Interaction between securitization gains and abnormal loan loss provisions: Credit risk retention and fair value accounting

• DOI: http://doi.org/10.1111/jbfa.12381
• Author: Qiuhong Zhao
• Published date: 01 July 2019
• Date: 01 July 2019

DOI: 10.1111/jbfa.12381

Interaction between securitization gains and

abnormal loan loss provisions: Credit risk

retention and fair value accounting

Qiuhong Zhao

College of Business, TexasA&M University

Corpus Christi

Correspondence

QiuhongZhao, College of Business, TexasA&M

UniversityCorpus Christi, 6300 Ocean Drive,

Unit5808, Corpus Christi, TX 78412, USA.

Email:qiuhong.zhao@tamucc.edu

Abstract

This study investigates whether managers use asset securitization

gains to substitute loan loss provision (LLP) management for earn-

ings management, and, if so, whether the percentage of credit risk

retained affects such a relationship. The literatureprovides evidence

that managers have used securitization transactions to boost earn-

ings. Using 2001–2014 data for a sample of bank holding companies,

I find that managers use securitization gains and LLPs as partial

substitutes and that earnings management from securitization gains

grows at an increasing rateto substitute income increasing LLP man-

agement as the level of risk retention increases. These findings are

consistent with the argument that the higher the level of risk reten-

tion, the greater the potential impact on achieving earnings targets,

given banks’ exerciseof discretion over securitization gains through

estimation of fair value of retained interest. In addition, I document

that the substitution effect between the two tools is non-existent

in the post-SFAS 166/167 period. Taken together, the findings have

timely implications for accounting standards by informing the effect

of risk retention that I measure through earnings management

techniques. Moreover, my findings provide additional support for

improved disclosures on assets-backedsecurities.

KEYWORDS

assets-backed securities, credit risk retention, fair value, securitiza-

tion gains, SFAS166/167

JEL CLASSIFICATION

M41

1INTRODUCTION

This study investigates whether managers use gains from asset securitizations to substitute loan loss provision (LLP)

management for earnings management, and, if so, whether the percentage of credit risk retention of securitized loans

affects such a relationship. Specifically, this study examines the role of increased risk retention in banks’ exerciseof

J Bus Fin Acc. 2019;46:813–842. wileyonlinelibrary.com/journal/jbfa c

2019 John Wiley & Sons Ltd 813

814 ZHAO

discretion over securitization gains among other accounting choices. This study also investigateshow banks that qual-

ify for sales treatment behave after the implementation of SFAS 166/167. Dechow,Myers, and Shakespeare (2010)

provide evidence that managers haveused securitization transactions to boost earnings through mis-valuing retained

interest. This study extends Dechow et al.’s (2010) research by examiningwhether the varying levels of retained con-

tractual interest affect securitizing banks’ ability to manage earnings along with other earnings management tech-

niques.

The credit risk retention requirement is one of the major elements of securitization reform in the Dodd-Frank Act

(2010). On 24 December 2016, the final credit risk retention rules mandated by the Dodd-Frank Act were put into

effect for all classes of asset-backed securities (ABS). Former Congressman and co-sponsor of the Act, Barney Frank,

suggested that this requirement is the ‘single most important part of the bill’ (New YorkTimes, 2013). Specifically, the

regulation requires that sponsors retain, generally,at least 5% of the credit risk for any asset that the sponsor, through

the issuance of ABS, transfers, sells, or conveys to a third party; however,if originators retain an amount of risk, only

the remaining risk (up to 5% total) will be allocated to sponsors. The percentage of retained interest can be lowered

based on underwriting standards; moreover,there is an exemption for qualified residential mortgages. The risk reten-

tion requirement was designed to fix the flaws in ABS markets where subprime loans were originated and then sold

into mortgage-backedsecurities transactions. It remains to be seen whether requiring sponsors of ABS transactions to

maintain ‘skin in the game’ and align interests between sponsors and investorsof securitizations through risk retention

will result in improved asset quality in ABS pools. This study is focused on the effect of varying levels of risk retention

and the potential impact of the Dodd-FrankWall Street Reform and Consumer Protection Act’s risk retention require-

ment, as measured through earnings management techniques.

Managers manage earnings through different channels. Forexample, managers manage LLPs (see Beatty, Chamber-

lain, & Magliolo, 1995; Liu & Ryan, 2006), realized gains or losses of available-for-sale securities (see Beatty & Harris,

1999;Dong, Ryan, & Zhang, 2014), securitization transactions (see Dechow & Shakespeare, 2009; Dechow et al., 2010),

hedging derivatives (see Kilic, Lobo, Raasinghe, & Sivaramakrishnan, 2013), and debt valuation adjustment (see Dong,

Doukakis, & Ryan, 2018) to boost or smooth earnings, regulatory capital and taxes. Specifically, Dechow and Shake-

speare (2009) argue that, compared to collateral borrowing treatment, the sales treatment for securitization provides

flexibility in valuing the retained asset and in determining the size of the gain recorded in the income statement. They

provide evidence that firms will time securitization to inflate earnings in order to meet/beat earnings thresholds. Fur-

ther, Dechow et al. (2010) provide evidencethat managers have used securitization transactions to boost or smooth

earnings and that CEOs are rewarded for the gains they managed on average. They argue that the determination of

fair value of retained interest is discretionary because the determination of an appropriate discount rate to estimate

the fair value is involvedwith forecasting default rates and prepayment rates. Therefore, findings about the role of the

increased risk retention in banks’ exerciseof discretion over securitization gains provide important implications.

LLP is the largest and most judgemental accrual estimate for banks (Beatty & Liao, 2014). In contrast, the use of

derivatives to manage earnings has been limited since the implementation of the derivative fair value accounting rule,

i.e., SFAS 133 (Choi, Mao, & Upadhyay, 2015; FASB, 1998; Kilic et al., 2013). Dong, Doukakis, and Ryan (2018) have

documented that managers use abnormal LLP,abnormal realized securities gains and losses (RSGL) and abnormal debt

valuation adjustments (DVA)as substitutes to manage earnings. However, banks with debt value adjustments consti-

tute a small sample, as documented in Dong et al. (2018) and in Cedergren, Chen, and Chen (2015). Cedergren et al.

(2015) identified 80 banks with debt valuation adjustments from 2007 to 2013. Their final sample consists of only 46

banks with debt valuation adjustments. Banks often report that valuation of debt or derivatives does not have any

material impact on their own credit risk (Deloitte, 2013). Therefore, this study is the first to explore the role of the

increased risk retention level in the interactionbetween gains from asset securitizations and LLP management.

I argue that managers engage in a trade-off between abnormal LLP and securitization gains. First, abnormal LLP

and securitization gains may not be cost-effective. Overstated LLP in one reporting period will have to be reversed,

with the adjustment being subsequently reported. Furthermore, monitoring activities by financial analysts, auditors,

credit rating agencies and boards might actually function as impediments to using abnormal LLP for boosting earnings,