Corporate decision making in the presence of political uncertainty: The case of corporate cash holdings

• Author: Ching‐Wai (Jeremy) Chiu,Anna‐Leigh Stone,Chak Hung Jack Cheng,William B. Hankins
• Published date: 01 May 2020
• Date: 01 May 2020
• DOI: http://doi.org/10.1111/fire.12205

DOI: 10.1111/fire.12205

ORIGINAL ARTICLE

Corporate decision making in the presence of

political uncertainty: The case of corporate

cash holdings

William B. Hankins1Anna-Leigh Stone2

Chak Hung Jack Cheng3Ching-Wai (Jeremy)Chiu4

1Department of Finance, Economics, and

Accounting, School of Business and Industry,

JacksonvilleState U niversity,Jacksonville,

Alabama

2Department of Economics, Finance, and

Quantitative Analysis, Brock School ofBusiness,

Samford University, Birmingham, Alabama

3George Dean Johnson, Jr.College of Business

and Economics, University of South Carolina

Upstate, Spartanburg, South Carolina

4Bank of England, London,UK

Correspondence

WilliamB. Hankins, Department of Finance,

Economics,and Accounting, School of Business

andIndustry, Jacksonville State University,115

CollegeStreet SW, Jacksonville, AL 36265.

Email:whankins@jsu.edu

Abstract

Using a quarterly panel of U.S. corporations over the period 1985–

2014, we show that corporate managers respond to political uncer-

tainty and economic policy uncertainty shocks in different ways. We

proxy for political uncertainty using the Partisan Conflict Index and

employ a prevalent empirical macroeconomic methodology to con-

struct structural shocks that are orthogonal to shocks captured by

the Economic Policy Uncertainty Index. Following a political uncer-

tainty shock, corporations increase cash but do not adjust invest-

ment. Alternatively,following an economic policy uncertainty shock,

firms appear to drawon cash and reduce capital spending to increase

research and development spending.

KEYWORDS

cash holdings, economic policy uncertainty, SVAR, U.S. partisan

conflict

JEL CLASSIFICATIONS

E32, G30, G32

1INTRODUCTION

In June 2016, 79% of the chief financial officers (CFOs) responding to the CFO Outlook Survey stated that the United

States faced “moderate-to-severe”political risk (Graham, 2016). Forty-seven percent of CFOs also indicated that they

would limit their business spending due to heightened political uncertainty.In 2013, half of all chief executive officers

(CEOs) responding to the Business Roundtable’s CEO Economic Outlook Survey claimed that political disagreement

within the federal government overthe upcoming budget negotiations and the looming debt ceiling crisis was likely to

have an adverse effect on their short-term hiring decisions. Recent public news reports document similar anecdotes

about how political uncertainty impacts business decisions. Forexample, Howard Shultz, the former CEO of Starbucks,

Financial Review.2020;55:307–337. wileyonlinelibrary.com/journal/fire c

2019 The Eastern Finance Association 307

308 HANKINS ET AL.

sent a well-publicized memo to employees urging better customer service in the face of, among other issues, “great

political uncertainty both at home and abroad” (Harwell, 2015). Jamie Dimon, the CEO of JPMorgan Chase, has dis-

cussedhis concern regarding uncertainty over health care, immigration, and infrastructure policy (Dimon, 2015). These

issues,particularly the first two, have elicited contentious partisan debate among members of Congress in recent years.

Given the anecdotal evidence, this paper directly addresses whether a political uncertainty shock impacts corpo-

rate cash holdings. Corporations hold cash for a variety of reasons, one of which is to serve as a buffer against neg-

ative shocks that could affect future cash flows (see Almeida, Campello, & Weisbach, 2004; Han & Qiu, 2007; Opler,

Pinkowitz, Stulz, & Williamson, 1999). If corporatemanagers anticipate that heightened partisan conflict would lower

theprobability that the government could effectively respond to an unanticipated negative economic shock, they might

direct their firms to hold more assets as cash to guard against potentially harmful events in the future.

We note that the estimation of political uncertainty is complicated by the existenceof economic policy uncertainty.

In this paper,our measure of political uncertainty is based on the Partisan Conflict Index (hereafter referred to as the

PC Index) first introduced in Azzimonti (2018a). Our measure of economic policy uncertainty is based on the news-

based Economic Policy Uncertainty Index (hereafter referred to as the EPU Index), developed byBaker, Bloom, and

Davis (2016). Although similar in construction, the PC Index is distinct from the EPU Index.In particular, the algorithm

from which the PC Index is constructed does not search for words directly related to economic policy, but instead

searches for words directly related to political disagreement. The EPU Index will only register episodes of political

uncertainty if the source of that conflict is economic uncertainty. Thus, the PC Index is a more complete measure

of political uncertainty. This distinction between the PC Index and the EPU Indexis important because there can be

instances where partisan conflict is low but economic policy uncertainty is high and vice versa. As an example, Azz-

imonti (2018a) notes that immediately following the September 11th terrorist attacks, economic policy uncertainty

was quite high, yet partisan conflict was low because most politicians had rallied around a common goal. Likewise,

following the failure of Lehman Brothers in 2008, the EPU Index exhibited a stark increase while the PC Index was

relatively stable. Conversely, instances where economic policy uncertainty is low,but partisan conflict is high, might

still lead investors and managers to alter their behavior.For example, Azzimonti (2018a) distinguishes between uncer-

tainty about the types of policies the governmentmight adopt, or whether they will adopt a meaningful policy at all, and

uncertaintyover the effects of policies that have already been adopted. Uncertainty over which policies will or might be

adopted stems from political uncertainty. Conversely, uncertainty regarding the effects of existing policy stems from

economic policy uncertainty.1

Toconstruct structural political uncertainty shocks that are orthogonal to economic policy uncertainty shocks, we

followthe macroeconomic literature by estimating a structural vector autoregression model (SVAR). Since Sims (1980),

vectorautoregressions have become a common tool to study dynamical relations between variables. Stock and Watson

(2001) stress that SVARs require identifying assumptions to allow correlations to be interpreted causally.We adopt

the recursiveness assumption that is widely used by researchers such as Christiano, Eichenbaum, and Evans (1999,

2005), and Bloom (2009), among many others. By constructing shocks that are orthogonal to one another, we have

additionalconfidence that partisan conflict shocks are not contaminated byshocks to economic policy uncertainty (and

vice versa) or other macroeconomic variables used in the SVARmodel. We then test the relationships between these

shocks and corporatecash holdings using data from COMPUSTAT quarterly data for all nonfinancial and nonutility U.S.

firms over the period 1985Q1–2014Q4.

We reach severalimportant findings. First, we provide robust empirical evidence that corporate managers shift an

economically significant amount of assets into cash holdings during periods of heightened political uncertainty, for

four quarters after the shock. More specifically,a partisan conflict shock similar to what the United States experienced

during 2010, when the Affordable Care Act was signed into law,would result in an increase in the cash ratio of approx-

imately 26 basis points, representing a 1.77% rise in averagecash-to-total assets.

1Previous research such as Pastorand Veronesi (2012) also shows that political uncertainty and impact uncertainty, which could be triggered by economic

policyuncertainty, are theoretically and empirically different.