroline Hoxby, Glenn Hubbard, Lawrence Katz, Jom-SteffenPischke, Nancy Rose, David Scharfstein, Robert Shimer, Andrei Shleifer, RichardThaler, and seminar participants at the University of California at Berkeley,Columhia University, the University of Chicago, Harvard University, the Mas-sachusetts Institute of Technology, Princeton University, and the National Bu-reau of Economic Research Corporate Einance Summer Institute 1999 for veryhelpful comments. We thank Kenneth Ayotte and Michael Mitton for excellentresearch assistance, Michael Haid for giving us access to his data set of oilcompanies, and David Yermack for giving us access to his data on executivecompensation. Einancial support was provided hy the Russell Sage Foundation,the Princeton Industrial Relations Section, and the Princeton Center for Economic


2/33

9 0 2 QUARTERLY JOURNAL OF ECONOMICSluck, therefore, cannot provide better incentives and will onlymake the contract riskier [Holmstrom 1979].^This paper starts by examining whether or not CEOs are infact paid for luck usin g th re e m ea su re s of luck.' Fi rs t, we performa case study of the oil industry where large movements in oilprices tend to affect firm performance on a regular basis. Second,we use changes in industry-specific exchange rate for firms in thetraded goods sector. Third, we use year-to-year differences inmean industry performance to proxy for the overall economicfortun e of a sector.'* For all thr ee m ea su res , we find th a t CEO payresponds significantly to luck. In fact, we find that CEO pay is assensitive to a lucky dollar as to a general dollar. Moreover, theseresults hold as well for discretionary components of paysalaryand bonusas they do for options grants.

These results are inconsistent with a simple contractingview. M otivated by prac titioners such as C rystal [1991], we pro-pose an alternative, skimming, which can explain these results[Bertrand and Mullainathan 2000a]. The skimming view alsobegins w ith the sep aratio n of ow nership an d control, bu t it argu esthat this separation allows CEOs to gain effective control of thepay-setting process itself. Both because of entre nch m ent, such aspacking the board w ith sup po rters , and because of the complexityof the pay process, many CEOs de facto set their own pay withlittle oversight from shareholders. Their pay level then becomesconstrained by an unwillingness to draw shareholders' attention.Pay for performance arises in the skimming view because goodperformance may ease these constraints, in essence creating

2. Note our emphasis on observable luck. In any model, given the random-ness of the world, CEOs (and almost everybody else) will end up being rewardedfor unohservable luck. Note also our emphasis on the fact that this predictionholds in simple agency models. As we will discuss shortly, complications to theagency model can in principle alter this result.3. Blanchard, Lopez-de-Silanes, and Shleifer [1994] present suggestive evi-dence on pay for luck hy showing tha t windfall gains from court rulings r aise thepay of CEOs. It is only suggestive since court rulings m ay not he luck hut ra th era resu lt of tbe CEO 's work. In other dom ains. Shea [1999] independ ently performsan exercise similar to ours for baseball players.4. This last test very much resembles the approach followed in the relativeperformance evaluation (RPE) literature [Gihhons and Murphy 1990; Janakira-


3/33

ARE CEOS REWARDED FOR LUCK? 9 0 3slack for the CEO. In other words, when the firm is doing well,shareholders are less Hkely to notice a large pay package. To theextent that lucky dollars create slack as readily as general dol-lars, pay for luck arises.

Finding pay for luck, however, does not necessarily single outthe skimming model. Complications to the agency model canm ake it such t h a t pay ing for luck is in fact optim al. For example,suppose that the value of a CEO's human capital rises and fallswith industry fortunes. One would then find that pay correlateswith luck because the CEO's outside wage moves with luck.Another possibility is that boards may tie pay to luck in order tomotivate CEOs to forecast or respond to luck shocks. SubsectionII.D discusses whether arguments such as these can really ex-plain the pay for luck relationship.To further differentiate skimming from these explanations,we empirically examine a direct implication of the skimmingmodel. Skimming should be less prevalent in better governedfirms. Well-governed firms, such as those with a large share-

holder pres ent on the board, limit the CEO 's ability to cap ture thepay process. We test this hypothesis using several measures ofgovernance: presence of large shareholders (on the board andoverall), CEO tenure (interacted with the presence of large share-holders to better proxy for entrenchment), board size, and frac-tion of directors th at are insid ers. Consistent w ith skimm ing, wegenerally find that the better governed firms pay less for luck.^These effects are strongest for the presence of large shareholderson the board. An additional large shareholder on the board re-duces pay for luck by between 23 and 33 percent. Large share-holders are especially important as CEO tenure increases, con-sistent with the idea that unchecked CEOs can entrenchthem selv es over time . If pay for luck we re optim al, we would ha veexpected well-governed firms to pay for luck as much as (if notmore than) poorly governed firms. For example, whether or not alarge shareholder is present, the CEO would have to be rewardedfor a rise in the value of his human capital. These findingssuggest that at least some of the pay for luck in poorly governedfirms is due to skimming by CEOs.


4/33

904 QUARTERLY JOURNAL OF ECONOM ICSI I . PAY FOR LUCK TE ST

II.A. Theoretical BackgroundA simple theoretical model will make more precise whatagency theory says about the reward for observable luck. Con-sider a standard agency setup where risk-neutral shareholderstry to induce a risk-averse top manager to maximize firm perfor-mance. Since the actions of the CEO can be hard to observe,shareholders will he unahle to sign a contract that specifies these

actions. Instead, shareholders will offer a contract to the CEOwhere her compensation level is made to depend on the firm'sperformance, hetp represent firm performance and a the CEO'sactions, which by assum ption are unobservahle to the sharehold-ers. Firm performance depends on the actions of the CEO and onrand om factors. We split th e rando m factors into two com ponents:those th at can be observed by shareh olders a nd those th at cannot.For an oil firm, the price of crude oil would be an observablerandom factor. Letting o be the observable factor and u be theunobservable noise term, we assume that performance can bewrit ten as p = a + ho + u.

Under some technical conditions (CARA utility and Brown-ian motion for the performance process), Ho lmstrom and Milgrom[1987] calculate the optimal incentive scheme for this model. Lets denote this incentive scheme. Since shareholders can only ob-serve two variables, p and o, the incentive scheme could at mostdepend on these two variables. In fact, shareholders will onlyreward CEOs for performance net of the observable factor:(1) s = a + p{p - So) - a + p(a + u).

In other words, the optimal incentive scheme filters the ob-servable luck from performance. This is because leaving o in theincentive scheme provides no added benefit to the principa l as , hydefinition, the agent has no control over o. Motivating her on oha s no incentive effects. Beyond providing no benefit, tyin g pay toluck actually costs the principal because the variance of theincentive scheme is higher, and th e principal m ust increase m ean


5/33

ARE CEOS REWARDED FOR LUCK? 9 0 5filter. In theory, these other components could adjust enough toundo the effect of the options value fiuctuating with luck. Suchadjustment may happen if a board were to monitor luck and altereach year's salary, bonus, and number of new options granted sothat the CEO's overall pay package remains free of luck.II.B. Empirical Methodology

W ithin the agency framew ork, most of th e empirical litera-ture on CEO pay estimates an equation of the form,y, + x, + ^x * X^ + e,,,w here J,, is total C EO com pensation in firm i at time t, perf^ isa perfo rm ance m ea su re, -y, ar e firm fixed effects, x, are tim e fixedeffects, and X,, are firm- and CEO-specific variables such as firmsize and ten ur e. The coefficient p cap ture s th e stre ng th of the payfor performance relationship.Performance is typically measured either as changes in ac-counting returns or stock market returns, and we will use both

m easures.^ In m easu ring compensation 3^,,, much of the litera turefocuses on the fiow of new compensation. Ideally, the compensa-tion in a given year would also include changes in the value ofunexercised options granted in previous years [Hall and Liebman1998]. Such a calculation requires data on the accumulated stockof options held by the CEO each year, wh ereas existing dat a s ets,including ours, contain only information on new options grantedeach year. Consequently, our compensation measure excludesthis component of the change in wealth. For our purposes, how-ever, this exclusion does not pose much of a problem. The changein wealth due to changing option values is mechanically tied toluck since options are not indexed. Thus, even if these data wereavailable, focusing on the subjective components of pay wouldstill be a natu ra l strategy. We discuss this issue at grea ter lengthin subsection II.D.

To estimate the general sensitivity of pay to performance, wewill follow the literature and estimate equation (2) using a stan-dard Ordinary Least Squares (OLS) model. To estimate the sen-sitivity of pay to luck, we need to use a two-stage procedure. In


6/33

9 0 6 QUARTERLY JOURNAL OF ECONOMICSsecond stage, we will see how sensitive pay is to these predictablechanges in performance. This two-stage procedure is essentiallyan Instrumental Variahles (IV) estimation where the luck vari-able is the instrument for performance.^Letting o be luck, the first equation we estimate is

6 * 0 , , + gi + c, + ax* X,, + e,,,where o^ represents the luck measure (oil price, for example).From this equation we predict a firm's performance using onlyinformation about luck. Call this predicted value per/;,. We thenask how pay responds to these predictable changes in perfor-mance due to luck:

The estimated coefficient Pic* indicates how sensitive pay is tochanges in performance that come from luck. Since such changesshould be filtered, basic agency theory predicts that p^^^.^ shouldequal 0.U.C. Oil Industry Study

We now turn to the oil industry as a case study of pay forluck. As Figures I and II show, the price of crude oil has fluctu-ated dramatically over the last 25 years. These large fluctuationshave caused large movements in industry profits. Moreover, theselarge fiuctuations in crude oil prices are likely to have beenbeyond the control of a single American CEO. For example, thesharp decline in cinide oil price at the end of 1985 was caused bySaudi Arabia's decision to reform its petroleum policy and toincrease production, an action hardly attributable (and neverattr ibu ted ) to the CEO s of Am erican oil firms. S imilarly, the largeoil price increase between 1979 and 1981 is usually attributed toan internal policy change by OPEC. Oil price movements there-fore provide an ideal place to test for pay for luck: they affect

7. One might wonder why we should use this procedure rather than simplyinclude o directly into the pay for performance equation (2) and run OLS to


7/33

ARE CEOS REWARDED FOR LUCK?

1977 Dollars per Barrel

9 0 7

21.0349

77 78 79 80 81 82 83 84 85 86 87 88 89 9O 91 92 93 94

FIGURE IReal Price of a Barrel of Crude Oil

performance, are measurable, and are plausibly beyond the con-trol of the CEOs.We use adata set on the pay and performance for the 51larg est Am erican oil comp anies between 1977 and 1994 to imple-

51 Largest OilCompaniesI I I I I I


8/33

9 0 8 QUARTERLY JOURNAL OF ECONOMICSI Change in Total Comp. Change in Oil Price

1977 1979 1981 1983 1985 1987 1989 1991 19931978 198O 1982 1984 1966 1988 1990 1992 1994

FIGURE IIIOil Industry CEO Pay and Crude Oil Price

ment the methodology of the previous section.^ Before moving toregression analysis, it is useful to look directly at how pay fluc-tuates compared with the movements inFigures I and II. InFig ure III we hav e graphe d chan ges in oil prices for each ye ar andchanges inmean log pay in the industry. Two striking factsemerge. First, pay changes and oil price changes correlate quitewell. In twelve of the seventeen years they are of the same sign:both are up, or both are down. This is suggestive of pay for luck.Second, the remaining five years where pay and oil prices move inopposite directions a re all yea rs in which th e oil price drops. T hishints at an asymmetry: while CEOs are always rewarded for goodluck, they may not always be punished for had luck.

This graphical analysis does not quantify pay for luck. Onecannot compare it with the size of the overall pay for perfor-mance. It also does not control for other firm-specific variablesthat might be changing over time. Table I follows the empirical


9/33

ARE CEOS REWARDED FOR LUCK? 909TABLE I

PAY FOR LUCK FOR OIL C E O S ( L U C K MEASURE IS LOG PRICE OF CRUDE OIL)DEPENDENT VARIABLE: In (TOTAL COMPENSATION)

Specification:

Accounting rateof returnLog (shareholderwealth)

A geAge^ * 100TenureTenure^ * 100Firm fixed effectsYear quadraticSample sizeAdjusted R'^

Greneral(1 ).8 2(.16).05

(.02)- . 0 4(.02).01(.01)- . 0 3(.02)

YesYes827

.7 0

Luck(2)

2.15(1.04).0 7

(.03)- . 0 5(.02).0 1(.01)- . 0 3(.01)

Y esYes8 2 7

General(3 ).38(.03).0 5

(.02)- . 0 4(.02).0 1(.01)- . 0 3(.02)YesYes827.7 5

Luck(4 ).35(.17),0 5

(.02)- . 0 4(.02).0 1(.01)- . 0 3(.02)YeeYes827

a Dependent variable is the logarithm of Wtal compensation. Performance meaaare is accounting rate ofreturn in columns (1) and (2) and the logarithm of shareholder wealth in columns (3) and (4). All nominalvariabips are expre.saed in 1977 do llars,b. Summary s tatistics for the sample of oil firms are available in Appendix 1.c. The luck regression (columns I2l and (4|l instrument for performance with the logarithni of the priceof a b arrel of trude oil in that year, expressed in 1977 dollars .d. Eaoh reg ression includes firm fixedfifFpctaand a quadratic inyear.e. Standard errors are in parentheses.

methodology presented in subsection II.B, which permits a moresystematic analysis. All regressions use log (total compensation)as dependent variable and include firm fixed effects, age andtenure quadratics , and a performance measure as dependentvariables.^ We also include a year quadratic to allow for the factthat CEO pay has been trending up during this period. Column(1) esti m ates the sensitivity of pay to a general change in account-ing performance. The coefficient of .82 suggests th at if an oil firmincreases its accounting return by one percentage point, totalcompensation rises by .82 * .01 = .0082 log point. Roughly,a one percentage point increase in accounting returns leads to a


10/33

9 1 0 QUARTERLY JOURNAL OF ECONOM ICScreases with age and to a lesser extent w ith tenu re. Both the ageand tenure profile are concave (the negative coefficient on thequadratic term).Column (2) estimates the sensitivity of pay to luck. As de-scribed above, we instrument for performance with the log of oilprice.^^ The coefficient in column (2) now rises to 2.15. Thissuggests that a one percentage point rise in accounting returnsdue to luck raises pay by 2.15 percent. Given the large standarderrors, one cannot reject that the pay for luck coefficient and payfor general performance coefficient are the same. One can, how-ever, strictly reject the hypothesis of complete filtering; oil CEOsare paid for luck that comes from oil price movements.

Columns (3) and (4) perform the same exercise for a marketm eas ure of performance, share hold er w ealth. The coefficient of.38 on column (3) sugge sts t ha t a 1 perce nt increase in s ha re-holder w ealth leads to roughly a .38 perce nt incre ase in CEO pay.In column (4) we find th at a 1 percen t increa se in sha reho lderwealth due to luck leads to .35 percent increase in CEO pay.Again, pay for luck matches pay for general performance.II.D. More General Tests

The oil industry case study, while instructive, raises thequestion of how genera lizable thes e resu lts a re. In this subsectionwe will exam ine luck shocks th a t affect a broad er s et of firms. Wefocus on two measures of luck: movements in exchange rates andmean industry performance. By affecting the extent of importpenetration and hence foreign competition, exchange rate move-ments can strongly affect a firm's profitability.^^ Movements inmean industry performance also proxy for luck to the extent thata CEO does not influence how the rest of her industry performs.As we mentioned before, this last instrument is more question-able. In practice, however, we find that mean industry move-ments operate exactly like exchange rate or oil price movements.

To implement these tests, we use compensation data on 792large corporations over the 1984-1991 period. The data set wasgraciously made available to us by David Yermack and Andrei


11/33

ARE CEOS REWARDED FOR LUCK? 9 1 1sation data were collected from the corporations' SEC Proxy,10-K, and 8-K fillings. Other data were transcribed from theForbes magazine annual survey of CEO compensation as well asfrom SEC Registration statements, firms' Annual Reports, directcorrespondence with firms, press rep orts of CEO hires and depa r-tu res , and stock prices published by Sta nd ard & Poor's. Firm swere selected into the sample on the basis of their Forbes rank-ings. Forbes magazine publishes annual rankings of the top 500firms on four dimensions: sales, profits, assets and market value.To qualify for the sample, a corporation must appear in one ofthese Forbes 500 rankings at least four times between 1984 and1991. In addition, the corporation m ust hav e been publicly t rad edfor four consecutive years between 1984 and 1991.

Yermack's data are attractive in that they provide both gov-ernance variables and information on options granted, not justinformation on options exercised. But the y do not include ch angesin the value of options held, which we must therefore excludefrom our compensation measure. If anything, this biases us to-ward understanding the amount of pay for luck. Since options arenot indexed, changes in the value of options held will covaryperfectly with luck. Including these changes in the compensationmeasure would only increase the measured pay for luck. Thisdata limitation, therefore, is less of a concern for our purposes.

Table II presents summary statistics for the main variablesof interest in the full Yermack data.^^ All nominal variables areexpressed in 1991 dollars. The average CEO earns $900,000 insalary and bonus. His total compensation is nearly twice thatamount at $1,600,000. The difference indicates the large fractionof a CEO's pay that is due to options grants. The average CEO isroughly 57 yea rs old and h as been CEO of the firm for nine y ear s.As far as governance goes, the average firm in our sample has1.12 large sh areh olde rs, of which less tha n a fourth are s itting onthe board. There are on average thirteen directors on a board.Forty-two percent of them are insiders.^^

12. In practice, depending on the required regressors, the various tests in the


12/33

912 QUARTERLY JOURNAL OF ECONOMICSTABLE II

SUMMARY STATISTICS: FUL L YERMACK CEO SAMPLE

Age of CEOTenure of CEOSalary and bonusIn (Salary and bonus)Total compensationIn (Total com pensation)Number of large shareholders (All)Number of large shareholders on boardBoard sizeFraction of insiders on board

Mean57.429.10901.696.621595.856.981.12

.2 413.45.42

S. D.6.848.08795.15.6 03488.32.8 11.42.744.54.1 9

a. Sample period is 1984-1991.b. All nominal variables are expressed in tho usand s of 1991 dollars.c. Column I ia thp mean for each variable, while column 2 is the stand ard deviation.d. A large shareh older is defined as someone who owns more than 5 percent of the common share s inthe company, pxclviding the CEO. Insiders here denote directors who are current or former officers of thecompany, relatives of corporate officers, or anyone who has a substantial business relationship withthe company. Relationships arising in the normal course of business would not be called an insider. Ourinsider definition corresponds to insider + gray in the original Yermack data.e. Total compensation eijuals salary plus bonus plus total v alue of options gran ts plus othercompensation.

O ur first gen eral m eas ure of luck focuses on exchange rat emovements. We exploit the fact that exchange rates between theU. S. dollar and other country currencies fluctuate greatly overtime. We also exploit the fact that different industries are affectedby different cou ntries' exchange ra te s. For exam ple, since the toyindustry may he more affected by Japanese imports while thelumber industry may be more affected by Bolivia, these twoind ustr ies m ay experience very different shocks in the s am e yea r.This allows us to construct industry-specific exchange rate move-ments which are arguably beyond a specific CEO's control sincethey are primarily determined by macroeconomic variables. Theexchange rate shock measure is based on the weighted average ofthe log real exchange rates for importing countries by industry.The weigh ts are th e sh ar e of each foreign country's im port in totalindu stry im ports in a base year (1981-1982). Real exchange ra tes


13/33

ARE CEOS REWARDED FOR LUCK? 913TABLE III

PAY FOR LUCK

!n (total In (totalDep. van: Cash comp In (cash) comp) In (cash) comp)

Specification: General Luck General Luck General Luck General Luck General LuckPanel A: Luck Measure is Exchange Rate Shock

Income .17 .35 1.021 (.16)

Im-ome I assets 2.13 2.94 2.36 4.39 _ _ _ _(.16) (1.28) (.28) (2.17)

In (shareholder .22 .32 .31 .57weal th) 1.021 (.13) (.03) (.23)

Sample size 1737 1737 1729 1729 1722 1722 1713 1713 1706 1706Adjusted ft^ .75 .75 .58 .75 .59

Panel B: Luck Measure is Mean Industry PerformanceIncome .21 .34 _ _ _

(.02) (.10)Income I assets 2.18 4.02 2.07 4.00 _ _ _

(.12) (.53) (.21) (.86)In (sh areholder .20 .22 .25 .29

weal th) (.01) (.12) 1.02) (.19)Sample size 4684 4684 4648 4648 4624 4624 4608 4608 4584 4584Adjusted R^ .77 .81 .70 ,82 .71

a. Dependent variable is the level of salary and bonus in columns 1 and 2. the logarithm of salary andbonus in columns 3. 4. 1. and 8 and the logarithm of total compensation in columns 5, 6, 9. and 10.Performiince measure is operating income before extraordinary items in columns i and 2 (in millions),operating income to total assets in columns 3 to 6 and the logarithm of shareholder wealth in columns 7 to10. All nominal variahles are expressed in real dollars.

b. In the luck regre.ssions in Panel A, the performance measure is inBtrumented with current and laggedappreciation and depreciation dummies and current and lagged exchange rate index growth. First-stageregressions are presented in Appendix 2.c In the luck regressions in Panel B, the performance measure is instrumented with the total assets-weighted average performance measure in the firm's two-digit industry (the firm itself is excluded from themean calculation).d. Each regression includes firm fixed effects, year fixed effect and demographic controls ^quadratics inage and tenure).e. Standard errors are Jn parentheses.

Panel A of Table III examines this luck measure. Note thatsince the exchange rate measure can only he constructed forindustries where we have imports data, the sample size is muchsmaller here than for our full sample. All regressions control for


14/33

9 1 4 QUARTERLY JOURNAL OF ECONOMICScompensation. Thus, relative to our standard specification, we donot ru n this reg ression in logs and do not include value of optionsgra nte d. Since profits are repo rted in millions and pay is reportedin tho us an ds , the coefficient of .17 in column 1 sug ges ts th a t$1000 increase in profits leads to a 17 cent increase in perfor-mance. Column 2 performs the same exercise for pay for luck: wein str um en t for performance usin g the exchang e ra te shocks.^' Asin th e oil case , we find a pay for luck coefficient th a t is of th e sa m eorde r of m ag nitud e as th e pay for ge neral perform ance coefficient.

Columns 3 through 6 run the more standard regressionw here we use the logarithm of pay and an accounting me asure ofperformance (operating income divided hy total assets). In col-umns 3 and 4 we use only cash compensation, while in columns 5and 6 we use total compensation. In hoth cases, we find thesensitivity of pay to luck to he abou t th e sam e as t he sensitivity ofpay to general performance. When accounting performance riseshy one percentage point, compensation (either total or cash) riseshy ahout 2 percent, whe ther t ha t rise was due to luck exchangerate movementsor not.Columns 7 through 9 replicate these four columns for marketmeasures of performance. Again, we find pay for luck thatmatches the pay sensitivity to a general shock. A 1 percentincrease in shareholder wealth raises pay (again either total orcash) of ahout .3 percent, irrespective of whether this rise wascaused hy luck or not.Two im porta nt points should he tak en away from th is pan el.First, the average firm rewards its CEO as much for luck as itdoes for a general movement in performance. There seems to bevery little if any filtering at all. Since we use a totally differentshock, these findings address theoretical concerns ahout the useof mean industry shocks (such as those raised in Gihbons andMurphy [1990] and Aggarwal and Samwick [1999h]) and showthat the lack of filtering ohserved in RPE findings generalizes toother sources of luck.Second, there is as much pay for luck on discretionary com-pon ents of pay (salary and honus) as the re is on othe r com ponents


15/33

ARE CEOS REWARDED FOR LUCK? 9 1 5such as options granted. This rules out the notion that pay forluck mechanically arises hecause firms commit (implicitly or ex-plicitly) to multiyear stock option plans where the number ofoptions grants is fixed ahead of time. Under such plans, as firmvalue rises, so does the value of precommitted options grants[Hall 1999]. Because salary and honus are the most subjectivecomponents of pay, finding pay for luck on these variahles is verysuggestive. Boards are rewarding CEOs for luck even when theycould filter it.

In P anel B of Tahle III, we replicate P anel A except th at ourmeasure of luck hecomes mean performance of the industry,which is meant to capture external shocks that are experiencedby all the firms in the industry. More specifically, as an instru-ment for firm-level rate of accounting return in a given year, weuse the weighted average ra te of accounting retu rn in th at ye ar inthe two-digit industry that firm belongs to, excluding the firmitself from th e calculation.^^ The w eight of a given firm in a givenyear is the share of its total assets in the aggregate "total assets"of the two-digit industry the firm helongs to. Similarly, as aninstrument for firm-level logarithm of shareholder wealth in agiven year, we use the weighted average of the log values ofshareholder wealth in the two-digit industry in that year, againexcluding the firm itself from the calculation and using totalassets to weight each individual firm.^^

As in Panel A, all regressions include firm fixed effects andyear fixed effects. We also control for a quadratic in CEO age anda quadratic in CEO tenure. The regressions include more thantwice the da ta points of Pa nel A hecause we can now use all firms,not only those in the trad ed goods sector. Pa nel B shows a p at te rnquite similar to Panel A. The pay for luck relationship in allspecifications again roughly matches the pay for general perfor-mance. Besides reinforcing the findings of Panel A, these latestfindings suggest that previous RPE results arose probably notbecause of mismeasurement of the reference industry or of theindustry shock but because of true pay for luck.


16/33

916 QUARTERLY JOURNAL OF ECONOMICSIII. WHY IS THERE PAY FOR LUCK?

The results so far clearly establish pay for luck. There areseveral reactions possihle to this evidence. First, one could take itas evidence of skimming. To understand how the skimming modelpredicts pay for luck, consider a CEO who has captured the payprocess. His primary worry in setting pay will be that outrageousskimming may cause otherwise passive investors to stand up andnotice. Good performance, however, provides the CEO with extraslack. For example, shareholders may scrutinize a firm moreclosely during bad times. This allows higher pay when perfor-mance is good and produces a positive link between pay andperformance, but for different reasons than in the contractingview. If good performance creates slack irrespective of whether itwas lucky, pay for luck will result.^'^

Altematively, one could argue that pay for luck is in factoptimal and that the evidence so far is consistent with the con-tracting view. One reason why pay for luck might be optimal isthat the CEO's outside option may in fact depend on luck. Whenthe oil industry enjoys a good fortune, the human capital of oilCEOs may simply become more valuable. Firms then pay their oilCEOs more simply to match their increased outside options.Thus , pay for luck is optimal here not as an incentive device, butmerely because the optimal level of pay increases with luck.^^Ohjections can be raised against this view. First, our sugges-tive evidence of asymmetry in pay for luck may be hard to recon-cile with this view. Average CEO compensation in the oil industry

always goes up when the price of crude oil goes up but does notalways go down when the price of crude oil goes down. In our fullsample, we performed a similar test using the industry luckshock. For accounting measures, we again found that pay re-sponds more to positive industry shocks than to negative ones(with no asymmetry on general pay for performance). For market

18 , The scrutiny of otherwise passive investors may be triggered by absoluteperformance for several reasons. The very nature of deciding where to pay atten-tion requires focusing on variables immediately at hand. Passive investors may


17/33

ARE CEOS REWARDED FOR LUCK? 9 1 7measures, we could not reject symmetry. Second, it is unclearwhy a CEO's human capital should become more valuable asindustry fortunes rise. In fact, it may be exactly in bad times thathaving the right CEO is most valuable. A priori, either relation-ship seems plausible. To test this assum ption, we examined turn -over in the CEO market. We found no statistically significantrelationship between a CEO's turnover and industry returns(after controlling for the firm's returns) and a point estimate thatwas negative. This suggests that, if anything, turnover is coun-tercyclical. Th ird, we teste d t he effect of th e ind ustr y's avera geCEO turn ov er rat e on pay for luck. If pay for luck were cau sed bymarket competition for CEOs, then industries with higher turn-over should exhibit the greatest pay for luck. For accountingmeasures, we found that industries with the highest turnover infact showed the least pay for luck. For market measures of per-formance, we found no relationship between industry turnoverand pay for luck. Of course, for the last two findings, one couldalways argue that competitive pressures operate through thethre at of turnover rath er th an through actual turnover. As awhole, thou gh, we have been u nab le to find positive evidence th a toutside bidding up of CEO wages could explain our results.

Another reason why pay for luck may be optimal is that onem ay w an t to provide incentives to th e CE O to forecast or respondto luck.^ Th is kind of arg um en t can be most readily e valua ted inour oil indu stry application. Suppose tha t a particularly talente dCEO in the oil industry understood the political subtleties of theA rab cou ntries and forecast the com ing of th e positive oil shock a tthe beginn ing of th e 1980s. By incre asing ou tpu t from existing oilwells, incre asing inv ento ries, or intensifjdng search for new wells,he could have increased his firm's profits when the shock didcome. Shouldn't shareholders reward this farsighted CEO? Theimportant point here is that those CEOs who were exceptional inhaving forecast should indeed be rewarded. But this is not whatwe te st for. We use none of the b etween firm v ariation in respon seto the oil shock. We merely test whether the average firm expe-riences a rise (or fall, for the negative shocks) in pay. Pu t an oth er


18/33

9 1 8 QUARTERLY JOURNAL OF ECONOMICSway, our results suggest that a CEO who responds to the shockexactly the same way as every other oil CEO is rewarded. Thiscannot be a reward for having forecast well. Again, one maywant to reward CEOs for exceptional responsiveness to shocks,but there is little reason to reward them for just averageresponsiveness.

A final set of responses to pay for luck would be to abandonthe literal contracting view and argue that filtering out luck issimply impossible. This might be because of cognitive complexityin understanding what is luck and what is not luck. Part of thiscognitive complexity m ay be a pu re information issue if th er e arenot enough data available to figure out the appropriate effect ofluck. For example, estimating the coefficient 6 in equation (1) m aysimply not be possible. Part of the cognitive complexity may bepsychological as in th e evidence on the fun dam ental attr ibu tionerror [Durell 1999].^^ None of the evidence so far directly refutesthis argument.III.A. The Effect of GovernanceW hile we hav e argued aga inst some of the vario us extensionsof the simple agency model, in the end we still believe that theymerit serious consideration. They suggest to us that the pay forluck finding does not per se rule out agency models. The resultsare also consistent w ith the idea t h a t filtering out luck is ju st notfeasible. Therefore, we now turn to testing a specific prediction ofthe skimming view rather than arguing against the other views.Since the skimming view emphasizes the CEO's ability to gaincontrol of the pay process, corporate governance should play animportant role in skimming. It is exactly in the poorly governedfirms w here we expect CEO s to most easily gain control of th e payprocess. This suggests that we should expect more pay for luck inthe poorly governed firms.^^

21. Another possibility is that luck is not contractible. In practice, we do notbelieve this is importan t for most of our findin gs. Firs t, it is difficult to believe t hatnoncontracting issues can explain our results in the oil industry case study: theprice of crude oil can easily be measu red and written into a contract. Second, even


19/33

ARE CEOS REWARDED FOR LUCK? 9 1 9To examine how pay for luck differs between well-governed

and poorly governed firms, we estimate two equations. First, inorder to provide a baseline, we ask how pay for general perfor-mance (not luck) differs between well-governed and poorly gov-erned firms. We estimate an OLS equation similar to equation (2)except th a t w e allow th e pay for performance coefficient to depen don governance:(3) y,, - (3 * perf, + e * (Gov,, * perf^,)

+ 7, + X + "A- * X,i + ac * Gov., + ,where Gov^f is a measure of governance. To understand thisequation, differentiate both sides with respect to performance toget dyi,/dperf,f = p + 0 + Gov^i. In words, this specificationallows the pay for performance sensitivity to be a function of thegovernance variable. A positive value for 6 would imply thatbetter governed firms show greater pay for performance.Equation (3) of course tells us nothing about pay for luck,merely about pay for performance. To get at pay for luck, wereestimate this equation using our two-stage instrumental vari-ables procedure.^^ We then compute an estimate of the effect ofgovernance on pay for general performance, 9 and an estimate ofthe effect of governance on pay for luck, 6^,,^^.

O ur tes t th en con sists in com paring B an d 6/_^^^. We willspeak of more pay for luck in poorly governed firms when poorlygoverned firms display more pay for luck relative to pay forgeneral performance. If poorly governed firms simply gave morepay for performance and pay for luck rose as a consequence, wewould not refer to this as more pay for luck. In practice, we willsee th at it is pay for luck th a t chang es with go vernance, while payfor performance hardly changes. We will also verify that theseperformance does not systematically correlate with governance. Second, for the-ories that rely on changes in the value of the CEO's human capital, it is unclearwhy these changes would happen more in the poorly governed firms. Finally, thepresumption that filtering is somehow cognitively impossible would clearly berefuted if some firms could filte r.23 . An extremely important caveat here: our approach allows for the possi-bility th at better governed fimis may have a different responsiveness of perfor-


20/33

92 0 QUARTERLY JOURNAL OF ECONOMICSTABLE IV

LARGE SHAREHOLDERS AND PAY FOR LUCK (LUCK MEASURE IS MEAN INDUSTRYPERFORMANCE) DEPENDENT VARIABLE: ln (Total Compensation)

Gouemance measure:Specification:

Income AssetsGovernance*Income/assetsln (shareholderwealth)Governance*Log (shareholder

wealth)GovernanceSample sizeAdjusted R'^

Large shareholdersGeneral

(1 )2,18(.238)

-.094(.094)

-.009(.01114610.695

Luck(2 )

4.59(.912)

-416(.204)

-018(.018)4610

General(3 )

.249(.018)

.001(.007)

-.017(.049)4570.706

Luck(4 )

.383(.219)-.066(.036).411

(,240)4570

Large shareholders on bGeneral

(5 )2,14(,217)

-,181(,176)

-,006(.021)4621,694

Luck(6 )

4.49(.882)

-1.48(.396)

,084(,0331

4621

Genera!(7 )__

,258(,017)-019(,016),100

1,108)4581,706

oardLuck

(8 )

,318(.199)-,076(,053),480

1,356)4581

a. Dependent variable is the logarithm of total conipenaation, performance measure isoperating incometo total as sets. All nominal variablps arc expressed inreal dollars,b. In all the luck regreasions, hoth theperformance measure and theinteraction of the performancemeasure with the governance measure are instrumpntpd. The instrum ents are theasset-weighted averageperformance in the two-digit industry and the interactions of the industry performance with t hat governancemeasure,c. "Large shareholders" indicates the number of blocks of at least 5 percent of the firm's common sha res,wh ether the block holder is or is not a director. "Large sha reholders on board" indicates the n umb er of blocksof at least 5 pprcent of the firm's common shares th;it are held by directors of the board,d. Each regression mcludes firm fixed eflects, ye ar fixed effects, a quadratic inage, and a quadratic intenure.e. Standard errors arein parentheses.

results are robust toallowing pay for luck tovary by firm size.Otherwise, one might simply worry that governance is a proxy forsize.^'*In Table IV, we implement this framework for the case oflarge shareholders. We ask whether thepresence of large share-holders affects pay for luck. Shleifer and Vishny [1986], amongothers, argue that large shareholders improve governance in afirm. A single inv estor who holds a large block of sh are s ina firm24. One must also be careful in interpreting the results from this exercise.


21/33

ARE CEOS REWARDED FOR LUCK? 9 2 1will have greater incentives to watch over the firm than a dis-persed group of small shareholders.^'' In our context, the idea oflarge shareholders fits most naturally as this matches the intui-tion of "having a principal around." Yermack data contain avariahle that counts the numher of individuals who own blocks ofat least 5 percent of the firm's common shares. When the CEOhappens to own such a block, we exclude this block from thecount. We further know w heth er these large shareho lders are onthe hoard or not. A priori, one might expect that large sharehold-ers on the board have the greatest impact. They can exert theircontrol not ju st thro ugh implicit pres sur e or voting, hu t also w itha direct voice on the board. Since the information is available, wewill consider the effect both of all large shareholders and of onlythose on the hoard.

The first four columns of Table IV use all large shareholdersas our measure of governance. All regressions include the usualcontrols. Column (1) estimates how the sensitivity of pay to per-formance depends on governance for accounting measures of per-formance. The first row tells us that a firm with no large share-holders shows a sensitivity of log compensation to accountingre t um of 2.18. An increase in accounting re tu m of one percentagepoint leads to an increase in pay of about 2 percent. The secondrow tells us that adding a large shareholder only weakly de-creases the sensitivity of pay to general performance, and thiseffect is not statistically significant. For example, a one percent-age point increase in accounting return now leads to a 2.09percent increase in pay when the firm has one large shareholder(compared with 2.18 in the absence of any large shareholder).Column (2) estimates how large shareholders affect pay forluck. ^ As before, th e first row tell s us th a t t he re is significant payfor luck. The second row here, however, tells us that this pay forluck diminishes significantly in the presence of a large share-holder. A one percentage point increase in accounting retu rn s dueto luck leads to roughly a 4.6 percent increase in pay when thereis no large shareholder but only a 4,2 percent increase in pay


22/33

9 2 2 QUARTERLY JOURNAL OF ECONOMICSwhen there is one more large shareholder. Each additional largeshareholder decreases this effect by .4 percent. This is a 10percent drop in the pay for luck coefficient for each additionallarge shareholder.^^Columns (3) and (4) estimate the same regressions usingm ark et m easu res of performance. In this case, the pay for generalperformance does not depend at all on the existence of a largeshareholder (a coefficient of .001 with a standard error of .007).We again find, however, that pay for luck diminishes with thepresence of a large shareholder. While the result is only signifi-cant at the 10 percent level, the economic magnitude is larger.The pay for luck coefficient now drops .066/.383 = 17 percent foreach large shareholder.

In columns (5) through (8) we repeat the above exercise butalter the governance measure. We now focus only on large share-holders on the board. Comparing columns (6) and (2), we see thatthe governance effect strengthens significantly with respect tothe filtering of accounting performance. We see that the pay forluck drops by 33 percent for each additional large shareholder.The results are very statistically significant. On market perfor-mance measures, we find the effect also rises but less dramati-cally. In column (8) the pay for luck drops 23 percent with eachlarge shareholder on the board. Moreover, this last result isinsignificant. In sum m ary , our findings in Tahle IV high light howlarge shareholders (especially those on the board) affect the ex-tent of pay for luck. Firms with more large shareholders show forless pay for luck.

The results in Table IV simply compare firms with largeshareholders with firms without. This ignores the effects of CEOten ure , ano ther im porta nt d eter m ina nt of governance. A commonbelief is that CEOs who have heen with the firm longer have hada chance to become entrenched, perhaps hy appointing friends onthe board. In this case, we would expect high tenure CEOs to


23/33

ARE CEOS REWARDED FOR LUCK? 923TABLE V

TENURE, LARGE SHAREHOLDERS, AND PAY FORLUCK (LUCK MEASURE IS MEANINDUSTRY PERFORMANCE) DEPENDENT VARIABLE: In (TOTAL COMPENSATION)

Any large shareholder on the board?No Yes No Yes

.J. . General Luck General Luck General Luck. General LuckSpecification: ^^^ ^ j (3) (4) (5) (6) (7) (8)

Income Assets

CEO tenure *Income/assetsLog (shareholder

wealth)CEO tenure 4%(Current)Appr. > 4%(Lagged)2% < Depr. < 4%(Current)2% < Depr. < 4%(Lagged)Depr. > 4%(Current)Depr. > 4%(Lagged)Exch. rate index growth(Current)Exch. rate index growth(Lagged)Sample sizeAdjusted R^F-stat{prob > F - stat)

Income(1 )

-56 .588(26.408)-15,428(24,271)-68,903(32.039)-12 .045(30.646)76.642(24.647)85.858(25.942)45,482(27.761)76,345(29,791)-19,273(167,134)216.140(175.302)1737.6223.48(,000)

Inc. to Assets(2 )

- . 006(.004),004(,004)- , 0 1 3(005),006(.005)- . 000(.004).010(.004).007(.005),017(,005)- , 000(.030),038(.031)1729.7002.6(.004)

Ln (Sh. Wealth)(3 )

- 0 3 9(.047)- . 0 2 7(.048)- . 034(.058).053(.055).153(.045),114(.047).094(,050).046(.054)- . 0 7 7(.302).237(.316)1713.8732.47(,006)

a. Dependent variable is the level of income in column (1), the ratio of operating income to total assetsin column (2) and the log value of shareholder wealth in column (3), Income and shareholder wealth areexpressed In millions of 1977 dollars, 2% < Appr. < 4% is dummy variable that equals 1 if the industry-


31/33

ARE CEOS REWARDED FOR LUCK? 931

P R I N C E T O N U N I V E R S I T Y , N A T I ON A L B U R E A U O F E C O N O M I C R E S E A R C H , A ND C E N T E RFOR E C O N O M I C P O LI C Y R ES EA R C HM A S S A C H U S E T T S I N S T I T U T E O F T E C H N O L O G Y A N D N A T I O N A L B U R E A U O F E C O N O M I CR E S E A R C H

REFERENCESAbowd, Joh n M., and David S.Kap lan, "Executive Com pensation; Six Questions T ha tNeed Answering," Journal of Economic Perspectives, XIII (1999), 145-168.Aggarw al, Rajesh K,, and Andrew A, Samw ick, "The Other Side of the Trade-Off:The Impact of Risk on Executive Compensation," Journal of Political Econ-omy, CVII (1999a), 65-105,Aggarwal, Rajesh K., and Andrew A. Samwick, "Executive Compensation, Stra-tegic Competition, and Relative Performance Evaluation; Theory and Evi-dence," Journal of Finance. LIV (1999b), 1999-2043.Baker, George, Robert Gibbons, and Kevin J. Murphy, "Subjective PerformanceMeasures in Optimal Incentive Contracts," Quarterly Journal of Economics,CIX(1994), 1125-1156.Baker, George, and Brian Hall, "Understanding Top Management Incentives:Firm Size, Risk and CEO Effort," Harvard Business School, 1998.Bay singer, Robert D,, and H enry N, Butler, "Corporate Go vernance and th e Boardof Directors: Performance Effects of Ch ange s in Board Composition," Jo ur na lof Law, Economics and Organization, I (1985), 101-124.Be rtrand, M arianne, "From the Invisible H andsh ake to the Invisible Han d? HowImport Competition Changes the Employment Relationship," National Bu-

reau of Economic Research Working Paper No. 6900, 1999,Bertrand, Marianne, and Sendhil Mullainathan, "Corporate Governance and Ex-ecutive Compensation: Evidence from Takeover Legislation," Princeton Uni-versity, 1999,Bertrand, Marianne, and Sendhil Mullainathan, "Agents with and without Prin-cipals," Am crfcan Economic Review. XC (2000a), 203-208,Bertrand, Marianne, and Sendhil Mullainathan, "Do CEOs Set Their Own Pay?The Ones w ithout Principals Do," Na tional B ureau of Economic R esearchWorking Paper No. 7604, 2000b.Blanchard, Olivier J,, Florencio Lopez-de-Silanes, and Andrei Shleifer, "What DoFirm s Do with Cash Windfalls?" Journal of Financial Econo mics, XXXVI(1994), 337-360.Brickley, James A., JefB*ey L, Coles, and Rory L, Terry, "Outside Directors and theAdoption of Poison Pills," Joumal ofFinancial Economics, XXXV (1994), 371-390.Byrd, John W,, and Kent A, Hickman, "Do Outside Directors Monitor Managers?"Journal of Financial Economics, XXXII (1992), 195-221,Crystal, Graef, In Search of Excess: The O vercompensa tion of American Executives(New York, NY: W.W, Norton Co,, 1991).Denis, David, and Jan Serrano, "Active Investors and Management TurnoverFollowing Unsuccessful Control Contests," Journal of Financial Economics,XL (1996), 239-266.Diamond, Peter, "Managerial Incentives: On the Near Linearity of Optimal Com-pensation," Journal of Political Economy, CVI (1998), 931-957,Durell, Alan, "Attribution in Performance Evaluation," Dartmouth College, 1999.Garen, John E,, "Executive Compensation and Principal Agent Theory," Journalof Political Economy, CII (1994), 1175-1199.Gibbons, Robert, and Kevin J, Murphy, "Relative Performance Evaluation forChief Executive Officers," Industrial and Lahor Relations Review, XLIII


32/33

9 3 2 QUARTERLY JOURNAL OF ECONOMICSHall, Brian J,, "The Design of Multi-Year Stock Option Plans," Journal of AppliedCorporate Finance. XII (1999), 97-106.Hall, Brian J ., and Jeffrey B, Liebman, "Are CEOs Really Paid Like B ureau crats?"Quarterly Journal of Econom ics, CXIII (1998), 653-691 .Hermalin, Benjamin, and Michael Weishach, "The Effects of Board Compositionand Direct Incentives on Firm Performance," Financial Management, XX(1991), 101-112,Himmelberg, Charles, and R. Glenn Hubbard, "Incentive Pay and the Market forCEOs: An Analysis of Pay-For-Performance Sensitivity," Columhia Univer-sity, 2000,Holmstrom, Beng t, "Moral Haz ard and Observability," Bell Journal of Economics,X (1979), 74-91 ,Holmstrom, Bengt, and Paul Milgrom, "Aggregation and Linearity in the Provi-sion of Intertemporal Incentives," Econonietrica, LV (1987), 303 -328 .Hu bbard, R. Glenn, and D arius Palia, "Executive Pay and Performance: Evidencefrom the U. S, Banking Industry," Journal of Financial Economics, XXXIX(1994), 105-130,Janakiraman, Surya N., Richard A, Lambert, and David F, Larcker, "An Em-pirical Investigation of the Relative Performance Evalu ation Hyp othesis,"Journal of Accounting Research, XXX (1992), 5369.Jensen, Michael, and Kevin J. Murphy, "Performance Pay and Top-ManagementIncentives," Journal of Political Economy, XCVIII (1990), 225-264,Joskow, Pau l L., Nancy L, Rose, and Cath erine D. Wolfram, "Political C on straint son Executive Compensation: Evidence from the Electric Utility Industry,"Rand Journal of Economics, XXVII (1996), 165-182., "Corporate Performance and Managerial Remuneration: An Empirical In-vestigation of Managerial Labor Contracts," Rand Journal of Economics,

XVII (1986), 59-76,-, "Executive Compensation," Handbook of Labor Economics. O, Ashenfelterand D. Card, eds. (Amsterdam, The Netherlands: North-Holland, 1999).Revenga, Ana L,, "Exporting Jobs? The Im pact of Impo rt C ompetition on Em ploy-ment and Wages in U, S, Manufacturing," Quarterly Journal of Economics,CVII (1992), 255-284.Rosenstein, Stuart, and Jeffrey G. Wyatt, "Outside Directors, Board Indepen-dence, and Shareholder Wealth," Journal of Financial Economics, XXVI(1990), 175-191,Shea, John, "Nominal Illusion: Evidence from Major League Baseball," Universityof Maryland, 1999.Shivdasani, Anil, "Board Composition, Ownership Structure, and Hostile Take-overs," Journal of Accounting and Economics, XVI (1993), 167198,Shleifer, An drei, and Robert Vishny, "Large Shareh olders and C orporate C ontrol,"Journal of Political Economy, XCIV (1986), 461-488,Tufano, Peter, "Who M anages Risk? An Em pirical E xam ination of Risk Man age-ment Practices in the Gold Mining Industry," Journal of Finance, LI (1996),1097-1137,W eishach, Michael, "Outside Directors and CEO Turnov er," t/ourna/ of FinancialEconomics, XX (1988), 4 31-4 60.Yermack, David, "Do Corporations Award Stock Options Effectively?" Journal ofFinancial Economics, XXXIX (1995), 237-269., "Higher Market Valuation of Companies with a Small Board of Directors,"Journal of Financial Economics, XL (1996), 185211.Zeckhauser, Richard J., and John Pound, "Are Large Shareholders EffectiveMonitors? An Investigation of Sh are Ow nership and C orporate Performance,"Asymmetric Information, Corporate Finance and Investment, R, Glenn Hub-bard, ed, (Chicago, IL; University of Chicago Press, 1990).


33/33

Date post:	05-Apr-2018
Category:	Documents
Upload:	keila2012nor
View:	217 times
Download:	0 times

Bertrand&Mullainathan.2001.Are CEOs Rewarded for Luck

Documents