Agency, Firm Growth, and Managerial Turnover

• Author: MIHAIL ZERVOS,M. CECILIA BUSTAMANTE,STÉPHANE GUIBAUD,RONALD W. ANDERSON
• Date: 01 February 2018
• DOI: http://doi.org/10.1111/jofi.12583
• Published date: 01 February 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 1 •FEBRUARY 2018

Agency, Firm Growth, and Managerial Turnover

RONALD W. ANDERSON, M. CECILIA BUSTAMANTE, ST ´

EPHANE GUIBAUD,

and MIHAIL ZERVOS∗

ABSTRACT

We study managerial incentive provision under moral hazard when growth opportu-

nities arrive stochastically and pursuing them requires a change in management. A

trade-off arises between the benefit of always having the “right” manager and the

cost of incentive provision. The prospect of growth-induced turnover limits the firm’s

ability to rely on deferred pay,resulting in more front-loaded compensation. The opti-

mal contract may insulate managers from the risk of growth-induced dismissal after

periods of good performance. The evidence for the United States broadly supports

the model’s predictions: Firms with better growth prospects experience higher CEO

turnover and use more front-loaded compensation.

WHEN OWNERSHIP AND CONTROL ARE SEPARATED, firm performance depends cru-

cially on having the right managers at the helm and incentivizing them prop-

erly. Over time, changes in business conditions may call for a change in top

management for the firm to seize new opportunities or overcome challenges.

But this may complicate the task of incentivizing incumbent managers. For in-

stance, if managers anticipate that their tenure at the firm will be short, they

will be reluctant to accept any form of deferred compensation, a standard fea-

ture of incentive contracts. The firm may therefore face a dilemma: By changing

management to adapt to evolving business conditions, it may increase the costs

of incentive provision.

To analyze this tension, this paper introduces the idea of growth-induced

turnover into a dynamic moral hazard framework. Growth-induced turnover

refers to the replacement of top management that is motivated by the need to

∗Ronald W.Anderson is with the London School of Economics. M. Cecilia Bustamante is with the

Robert H. Smith School of Business, University of Maryland. St´

ephane Guibaud is with SciencesPo.

Mihail Zervos is with the London School of Economics. Weare especially grateful to Bruno Biais (the

Editor) and two anonymous referees for helping us improve the paper significantly. We also thank

Catherine Casamatta, Peter DeMarzo, Mike Fishman, Denis Gromb, Zhiguo He, Andrey Malenko,

Guillaume Plantin, Jean-Marc Robin, Antoinette Schoar,our discussants David Dicks, Oliver Spalt,

Sergey Stepanov, and John Zhu, seminar participants at Boston University, ´

Ecole Polytechnique,

London School of Economics, Queen Mary University of London, SciencesPo, Toulouse, and UT

Austin, and conference participants at the meetings of the CEPR European Summer Symposium

in Financial Markets, China International Conference in Finance, European Finance Association,

European Finance Summit, International Moscow Finance Conference, S´

eminaire Bachelier Paris,

and Western Finance Association for helpful comments. We have read the Journal of Finance’s

disclosure policy and have no conflict of interest to disclose.

DOI: 10.1111/jofi.12583

419

420 The Journal of FinanceR

have managers who possess the appropriate skill set and experience to lead

the firm in its current circumstances. This may involve, for instance, adopting

new production techniques, making acquisitions, launching a new product, or

expanding into new markets. If the incumbent lacks the vision or skills neces-

sary to implement such transformations, appointing new management is the

only way for the firm to successfully pursue its course.1Atthesametime,

proper dynamic incentive provision requires a combination of deferred com-

pensation and a threat of dismissal following poor performance, both of which

constitute agency costs. By introducing the possibility of managerial turnover

for the sake of growth as well as for discipline, we show how these costs are

affected. The main insight of the paper is that the prospect of growth-induced

dismissal effectively increases managers’ impatience, which increases agency

costs and generates a tendency to front-load compensation. In fact, the firm

may actually be better off ex ante by committing to pass up otherwise attrac-

tive growth opportunities in some circumstances. More generally, our analysis

delivers empirical predictions on the effects of a firm’s growth prospects on

managerial turnover and compensation, predictions that we show are broadly

supported in the data.

Although our analysis is set up in a continuous-time stationary environment,

we first develop the theory in the context of a two-period model. The simplicity

of the framework enables us to distill most of the economics of the paper in

a transparent way. In particular, the trade-off faced by the firm between the

benefit of having a manager who is able to seize new opportunities and the

cost of incentive provision appears starkly in this setting. Moreover, the key

empirical implications of the theory are derived analytically.

In the continuous-time model, a long-lived firm is run by a sequence of risk-

neutral managers protected by limited liability.A moral hazard problem arises

because, while they are in charge, managers can divert cash flows for their

own private benefit. The firm can fire the incumbent manager at any time and

replace him at a cost. Fleeting growth opportunities arrive stochastically over

time, and a change in management is needed to seize them. If the firm decides to

pursue an opportunity, it pays the costs associated with replacing the manager

and its size (or profitability) increases. A long-term incentive contract is signed

between the firm and its successive managers at the time they are hired.

As in previous dynamic contracting studies, we show that optimal com-

pensation and turnover policies in this environment can be described in

terms of a state variable that coincides with the agent’s expected discounted

compensation, referred to as his contractual promise. The manager receives

1In some circumstances, a change in management may be required to avoid decay, rather than

to pursue growth—for example, when by sticking with the status quo, the firm would fail to face

up to a disruptive competitive threat. For instance, in his narrative of the battle waged in Canada

around 1820 between the long-established Hudson Bay Company (HB) and its upstart rival North

West Company (NW), Roberts (2004, p. 5) recounts that “HB did respond to the threat, essentially

by copying NW’s new approach. It did so, however, only after the leaders of the firm had been

replaced by new ones who understood the nature of the threat and were not tied to the old ways that

had worked so well for so long.” (emphasis added)

Agency, Firm Growth, and Managerial Turnover 421

cash compensation only when his promise reaches an endogenous bonus thresh-

old. When the manager’s promise lies below this threshold, cash compensation

is deferred, and the promise is increased at a contractually specified rate plus

a positive or negative adjustment based on the firm’s current performance. If

the firm suffers a sustained period of poor performance, the manager’s promise

can be lowered sufficiently to reach zero, the firing threshold, at which point

the incumbent is replaced by a new manager who receives an initial promise

that is no less than his exogenous reservation value.

In contrast with other studies, the manager’s contract in our framework is

also contingent on the presence or not of a growth opportunity. If no growth

opportunity becomes available, the manager continues his tenure so long as his

promise stays above the firing threshold, and he is compensated with bonuses

and performance-related changes in his promise as described above. If a growth

opportunity arises and the firm takes it, the manager is replaced. However, not

all growth opportunities are seized by all firms—even though they would be

under first-best. Specifically, we show that, depending on the characteristics of

the firm and its environment, the optimal growth policy can be one of two types.

For some firms, it is optimal to pursue all growth opportunities as they come.

For other firms, it is optimal to forgo opportunities that arise after periods of

good performance, that is, when the incumbent manager’s promise is above a

certain growth threshold. We refer to these two types of firms as high-growth

firms and low-growth firms, respectively. In effect, optimal incentive provision

in low-growth firms calls for some degree of job protection against the risk of

growth-induced termination. Intuitively, the reason job protection is granted

after a spell of good cash flows is that losses due to agency problems are re-

duced after good performance, thus increasing the value of continuing with the

incumbent manager net of the forgone benefit of growth. In high-growth firms,

the benefit of growth always dominates.

Under the optimal contract, managerial compensation is affected by the pos-

sibility of growth-induced turnover through the drift of the manager’s promise

during his tenure. In the absence of growth opportunities, this drift would sim-

ply be equal to the manager’s discount rate. The key novelty in our setup is

that, whenever the firm stands ready to seize an opportunity that might become

available, the drift rate needs to be augmented to compensate the manager for

the risk of growth-induced termination, with the drift modification depending

on the arrival intensity of growth opportunities. This upward adjustment of the

drift when the firm stands ready to pursue a growth opportunity explains why

firms with better growth prospects tend to have more front-loaded compensa-

tion. It also sheds light on why low-growth firms grant job protection when past

performance has been good but not if it has been bad. A higher drift is indeed

less costly to the firm after poor performance, when the manager’s promise

is close to the firing threshold, as it reduces the likelihood of a subsequent

inefficient disciplinary turnover.

Our analysis explicitly allows for the possibility of lump-sum payments, and

we show that severance pay is suboptimal in our setting even in the case of

growth-induced turnover. Indeed, it is always better for the firm to increase the