Window Dressing of Short-Term Borrowings

Window Dressing of Short-Term Borrowings

Edward L. Owensa and Joanna Shuang Wub William E. Simon Graduate School of Business Administration, University of Rochester,

Rochester, NY 14627, United States

February 2012 First Draft: March 2011

_______________________________

We thank Jose Berrospide, Dan Collins, Anya Kleymenova, Anzhela Knyazeva, Stephen Ryan, Jerry Zimmerman, and workshop participants at George Washington University, London Business School, New York University, University of Rochester, the University of Minnesota 2011 Empirical Research Conference, and the Fifth Annual Toronto Accounting Research Conference for helpful comments and suggestions. a Tel.: +15852751079; email: edward.owens@simon.rochester.edu. b Tel.:+15852755468; email: wujo@simon.rochester.edu.

Window Dressing of Short-Term Borrowings

Abstract We investigate bank holding companies' window dressing of quarter-end short-term borrowings. We find evidence of downward window dressing of short-term borrowings through repo and federal funds liabilities that appears material for a large fraction of the sample. Such downward window dressing is more pronounced at banks with a higher concentration of short-term borrowings in their total liability structure and lower capital adequacy ratios, consistent with risk-masking incentives. Such window dressing is also more pronounced at banks with greater management compensation sensitivity to ROA and at banks that borrow in private debt markets, consistent with contractual incentives. Finally, we document a negative equity market reaction to the release of regulatory filings that indicate unexpected downward window dressing of shortterm borrowings. The potential implications of our findings go beyond bank holding companies and the financial industry, and bear relevance to recent SEC deliberations regarding short-term borrowing disclosure regulation.

JEL Classification: G14; G21; G28; M40 Keywords: Window dressing; Short-term borrowing; Sale and repurchase agreement; Bank holding companies; Disclosure regulation

1. Introduction The recent financial crisis brought into focus financial institutions' risk-taking behavior,

and raised concerns about whether their end-of-quarter balance sheets are accurate depictions of their risk levels during the quarter.1 Coincident with these concerns, the Securities and Exchange Commission (SEC) unanimously voted on September 17, 2010 to propose rules requiring both financial and non-financial public companies to provide enhanced disclosure of short-term borrowings such as repurchase agreements (repos), federal funds purchased, and commercial paper.2 The SEC is particularly concerned with disclosures related to short-term borrowings, as the levels of such borrowings can vary significantly during a reporting period, potentially making end-of-period balances less representative of risk exposure during the period. Moreover, the SEC points out that short-term borrowings form a critical component of firms' liquidity and capital resources, and that recent events have shown that such sources of funding can be severely affected by market illiquidity.

Even though the spotlight has been on the financial industry, similar issues can arise in other industries that rely on short-term financing arrangements to fund operations. In its proposed rule, the SEC suggests that the inherent riskiness associated with short-term borrowings may warrant enhanced disclosure of within-quarter exposure to these risks by all SEC registrants. Presently, only commercial banks and bank holding companies (BHCs) are required to disclose quarterly averages of certain financial variables, including a key source of their short-term funding - repurchase agreements and federal funds. Appendix A summarizes the

1 For example, a Wall Street Journal article on April 9, 2010 titled "Big banks mask risk levels" reports that during 2009 a group of 18 large banks in aggregate substantially lowered their quarter-end repo liabilities compared to the levels during the quarter. 2 SEC Release Nos. 33-9143 and 34-62932 (Sept. 17, 2010); File No. S7-22-10 (to be codified at 17 C.F.R Parts 229 and 249).

1

current disclosure requirements for BHCs from the Federal Reserve and the SEC as well as the SEC's proposed rule.

In this study, we investigate BHCs' downward window dressing of quarter-end short-term borrowings, where we define window dressing as a discretionary short-term deviation around quarter-end reporting dates of a financial variable from its quarterly average level.3 In so doing, this study provides the first empirical evidence on the window dressing of short-term borrowings and the stock market reaction to the public release of regulatory filings (i.e., Y-9C filings) from which window dressing may be detected. Even though our analysis is based on Y-9C filings by BHCs, the implications are broader and may extend to other industries.

Incentives for managers to downward window dress short-term borrowings can come from several sources. First and foremost, as pointed out by the SEC short-term borrowings such as repo liabilities are inherently risky. In particular, when market liquidity is low, firms that rely heavily on short-term borrowings are more susceptible to increases in borrowing rates or other unfavorable terms. Further, when markets are unstable it may be difficult to roll over short-term borrowings. Therefore, banks may have a direct incentive to decrease the reported quantity of such short-term borrowings in their financial statements, ceteris paribus. Managers may thus engage in downward window dressing in an attempt to mask the true risk level of the firm in hopes of obtaining higher valuations for the firms' securities and better terms with transaction counterparties. Second, regulatory capital requirements may provide both direct and indirect repo liability window dressing incentives. A direct incentive comes from the fact that the amount of margin ("haircut") in a repo transaction imposes additional risk-based capital requirements on the borrower. Therefore, banks may decrease their repo borrowings around quarter-end to directly

3 We describe our empirical measure of window dressing in detail in Section 4.1. As is standard in the literature, we refer to cases where the quarter-end value is less than (greater than) its quarterly average level as downward (upward) window dressing.

2

reduce required risk-based capital. Indirectly, banks with lower capital adequacy ratios appear riskier, and therefore may have added incentive to reduce their disclosed levels of risky shortterm borrowings. Finally, window dressing may arise from contractual incentives. By taking on additional borrowing during the quarter, a bank expands its balance sheet and the base from which earnings are produced. If managers and other employees are compensated based on earnings relative to the end-of-quarter asset base and risk levels, downward window dressing of short-term borrowings can boost their compensation relative to compensation that would be awarded in the absence of window dressing. Furthermore, participation as a borrower in private debt markets can provide an additional window dressing motive because of a desire to avoid financial covenant violations.

Using a sample of publicly traded BHCs, we find evidence of significant downward deviations in quarter-end short-term borrowings levels, in particular, repo and federal funds liability accounts, that appear material in a substantial fraction of firm-quarter observations, especially among the largest BHCs. Consistent with our predictions, we find that BHCs with a higher concentration of short-term borrowings in their liability structure (i.e., those that likely have greater incentives to mask their risk levels) have larger downward deviations in quarter-end repo and federal funds liabilities. In addition, we find evidence that the magnitude of these quarter-end downward deviations is greater for relatively large BHCs with lower regulatory capital adequacy ratios. We further document the magnitude of downward quarter-end deviations in short-term borrowings is larger for banks with greater sensitivity of CEO total compensation to return on assets, a performance measure typically used in compensation contracts that may be boosted by window dressing activities. Finally, we provide some evidence that such downward window dressing is more pronounced for firms that borrow in the private debt market.

3

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download