Evaluating the SDC Mergers and Acquisitions Database

• Date: 01 November 2014
• DOI: http://doi.org/10.1111/fire.12057
• Published date: 01 November 2014

The Financial Review 49 (2014) 793–821

Evaluating the SDC Mergers and

Acquisitions Database

Beau Grant Barnes∗

Washington State University

Nancy L. Harp

Clemson University

Derek Oler

TexasTech University

Abstract

We compare 20 years of data from Thompson Financial SDC Platinum (SDC)’sMergers

and Acquisitions database with a hand-collected database, providing evidence on the complete-

ness and accuracy of SDC data across time. We find that our hand-collected data is generally

more accurate than SDC, but SDC’saccuracy and coverage improves over time. Our investiga-

tion of discrepancies between the databases finds that SDC is more prone to errors on smaller,

high book-to-market acquirers with weak announcement period market responses. Preliminary

analyses suggest that this potential bias is not significant, but could affect inferences when

examining smaller, high book-to-market firms.

Keywords: SDC, mergers, acquisitions

JEL Classifications:G34

∗Corresponding author: Department of Accounting, WashingtonState University, 237G Todd Hall, Pull-

man, WA99164-4729; Phone: (509) 335-4472; Fax: (509) 335-4275; E-mail: beau.barnes@wsu.edu.

We are grateful for the contribution of an initial data set of 947 acquisitions provided by TimLoughran.

We are also grateful for the research assistance of Sari Ruben, Andrea Caggiano, Khary Barnes, Zenyep

Eroglu, and Troy Engstrom in the development of this database. We are also thankful for the thoughtful

comments received during the review process.

C2014The Eastern Finance Association 793

794 B. G. Barnes et al./The Financial Review 49 (2014) 793–821

1. Introduction

Researchers require data that is relatively complete and accurate to draw mean-

ingful conclusions. The Thompson Financial SDC Platinum (SDC) database has fa-

cilitated a significant amount of research into mergers and acquisitions (MA) and has

been cited regularly in major accounting and finance journals over the past decade.1

Some authors express concern over the completeness of the information provided

by SDC, but there is little agreement on the precise year that data in SDC becomes

reliable enough to warrant its use in research. For example, Aboody, Kasznik and

Williams (2000, p. 272) report that they begin their sample with 1991 because “this

is the first year for which the acquisition data provided by SDC are complete.” In

contrast, Baker and Savasoglu (2002, p. 97) use 1981 as their starting year, noting

that “prior to 1981, SDC does not provide full coverage of mergers and acquisitions.”

Other papers using SDC typically report a starting year within these two extremes,

in one case stating that “coverage is more likely to be spotty in the 1980s than in the

1990s” (Pontiff and Woodgate, 2008, p. 931).

The accuracy of data reported in SDC’s MA database has also been questioned.

For example, Bharadwaj and Shivdasani (2003) report using the acquirer’s 14D-1

Securities and Exchange Commission (SEC) filings as a better source of information

for financing data, where available, because SDC’s data appears to be less reliable.

Faccio and Masulis (2005, p. 1351) use the SDC MA database for mergers between

1997 and 2000 and note that they “ ...found (and fixed) a number of mistakes in the

SDC database” when attempting to calculate merger size. While concerns over SDC

accuracy have accumulated in the literature, there has been little empirical work that

addresses the degree of completeness and accuracy of the SDC MA database.2

Our paper complements the recent work of Netter, Stegemoller and Wintoki

(2011), who examine the effect of various screens placed on data from SDC. They

show that the conclusions drawn by researchers on acquisition waves and the market

response to acquisition announcements are sensitive to the screens used (e.g., if

private target firms are excluded, or if only acquirers with Compustat data available

are used). While they focus on the effect of broad and commonly used data screens,

we focus on observations from SDC that meet the more commonly used screens (both

the target and acquirer must be included in the CRSP database, the acquirer must

also be included in Compustat, and the target must be 100% acquired and delisted).

Our study also contrasts Netter, Stegemoller and Wintoki (2011) in that we focus on

providing detailed information about the accuracy and completeness of SDC across

1A search for articles in the Journal of Finance, Journal of FinancialEconomics, Journal of Accounting

and Economics, Journal of Accounting Research, Review of Accounting Studies, Review of Financial

Studies,andtheAccounting Review from 2000–2010 yields 52 MA studies that reference SDC.

2One exception regarding accuracy is Fuller,Netter and Stegemoller (2002) who, in their investigation of

1990s acquiring-firm returns, randomly sample 500 acquisitions and attempt to verify the corresponding

SDC announcement date. They report an SDC error rate of 7.4%.