A Note on Capital IQ's Credit Line Data

• DOI: http://doi.org/10.1111/fire.12111
• Published date: 01 August 2016
• Date: 01 August 2016

The Financial Review 51 (2016) 435–461

A Note on Capital IQ’s Credit Line Data

Ani Manakyan Mathers∗

Salisbury University

Emanuela Giacomini

University of Florida

Abstract

Empirical work in finance is increasingly using Capital IQ’s detailed data on capital

structure. We compare the Capital IQ credit line data to hand-collected data for a random

sample of firms. Missing values in Capital IQ are prevalent, so the data set underreports the

importance of corporate credit lines. When data are reported, Capital IQ often differs from

hand-collected credit line activity. We suggest methods for correcting the errors in the Capital

IQ data, note which portions of the data are most reliable, and quantify the effects of Capital

IQ’s underreporting by examining the tradeoff between cash and lines of credit.

Keywords: Capital IQ, credit line

JEL Classifications: G21, G32

∗Corresponding author: Perdue School of Business, Salisbury University, 1101 Camden Ave, Salisbury,

MD 21801; Phone: 410-548-5695; Fax: 410-677-5374; E-mail: ammathers@salisbury.edu.

We would like to thank the editors and two anonymous referees for their thoughtful comments and

suggestions. We also appreciate the many useful comments we received from seminar participants at

Salisbury University, and discussants and participants at the 2014 Southwestern Finance Association

meeting and the 2015 Eastern Finance Association meeting.

C2016 The Eastern Finance Association 435

436 A. M. Mathers and E. Giacomini/The Financial Review 51 (2016) 435–461

1. Introduction

Capital IQ recently began compiling detailed data on firms’ capital structure,

including data from 2001 to present on drawn and available amounts on corporate

credit line contracts. These data became widely available to academics around 2010,

resulting in a growing number of papers that are making use of the detailed data.

For example, Ippolito and Perez (2012), Colla, Ippolito and Li (2013), Nini (2013),

Acharya, Almeida, Ippolito and Perez (2014a,b), Berg, Saunders and Steffen (2016),

Dudley and Qie (2014), McCumber (2014), Nikolov, Schmid and Steri (2015), and

Tengulov (2015) look at drawn amounts outstanding as reported in Capital IQ in

various analyses of firms’ debt composition. Sun (2014) and Garcia-Appendini and

Montoriol-Garriga (2013) employ Capital IQ to collect data on unused lines of credit

as part of a measure of financial slack. Ippolito and Perez (2012) and Acharya,

Almeida, Ippolito and Perez (2014a,b) utilize Capital IQ to identify whether a firm

has access to credit lines. In this paper, we compare the drawnand undrawn (available)

amounts in Capital IQ to hand-collected data on credit line usage for a random sample

of firms. The goal is to investigate the accuracy of Capital IQ’s reported data as

compared to carefully hand-collected data and to provide suggestions for researchers

who wish to use the Capital IQ credit line data.

We identify severalconcerns with the Capital IQ data on credit line usage. First,

Capital IQ often reports missing values when there are data available on credit line

usage and availability in the company’s 10-K filings. Worse, this lack of reporting is

not consistent over time. For example, in the case of Kimberly Clark Corporation,

Capital IQ reports nonmissing data only in 2006. In all other years from 2001 to 2011,

we find varying amounts of availability on credit lines reported in 10-K filings with

the Securities and Exchange Commission (SEC), but Capital IQ reports only blanks.

As shown in Appendix A, Capital IQ reports a nonmissing value of the drawn amount

outstanding for 2001 and 2005–2010 for Alaska Air Group. From 2002 to 2004 and

2011, Capital IQ reports a missing value for the drawn amount outstanding despite

the line being fully drawn at the end of fiscal year 2002 and the data appearing in

the same section, and footnote within the 10-K filing. We particularly emphasize the

Alaska Air Group example because part of the difference in reporting from Capital

IQ to the hand-collected data is a difference in judgment. We do not include the

asset-backed pre-delivery payment (PDP) credit line that is solely available to pay

for aircraft, but Capital IQ does include drawn amounts from that line in 2005–2008.

This difference in judgment is reasonable; however, Capital IQ is inconsistent. They

report only blanks for the Undrawn amount of the line available in every year from

2001 to 2011, despite the positive availability on the PDP and the existence of an

undrawn bank line of credit. Therefore a researcher using Capital IQ’s Undrawn

Revolving Credit variable would incorrectly report that Alaska Air Group had no

credit line availability in 2003 and from 2005 to 2011.

We also findthat using Capital IQ’s reported availability far underrepresents the

number of firm-year observations in which the sampled companies have access to