CORPORATE LIQUIDITY AND NBER RECESSION ANNOUNCEMENTS

• Date: 01 September 2019
• Published date: 01 September 2019
• DOI: http://doi.org/10.1111/jfir.12191
• Author: Benton E. Gup,Anna‐Leigh Stone

The Journal of Financial Research Vol. XLII, No. 3 Pages 637–669 Fall 2019

DOI: 10.1111/jfir.12191

CORPORATE LIQUIDITY AND NBER RECESSION ANNOUNCEMENTS

Anna‐Leigh Stone

Samford University

Benton E. Gup

University of Alabama

Abstract

Using announcement memos released by the National Bureau of Economic Research

(NBER), we show that corporations increase liquidity during the quarter the NBER

announces a peak in the business cycle. This reaction is primarily restricted to

memos about peaks in the business cycle, whether it is a preliminary announcement

or an official confirmation. Federal Open Market Committee and other monetary

policy news, real‐time data releases, and tightening credit conditions do not drive the

increase. This finding adds to the precautionary cash holdings literature by

suggesting that corporations adjust liquidity not necessarily before recessions, but

upon confirmation of a recession.

JEL Classification: G01, G30, G32

I. Introduction

During the Great Recession, corporations came under scrutiny for allegedly hindering

economic recovery by increasing stockpiles of cash to record levels. According to the

Federal Reserve Flow of Funds, U.S. corporate cash crossed the $2 trillion mark in the

second quarter of 2011 and has continued to increase. As such, several theories have

arisen to explain why companies in the United States and around the world are holding

more cash. One suggestion is that companies hold larger stockpiles of cash for

precautionary measures during declines in economic activity.

1

Whereas the literature shows that cash holdings demonstrate a countercyclical

relation across business cycles (see Almeida, Campello, and Weisbach 2004; Ferreira,

Custodio, and Raposo 2005; Baum et al. 2006; Duchin, Ozbas, and Sensoy 2010;

Gao and Xu 2018), our article focuses on recessions and documents exactly when firms

increase cash along the business cycle. Previous studies examine cash holdings across

periods of weak macroeconomic performance; however, these periods can last several

The authors wish to thank the editor, Erik Devos, an anonymous referee, William Hankins, Junsoo Lee,

Jeffrey Wooldridge, and seminar participants at The University of Alabama for their helpful comments and

suggestions.

1

Jacob Goldstein, “Companies Have Been Holding More Cash for Decades,”NPR Planet Money Blog

(September 20, 2011), http://www.npr.org/blogs/money/2011/09/19/140605375/companies‐have‐been‐piling‐

up‐cash‐for‐decades.

637

© 2019 The Southern Finance Association and the Southwestern Finance Association

quarters and are often not identified as a recession until after the fact.Instead of

focusing on how cash holdings behave across recessions, we focus on how cash

holdings behave when the recession is announced.

Using corporate panel data and several macroeconomic indicators during

1976–2015, we document that cash holdings display a U‐shaped pattern across recessions,

declining during early periods of the recession and then increasing. More important,

though, we find that cash holdings actually increase across the period when the National

Bureau of Economic Research (NBER) Business Cycle Dating Committee announces a

start (peak) and end (trough) of a recession, with the primary increase coming during the

quarter in which the peak announcement is made. Specifically, we find an increase in cash

holdings during the quarter of a peak announcement of 1.3%, or approximately $1.8

million. This increase is not present when looking at the quarter before or after the peak

announcement. Additionally, these findings suggest that the macroeconomic indicators

used in past studies may not provide a complete picture of changes in corporate liquidity

during recessions. By documenting when firms increase cash, we are better able to identify

to what extent announcements about the business cycle influence cash holdings and the

macroeconomic information firms use to make this decision. These findings also help

identify the point in the business cycle when corporate managers decide conditions are

severe enough to change operations and begin accumulating additional cash.

We offer three explanations for why corporations appear to adjust cash holdings

concurrently with announcements by the Business Cycle Dating Committee: (1) companies

value the confirmation made by experts on the committee and increase cash holdings

accordingly; (2) at the time a recession is announced, corporate officers have observed the

same trends in the economy and have adjusted cash holdings accordingly; and (3)

corporations are not responding directly to the announcements, but instead are responding

to tightening conditions brought on by a confirmation that the economy is in a recession.

Figure I, which plots cash as a percentage of corporate assets from January

1976 to December 2015, provides support that examining NBER recessionary periods

and NBER peak and trough announcements yields differing results. The figure reports

both the mean and median of cash holdings, which are obtained by using all

nonmissing observations in Compustat. Panel A shades the actual recession periods as

determined by the NBER, and Panel B shades the dates between the NBER Business

Cycle Dating Committeeʼs peak announcement and its trough announcement. Panel A

shows that declines in both the mean and median of cash holdings occur during

recessions, except for the 1982 recession. Panel B shows that both the mean and

median of cash holdings increase at the peak announcement. Specifically, for most of

the series, we see an increase during the quarter a peak announcement is made,

regardless of the lag between the actual beginning of the recession and its

announcement. Figure I also reveals that except for the most recent recession, cash

holdings increase to levels greater than those before the start of a recession. During the

most recent recession, cash holdings increase to their precrisis level. This postrecession

increase possibly indicates that companies were not holding cash just because of the

recession, but because the cash management strategies of companies changed.

To understand how the response to NBER announcements have developed over

time, we examine four 10‐year intervals surrounding each recession so that the recessions

638 The Journal of Financial Research

can be examined independently of each other. The findings show that the two most recent

recessions are driving the increase in cash holdings during the quarter of a peak

announcement. Upon further examination, using annual data, we find that part of the

reason for this finding is our definition of cash holdings, which includes short‐term

investments, and the trade‐off that businesses face between interest rates and liquidity.

During the last two recessions, when interest rates were at their lowest levels in 40 years,

we find that on average firms were holding more cash than investments. During the 1980s

and 1990s, however, firms increased their holdings of short‐term investments because

higher interest rates increased the opportunity cost of holding cash.

As a placebo test, we also examine several other announcements made by the

NBER to see whether similar patterns arise. We first look at “preliminary

announcement”memos. These memos do not formally announce the start of a

recession, but detail the continuing efforts of the committee to monitor conditions to

identify whether a peak has occurred. There are three quarters during which the NBER

releases four preliminary announcements, and cash holdings increase during each

quarter. Surprisingly, the increase lasts only for one quarter before reverting to

preannouncement levels. We also examine three other types of announcements made

by the NBER: preliminary trough announcements, trough announcements, and

procedural memos. We find that firms hold less cash than average during these

announcements. These placebo tests confirm that corporations are not increasing cash

in reaction to just any announcement made by the NBER, but that the increase happens

during the quarter the NBER announces a peak or the potential onset of a peak.

To determine which type of firm might be driving the results, we compare the

cash holdings of financially constrained firms and unconstrained firms during the

quarter of a peak announcement. However, using several common measures of

financial constraint and one based on bank dependence, we find that constrained and

unconstrained firms show no statistically significant difference in increased cash

Figure I. Cash as a Percentage of Total CorporateAssets. This figure shows cash as a percentage of total assets

from January 1976 to December 2015. The values are the mean and median values of cash holdings of all

firms in Compustat (excluding financial firms and utilities) with nonmissing observations. The shaded

areas represent recessions as defined by the National Bureau of Economic Research (NBER) (Panel A)

and periods between NBER peak and trough announcement memos (Panel B).

639Corporate Liquidity