Asset‐Level Transparency and the (E)valuation of Asset‐Backed Securities

• Published date: 01 June 2022
• Author: JED J. NEILSON,STEPHEN G. RYAN,K. PHILIP WANG,BIQIN XIE
• Date: 01 June 2022
• DOI: http://doi.org/10.1111/1475-679X.12389

DOI: 10.1111/1475-679X.12389

Journal of Accounting Research

Vol. 60 No. 3 June 2022

Printed in U.S.A.

Asset-Level Transparency and the

(E)valuation of Asset-Backed

Securities

JED J. NEILSON,∗STEPHEN G. RYAN,†K. PHILIP WANG,‡

AND BIQIN XIE∗

Received 26 February 2019; accepted 27 May 2021

ABSTRACT

As of November 2016, SEC Regulation (“Reg”) AB II requires issuers of cer-

tain types of asset-backed securities (“ABS”) to disclose the credit-risk at-

tributes of each asset in the underlying pool, a substantial expansion of prior

disclosure requirements. We examine how ABS issuers’ asset-level disclosures

under Reg AB II affect the (e)valuation of ABS by investors and credit rating

agencies. Using difference-in-differences models that compare affected and

unaffected types of ABS, we find that these disclosures improve the ability of

initial ABS yields and credit ratings to predict the performance of the underly-

ing assets. These results are concentrated in deals with above-median risk lay-

ering in the underlying assets and complexity in the tranching of credit risk.

We further find that asset-level disclosures are associated with lower yields.

∗Smeal College of Business, Pennsylvania State University; †Stern School of Business, New

York University; ‡Fisher School of Accounting, University of Florida.

Accepted by Haresh Sapra. This paper has benefited from comments and suggestions

from Jannis Bischof, Sam Bonsall, Mei Cheng (discussant), Kimberly Cornaggia, Yiwei Dou,

Kai Du, Mark Flannery, Tracie Frost (discussant), Steve Huddart, Christian Laux, Christian

Leuz, Chris James, Mark Ma, Karl Muller, Hong Qu, and Yuehua Tang, and the anony-

mous associate editor and reviewer. We thank seminar and conference participants at the

American University, University of Iowa, University of California, Riverside, the 2018 BYU

Accounting Research Symposium, the 2018 Florida Accounting Symposium, the 2019 FARS

Midyear Meeting, Florida State University, Pennsylvania State University, Temple Univer-

sity, and University of Florida. An online appendix to this paper can be downloaded at

http://research.chicagobooth.edu/arc/journal-of-accounting- research/online-supplements

1131

© 2021 The Chookaszian Accounting Research Center at the University of Chicago Booth School of

Business.

1132 j. j. neilson, s. g. ryan, k. p. wang, and b. xie

Lastly, we provide evidence that most prospective ABS investors download

asset-level information during the price formation period prior to ABS is-

suance.

JEL codes: G21, G24, G28, L1, M41, M48

Keywords: asset-backed securities; asset-level disclosures; Reg AB II; trans-

parency; risk layering; credit rating quality

“Sunlight is said to be the best of disinfectants.”

Louis Brandeis, Other People’s Money and How the Bankers Use It

[1914, Ch. V]

1. Introduction

We examine the impact of transparency about the credit-risk attributes

of individual underlying assets on the (e)valuation of asset-backed securi-

ties (“ABS”) by investors and credit rating agencies. We focus on the ability

of initial yields and credit ratings to predict the future credit performance

of the underlying assets. This research question is important because the

opacity of the assets underlying ABS is widely cited as a primary cause of

the 2007−09 financial crisis (Acharya et al. [2009], Scott and Taylor [2009],

Gorton [2010]). The Financial Crisis Inquiry Report [2011, p. xix] con-

cludes that “a combination of excessive borrowing, risky investments, and

lack of transparency put the financial system on a collision course with cri-

sis.” Ashcraft and Schuermann [2008] explain how information about the

underlying assets is lost in each step of the securitization process: at as-

set origination, when originators sell assets to the issuers, and when issuers

package the assets in complex structured deals and sell the ABS to investors.

Owing to their position at or near the end of this chain, ABS investors have

a poor understanding of underlying asset quality and the risks of ABS (Co-

val, Jurek, and Stafford [2009], Gorton [2010]).

As part of the post–financial crisis effort to reform the securitization pro-

cess, the Dodd-Frank Wall Street Reform and Consumer Protection Act

(“Dodd-Frank Act”) directed the SEC to adopt regulations requiring ABS is-

suers to provide asset-level disclosures (section 942[b]). To implement this

mandate, the SEC developed Regulation AB II (“Reg AB II”), which it is-

sued in September 2014. Reg AB II requires issuers of certain types of ABS

to disclose asset-level credit-risk attributes as of November 23, 2016 (SEC

[2014]).

Reg AB II’s asset-level disclosure requirements represent a promising set-

ting to address our research question for two reasons. First, these require-

ments constitute the first and most significant postcrisis expansion of pub-

lic information about the assets underlying ABS. The prior Regulation AB

(“Reg AB”) only required ABS issuers to provide ABS investors with pool-

level summary statistics for relatively few individual credit-risk attributes,

(e)valuation of asset-backed securities 1133

such as borrower FICO credit scores and loan-to-value ratios. Although use-

ful, such one-dimensional pool-level statistics suppress multiattribute fea-

tures of the underlying assets, notably risk layering (Ryan [2018]). Under

Reg AB II, issuers disclose the exact values of more numerous credit-risk at-

tributes of each asset in the pool, thereby revealing these features.1Second,

Reg AB II’s asset-level disclosure requirements substantially increase disclo-

sures for only certain types of ABS.2The limited scope of these require-

ments enables us to employ a difference-in-differences research design.

Prior empirical research finds that investors did not fully appreciate the

risks of ABS prior to the financial crisis, and as a consequence bore substan-

tial losses during the crisis (Coval, Jurek, and Stafford [2009]). Disclosure

theory generally predicts that public disclosure improves price efficiency

by driving prices closer to fundamental values (Gao [2008]) and improving

investors’ prediction of future payoffs (Goldstein and Yang [2017]). Based

on this prior research, we expect ABS issuers’ asset-level disclosures under

Reg AB II to improve the accuracy of investors’ valuations of the subject

ABS. However, such improvement might not materialize owing to the im-

materiality of individual loans (Ally Financial Inc. et al. [2011]), the highly

disaggregated nature of asset-level disclosures, and investors’ information-

processing constraints (Richardson, Ronen, and Subrahmanyam [2011], p.

482).

Prior research identifies overly optimistic credit ratings of ABS as a key

contributor to the financial crisis (Duyn and Chung [2008], Jones, Tett,and

Davies [2008], Ashcraft, Goldsmith-Pinkham, and Vickery [2010], Jiang,

Wang, and Wang [2018]). Asset-level disclosures may not directly improve

rating quality because rating agencies already had unrestricted access to

granular nonpublic information from the issuers, and none of the big

three rating agencies substantially changed their rating methodologies af-

ter the enactment of the asset-level disclosure requirements.3However,

public disclosure of asset-level information can indirectly improve credit

ratings through enhanced market discipline, as these disclosures improve

1Appendix A provides examples of auto ABS issuers’ disclosures before and after the effec-

tive date of Reg AB II’s asset-level disclosure requirements.

2As discussed below, the limited effect of Reg AB II’s asset-level disclosure requirements to

date is attributable to (1) only certain types of ABS deals being subject to these requirements,

(2) inactive public markets for some of the subject deal types after the effective date of the re-

quirements, and (3) issuers providing asset-level disclosures in practice for subject commercial

mortgage-backed securities (CMBS) deals prior to this effective date.

3Our discussions with auto ABS credit analysts at one of the big three credit rating agencies

confirm that, prior to Reg AB II, these analysts had access to more granular data (including

asset-level data) than was publicly available, because they could request and expect to receive

such information from auto ABS issuers. The analysts further state that they were given all the

information that they felt was necessary to analyze auto deals prior to Reg AB II. Of the big

three credit rating agencies (S&P,Moody’s, and Fitch), only Fitch updated its auto ABS rating

methodologies after the effective date of the asset-level disclosure requirements, and even

it states that “this updated criteria report is substantially unchanged from the prior criteria”

(Fitch [2017]).