Accounting is conservative. Under Generally Accepted Accounting Principles (GAAP), a firm should recognize losses before the losses are realized but postpone the recognition of gains until the gains are realized. In other words, the conservative accounting principle promotes timely loss recognition at the expense of timely gain recognition. However, in debt contracting, where lenders and borrowers often negotiate a set of customized accounting rules, more and more loan agreements have included cost savings and synergy add-backs (cost-synergy add-backs hereafter). These add-backs allow borrowers to include unrealized cost savings and synergy gains in contractual earnings calculations, thereby suggesting timely gain recognition in debt contracting. The prevalence of cost-synergy add-backs in the past few years has drawn regulatory attention and made media headlines. For example, the Financial Stability Board (2019) warns that cost-synergy add-backs are opaque and may cause market mispricing. My recent study aims to provide timely empirical evidence on the determinants of the inclusion of cost-synergy add-backs in loan contracts and the implications of this adjustment for debt-contracting efficiency.
Theoretical development
Cost-synergy add-backs increase contractual earnings. However, these add-backs may or may not improve the informativeness of contractual earnings. Cost savings and synergies are gains from structural changes such as merger and acquisitions (M&As), leveraged buyouts (LBOs), and other cost reduction activities. Such structural activities often require significant spending upfront and generate potential gains afterwards. Under conservative accounting, firms are not allowed to recognize future cost savings and synergies even though the expenses incurred to generate these future gains have been recognized. By recognizing cost savings and synergy gains in a timely fashion, cost-synergy add-backs may correct the accounting distortion about future gains under conservative accounting and increase earnings informativeness. On the other hand, due to the unverifiable nature of future gains, cost-synergy add-backs may overstate earnings and decrease earnings informativeness.
By affecting earnings, cost-synergy add-backs further affect debt-contracting efficiency. According to incomplete contract theory, an efficient contract allocates control rights to borrowers when borrowers have good performance and to lenders when borrowers have poor performance. The allocation of control rights depends on debt covenants, which often use contractual earnings as an input. There are two types of allocation error: false positives that occur when covenants are violated even though the borrower has good performance and false negatives that occur when covenants are not violated even though the borrower has poor performance. An efficient contract minimizes the sum of costs from false positives and false negatives. Based on this efficiency notion, Gigler et al. (2009) suggest that optimal accounting in debt contracting should be informative about borrowers’ performance. By contrast, conservative accounting reduces false negatives but increases false positives, which may ultimately reduce contract efficiency. Therefore, if
cost-synergy add-backs improve earnings informativeness, they may reduce the probability of false positives and increase contract efficiency. However, if cost-synergy add-backs overstate earnings, they may increase the probability of false negatives and reduce contract efficiency.
Everything else equal, borrowers should prefer including cost-synergy add-backs in loan agreements because these add-backs give them the option to increase contractual earnings. But why do lenders allow the inclusion of these add-backs? To answer this question, it is important to understand lenders’ monitoring incentives. If lenders perform monitoring, borrowers are likely to include credible cost-synergy add-backs, which in turn improve earnings informativeness and reduce false positives (the earnings informativeness hypothesis). However, lenders may lack strong monitoring incentives when they reach for yield in an overheated market. Reaching for yield is a practice where lenders take on excessive risks to maintain high nominal yields despite the low risk-adjusted returns. Prior research finds that lenders may take on excessive risks through complex contract design that weakens creditor protection (Kaplan and Stein 1993; Stein 2013; Ivashina and Vallee 2018). Therefore, if lenders reach for yield in an overheated market, they may accept cost-synergy add-backs to boost nominal yields even though borrowers may use these add-backs to inflate earnings without lender monitoring (the market overheating hypothesis).
I exploit recent developments in private lending to identify lenders’ incentives and differentiate these two hypotheses. The loan market has evolved into two distinct market segments—the leveraged and non-leveraged loan markets. In the non-leveraged loan market, banks are the major players that perform extensive monitoring of borrowers and benefit from improved earnings informativeness (Roberts and Sufi 2009; Chava et al. 2019). I thus predict that the inclusion of cost-synergy add-backs in non-leveraged loans is consistent with the earnings informativeness hypothesis. The leveraged loan market, on the other hand, is a rapidly developing market where non-bank lenders such as collateralized loan obligations (CLOs) and loan funds dominate. The excessive risk-taking by non-bank lenders and weak bank monitoring may result in potential market overheating (Stein 2013; Wang and Xia 2014; Ivashina and Vallee 2018; Aramonte et al. 2019). Leveraged lenders may thus accept cost-synergy add-backs to reach for yield. I therefore predict that the inclusion of cost-synergy add-backs in leveraged loans is consistent with the market overheating hypothesis.
Empirical findings
I extract data on the inclusion of cost-synergy add-backs in loan agreements for U.S. loans issued between 1998 and 2019. I first document the growing presence of cost-synergy add-backs. The proportion of loans with cost-synergy add-backs gradually increased from close to zero in 1998 to 18% in 2007 before dropping to 10% in 2008. However, since 2010, these add-backs have regained momentum, appearing in about 33% of loans in 2019.
Next, I provide validation analyses on the differences in lenders’ incentives between the two markets. I find that leveraged loans are more likely to include cost-synergy add-backs and are less likely to impose restrictions on these add-backs than non-leveraged loans, despite the fact that leveraged loans are taken by riskier borrowers. Further, I find that lead lenders’ monitoring incentives in leveraged loans, when contracting on cost-synergy add-backs, are likely undermined by their lack of skin in the game and their profit from syndication fees. I do not find such evidence in non-leveraged loans. These results suggest that leveraged lenders have weaker monitoring incentives than non-leveraged lenders when contracting on cost-synergy add-backs.
I then develop three ex-ante predictions to test the two hypotheses separately in each of the two markets. I first examine the determinants of loan agreements including cost-synergy add-backs. I find that the inclusion of cost-synergy add-backs in non-leveraged loans increases with the informativeness of borrowers’ past contractual adjustments. By contrast, while the inclusion of cost-synergy add-backs in leveraged loans is not associated with the informativeness of borrowers’ past contractual adjustments, it is positively associated with the extent of market overheating captured by the total market share of leveraged loans, total CLO and loan fund flows, and loan-level investor pressure during loan syndication. These findings suggest that the inclusion of cost-synergy add-backs is driven mainly by lenders’ demand for informative earnings in the non-leveraged market and the extent of market overheating in the leveraged loan market.
Second, I test the relation between the inclusion of cost-synergy add-backs and covenant design. Consistent with the earnings informativeness hypothesis, I find that non-leveraged loans with cost-synergy add-backs tend to include more performance covenants and fewer capital covenants. By contrast, I report that the inclusion of cost-synergy add-backs in leveraged loans is associated with weaker creditor protection manifested in fewer performance and capital covenants, less strict financial covenants, and fewer conservative modifications in net worth covenants.
Third, I examine how the inclusion of cost-synergy add-backs affects loan pricing. In the non-leveraged loan sample, I find that lenders charge higher interest rates to compensate for the risks of false negatives and the incremental monitoring costs associated with cost-synergy add-backs. By contrast, in the leveraged loan sample, the inclusion of cost-synergy add-backs is not associated with higher interest rates, consistent with the idea that risks taken by reaching-for-yield lenders are not well compensated for in an overheated market. Overall, results on the three ex-ante predictions collectively support the earnings informative hypothesis in the non-leveraged market and the market overheating hypothesis in the leveraged loan market.
I supplement these ex-ante analyses by providing ex-post evidence on the implications of cost-synergy add-backs. I first find that the expected cost savings and synergy gains added back in realized contractual earnings are negatively associated with borrowers’ future operating cash flows in the leveraged loan sample, indicating that cost-synergy add-backs signal worse future performance in this market. Next, I show that the inclusion of cost-synergy add-backs in leveraged loans, but not in non-leveraged loans, is associated with higher subsequent goodwill impairment. This finding shows that expected cost-synergies in leveraged loans are less realizable. Lastly, I find that cost-synergy add-backs reduce false positives but do not increase false negatives when they are included in non-leveraged loans. However, these add-backs increase false negatives but do not reduce false positives when they are included in leveraged loans. I thus conclude that the inclusion of cost-synergy add-backs in non-leveraged (or leveraged) loans enhances (or distorts) contract efficiency.
Contribution
This paper provides the first large-sample empirical evidence on the implications of cost-synergy add-backs in loan contracts. My findings complement prior research examining lenders’ desired accounting properties through contractual adjustments. A maintained assumption in the prior studies is that lenders use contractual adjustments to facilitate monitoring. My evidence that cost-synergy add-backs are an inefficient contract design in the leveraged loan market suggests that this assumption may not hold when the market overheats. The results also highlight the importance of identifying lender incentives in examining the role of contractual accounting measures.
This study should be of interest to accounting standard setters. GAAP accounting does not allow timely gain recognition. However, my findings suggest lenders may desire timely gain recognition. This paper also carries important policy implications for banking regulators. Detecting potential market overheating is a critical step in maintaining financial stability. My results show that the inclusion of cost-synergy add-backs signals market overheating in leveraged lending, which calls for close monitoring of the development of the leveraged loan market.
Shushu Jiang is an accounting PhD candidate at the University of Toronto, Rotman School of Management.
This post is adapted from her paper, “Timely ‘Gain Recognition’ in Debt Contracting—Evidence from Cost Savings and Synergy Add-backs”, available on SSRN.
