Moral Hazard, Agency Cost, and Firm Growth

• Author: Rui Li,Mengying Wang
• DOI: http://doi.org/10.1111/irfi.12233
• Published date: 01 September 2020
• Date: 01 September 2020

Moral Hazard, Agency Cost,

and Firm Growth

RUI LI AND MENGYING WANG

Department of Accounting and Finance, University of Massachusetts Boston,

Boston, MA

ABSTRACT

We study how the agency cost implied by the moral hazard problem in a

firm dynamics model affects the life cycle growth pattern of firms. In the

early stage of a firm’s growth, the agency cost restricts the firm’s capital input

and diminishes over time, so that the firm’s growth is driven by efficiency

improvements and an exogenous progress in productivity. In the long run,

when the firm loses its potential to improve efficiency, growth is driven only

by the progress in productivity. As a result of this growth mechanism, consis-

tent with the data, the growth rate and its vola tility, as well as Tobin’sQ,

decrease with age and size. Moreover, the cross-sectional distributions of firm

size and managerial compensation obey a power law, as they do in the data. In

addition, the model provides novel implications for how the characteristics of

the production technology and the preferences of the economic agents affect

the growth pattern of firms, and these implications are poten tially testable.

JEL Codes: G30

Accepted: 8 August 2018

I. INTRODUCTION

The empirical literature has documented two sets of prominent stylized facts:

First, older or larger firms grow more slowly than their younger or smaller coun-

terparts, their growth rates are less volatile, and they have a smaller Tobin’s

Q. Second, the cross-sectional distributions of firm size and managerial com-

pensation are highly skewed to the right and obey a power law. Understanding

the growth mechanism behind these two sets of facts is important because they

are direct outcomes of firm growth, have been documented for decades, and are

very robust across different data sets and time. However, simultaneously

explaining those empirical facts in a frictionless economy would be challeng-

ing.

1

Therefore, in this paper, we introduce moral hazard into an otherwise

standard firm growth framework to understand the underlying growth mecha-

nism for the empirical facts.

1 For example, in such an economy, a constant returns to scale production technology would

imply i.i.d. firm growth, which gives rise to the power law of size distribution (Gibrat’s law),

but it rules out any age and size dependence of firm growth and Tobin’sQ.

© 2018 International Review of Finance Ltd. 2018

International Review of Finance, 20:3, 2020: pp. 639–664

DOI: 10.1111/irfi.12233

Because of the moral hazard problem, in our model, when a new firm is estab-

lished, it is subject to a significant agency cost, which restricts it capital input to

be lower than the efficient level. In the early stage of its life cycle, the agency

cost diminishes so that the firm’s growth is fast and is driven by efficiency

improvements and an exogenous progress in productivity. We say the firm is in

the inefficient state. The firm is expected to lose its potential to improve efficiency

and reach the efficient state in the long run. In the efficient state, firm growth is

slow and driven only by the progress in productivity progress. As a result of this

two-state growth mechanism, the growth rate and its volatility decrease with age

and size. Since large firms keep growing because of productivity progress, the

unbounded growth gives rise to a power law in the firm size and managerial

compensation distributions. Moreover, the decreasing returns to scale produc-

tion function in our model imply negative size and age effects on Tobin’sQ.In

addition, the model provides novel implications for how the parameters related

to the production function and the agents’preferences affect the life cycle

growth pattern of the firms, and these implications are potentially testable.

In our model, a shareholder delegates her firms to a group of managers, with

each manager operating one firm. Over a long time horizon, the firms produce

consumption goods by combining their production technology with capital. We

assume that the firms experience exogenous and unbounded productivity

growth. Moral hazard arises because a firm’s income is subject to random shocks

that are not observable to the shareholder. Hence, the manager could divert the

firm’s income for consumption by reporting a loss from a negative shock. To

deter hidden diversions, the shareholder has to punish the manager by cutting

managerial pay if a low level of income is realized. However, the manager is pro-

tected by limited liability so that liquidation of the firm has to be used as an ulti-

mate punishment after a large amount of reported losses. The punishments have

to be no less than the reported losses, which are proportional to the scale of pro-

duction and the volume of income overseen by the manager. Consequently, the

liquidation probability increases with the capital employment, and then an

agency cost arises, which restricts the firm’s capital at an inefficiently low level.

In our model, the agency cost decreases with the manager’s stake in the firm,

which is measured by the present value of his future compensation relative to

the production scale, because a larger stake on the manager’s part allows the firm

to punish the manager for larger losses without triggering liquidation.

To alleviate the agency cost and raise the capital input, the contract raises

the manager’s stake in the firm by deferring his compensation. Hence, in the

early stage of its life cycle, a firm grows at a higher rate because its growth is

driven by both efficiency improvements and the exogenous productivity pro-

gress. Once capital reaches its first-best level, as expected in the long run when

the manager’s stake reaches an upper bound, the firm loses its potential to

improve efficiency, and its growth is driven only by the productivity progress.

2

2 Notice that the efficient state is not absorbing, as the manager’s stake could decrease when

negative shocks hit and the capital input could become inefficient again.

© 2018 International Review of Finance Ltd. 2018640

International Review of Finance