Liquidity Constraints, Competition, and Markup Cyclicality

• Published date: 01 August 2016
• Author: Claudio Raddatz,Matías Braun
• Date: 01 August 2016
• DOI: http://doi.org/10.1111/fima.12121

Liquidity Constraints, Competition,

and Markup Cyclicality

Mat´

ıas Braun and Claudio Raddatz∗

This article provides evidence on the relation among financial constraints, competition, and the

cyclicality of markups. Based on a long series of industry data from a largenumber of countries,

we find that markups increase in conjunction with the business cycle in environments with higher

short-term financial constraints (liquidity constraints) and more competition. The evidence also

suggests that these two elements complement each other: the procyclicality of markups in firms

facing both high competition and high liquidity constraints is higher than that explained by each

determinant independently.

There is ample evidence that financial constraints affect the production of firms experiencing

shocks (see, e.g., Gertler and Gilchrist, 1994; Bernanke, Gertler, and Gilchrist, 1996). As stressed

by Braun and Larrain (2005), the response of firms to negative shocks depends on their reliance

on financial markets. In this situation, financially constrained f irms that are unable to raise funds

for investment in physical or working capital are forced to trim production or even shut down

entirely. The flipside of this mechanism is that these firms will also likely try to raise as much

internal funds as possible when facing difficulties. One way of doing it is through pricing. This

implies that the behavior of prices and markups around the business cycle will likely depend on

the degree of financial constraints faced by fir ms.1

The competitive environment is another major determinant of this behavior, as pricing and

markup decisions are critically affected by the strategic interaction between a firm and its

competitors. Moreover, these two elements are not necessarily independent because the mar-

ket structure may alter the ability or willingness of firms to ease liquidity constraints through

pricing decisions. For instance, if firms build market share by keeping prices artificially low

during normal times, markups should decline less during downturns for firms whose f inancial

constraints become binding. This is because the incentive of these firms to invest in market share

declines during recessions (Chevalier and Scharfstein, 1995, 1996). However, it is also possible

that price wars that take place during downturns cause markups to decline more in downturns

as firms race to the bottom to create revenues. These price wars can be triggered by cyclical

changes in the elasticity of demand or other changes in the competitive environment, but they

may also be part of the strategy of some firms that try to drive financially weaker competitors off

the market, as proposed by the “long purse” view of predation. Under this view, the presence of a

financially constrained competitor during a downturn may prompt its financially unconstrained

rivals to pursue predatory strategies to capture the market, also leading to markups that are more

Weare grateful to the Raghavendra Rau (Editor) and an anonymous refereefor their valuable suggestions. Wealso thank

Carlos Alvarado and Rolando Campusano fortheir able assistance. The views in this article are the authors’ and do not

necessarily represent those of the CentralBank of Chile.

∗Mat´

ıas Braun is at Escuela de NegociosUniversidad Adolfo Ib ´

a˜

nez & IM Trust–CredicorpCapital. Claudio Raddatz is

at Banco Central de Chile.

1Of course, the cyclical behavior of marginal costs also matters.

Financial Management •Fall 2016 •pages 769 – 802

770 Financial Management rFall 2016

procyclical in industries with these characteristics (Telser, 1966; Bolton and Scharfstein, 1990;

see also Xu and Byoun, 2015).

This article relies on a large international cross-industry data set to study how the presence

of short-term financial constraints (hencefor th, liquidity constraints) and the degree of product

market competition increase or reduce the cyclicality of markups. We follow a difference-in-

differences identification strategy based on Rajan and Zingales (1998) to provide systematic

empirical evidence on whether the strength of financial constraints makes markups more or less

pro- or countercyclical and howthis phenomenon relates to the competitive environmentprevalent

in each industry.

Understanding how firms’ pricing decisions respond to the financial restrictions they face

under different competitive environments provides relevant information on the type of firms that

are most affected by lack of proper financing. This is important for correctly interpreting and

quantifying the well-documented effects of financial constraints on real variables such as growth,

investment, and capital structure, as firms that are better able to alter their pricing and markup

behavior in response to financial constraints during the aggregate business cycle may be less

forced to alter production and investment plans.

Despite relying on a variety of empirical approaches, various ways of measuring markups,

and different levels of data aggregation, the existing empirical evidence on the relation between

markup cyclicality and financial constraints is still inconclusive.2For instance, Chevalier and

Scharfstein (1995, 1996) and Campello (2003) find that markups are more countercyclical in

more financially constrained environments, or when industry debt is high. However, Botasso

and Sembenelli (2001) and Busse (2002) show that firms facing tighter financial constraints are

more likely to cut prices and start price wars. A few recent papers study how market structure

and financial conditions jointly determine fir ms’ nonfinancial decisions and document a relation

between these two dimensions (Hoberg and Phillips, 2010; Makaew and Maksimovic, 2013).

Some of them suggest that the cyclicality of profitability is higher when the degree of competition

is higher and financial constraints are more prevalent (Pontuch, 2011).

Using data from a large number of countries with different degrees of financial market depth

provides much more variation in the degree of financial constraints and product market competi-

tion than used in previous single-country studies. This increases the power of statistical tests and

reduces external validity concerns. It also allows us to tackle some of the measurement issues

that plague this literature because the additional dimensions of the data can be used to control

for some sources of measurement problems in nonparametric ways. Indeed, any measurement

problem that affects similarly the cyclicality of markups for all industries in a given country or

the same industry across countries can be controlled for in our setting.

Our results show that markups, as measured by an industry’s price-cost margin, are procyclical

on average, with an estimated cyclicality of 1.2. This means that a 1 percentage point increase

(decrease) in the real gross domestic product (GDP) growth rate is associated with a 1.2% increase

(decrease) in the level of markups. We also find that markups are more procyclical in industries

facing a higher level of competition, which is consistent with models where dynamic collusion

is more difficult to sustain when markets are booming. Markups are also more procyclical in

industries facing higher liquidity constraints (i.e., those with high needs for short-term external

financing that are located in less financially developed countries). This is consistent with either

a setting where constrained firms are willing to lower prices during recessions to generate cash

or a setting where a few unconstrained firms prey on their constrained competitors who are

unable to fight off competition. The estimated cyclicality increases from 0.45 to 1.39 when

2In fact, there is not even agreement on whether markups are pro- or countercyclical.

Braun & Raddatz rLiquidity Constraints, Competition, and Markup Cyclicality 771

moving from a state of lowliquidity constraints and low competition to one with more financially

constrained firms and higher competition. This heterogeneity could explain the differences in

findings across studies focusing on various countries or industries. For instance, US industries

may be more countercyclical than our average industry because of their lower degree of liquidity

constraints.3A large set of robustness checks indicate that these findings are unlikely to be

driven by mechanisms different from the interaction among liquidity constraints, competition,

and markup cyclicality.

Also, some of our results suggest that the two forces at play—competition and finance—

are not independent. In particular, we show that liquidity-constrained industries facing higher

competition have more procyclical markups than those facing weaker competition. We also

find that the presence of financially unconstrained f irms with a relatively low market share in

an industry tends to make markups more procyclical, especially when the average firm in that

industry is likely to be liquidity constrained. These results are in line with the presence of predatory

incentives in liquidity-constrained industries with high competition. Financially stronger firms

in these industries have higher incentives to lower their prices during recessions to predate their

weaker competitors.

Our analysis of the data provides evidencein f avor of many of the traditional models. However,

none of them can explain all of our results at the same time. Instead, our findings indicate that

several mechanisms proposed bythe theoretical literature are relevant in explaining the cyclicality

of markups.

In the next section we provide the theoretical background for our research questions. Section

II describes the data and methodology used in our analysis. Section III presents and discusses

the main results. Section IV provides some robustness exercises. Section V extends our results

to address the mechanisms behind our main findings. We discuss how our results relate to the

theoretical literature in Section VI and then conclude in Section VII.

I. Theoretical Underpinnings

The cyclical behavior of markups has long been of interest to various fields of economics.

In macroeconomics, it has been recognized at least since Keynes that a countercyclical markup

may help reconcile the countercyclical marginal product of labor under fixed technology with the

procyclical real wages observed in the data. In industrial organizations, the behavior of markups

is closely tied to the literature on dynamic oligopoly and price wars, and in corporate finance, the

role of financial constraints and product market competition on markups and profits has been a

focus of attention. In this section, we review the theoretical models that relate competition and

financial constraints to markup cyclicality to establish a theoretical basis for the interpretation of

our results.

A. Competition and Markup Cyclicality

In a perfectly competitive environment, firms charge their marginal cost and there are no

markups. Thus, most theories of markup cyclicality come from models of imperfect competition.

3Our study of the determinants of markup cyclicality across many countries is itself a contribution because few papers

explore these determinants in a systematic way. One notable exception is Oliveira Martins and Scarpetta (1999). After

correcting for a number of measurement problems, they conclude that markups are strongly countercyclical in the United

States but not so much in the other G-5 countries they study.As mentioned above, our broader sample includes countries

where the average firm is much more likely to be financially constrained than in theirs, which gives us much more

variation in this dimension and is a relevant extension of their work.