Skip to main content
SpringerLink
Log in

Quarter-end repo borrowing dynamics and bank risk opacity

  • Published:
Review of Accounting Studies Aims and scope Submit manuscript

Abstract

We investigate the extent to which banks’ quarter-end borrowings in the repurchase market deviate from within-quarter levels, and associated factors. Quarter-end repo liabilities are materially lower than within-quarter averages for a large fraction of sample banks. These deviations are more pronounced at banks with a higher concentration of repo borrowings in their liability structure and with larger absolute trading gains or losses. Furthermore, the association with trading activity is mitigated when banks are better capitalized. We also find that these deviations are associated with bank depositor and borrower behavior. Together, the evidence suggests that deviations reflect both active window dressing and passive customer-driven liquidity dynamics. We document that unexpected downward quarter-end deviations in repo liabilities are associated with adverse short-term capital market consequences. Over the long term, banks with more frequent downward quarter-end deviations exhibit higher credit risk, but we find mixed evidence for equity market valuation multiples.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Notes

  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. Coincident with these concerns, in 2010 the Securities and Exchange Commission (SEC) proposed rules requiring both financial and nonfinancial public companies to provide enhanced disclosure of short-term borrowings such as repurchase agreements and commercial paper (SEC Release Nos. 33-9143 and 34-62932).

  2. The repo market in the US went through major disruptions during the recent financial crisis. Gorton and Metrick (2012) report that repo haircuts increased from close to zero (e.g., a $100 loan is secured with $100 worth of securities) in early 2007 to nearly 50 % (e.g., a $100 loan requires $150 of collateral) in late 2008. Furthermore, at the height of the crisis, lenders refused to accept anything but the safest of collateral, causing segments of the repo market other than Treasuries to dry up.

  3. Appendix 2 summarizes the current disclosure requirements for bank holding companies from the Federal Reserve and the SEC.

  4. At the time of this writing, there has been no change in the rule’s “proposed” status.

  5. The Bank Holding Company Act of 1956 defines a bank holding company as any company (including a commercial bank) that has direct or indirect control of a commercial bank.

  6. The reporting size threshold was $150 million before 2006. Furthermore, only the top-tier bank within a bank holding company hierarchy is required to file Y-9C post-2006.

  7. Extant literature suggests that bank regulators and the SEC have neither devoted large amounts of resources to monitor window dressing revealed in bank regulatory filings nor imposed severe penalties when such activities are detected (e.g., Allen and Saunders 1992).

  8. http://www.ffiec.gov/nicpubweb/nicweb/nichome.aspx.

  9. For example, a Wall Street Journal article on April 9, 2010, titled “Big banks mask risk levels,” reports: “Financial firms use cash from (repo) loans to buy securities, then use the purchased securities as collateral for other loans, and buy more securities. The loans boost the firms’ trading power, or ‘leverage,’ allowing them to make big trades without putting up big money. This amplifies gains-and losses, which were disastrous in 2008.”

  10. Constrained by data availability from the Call Report, Allen and Saunders (1992) use the average over the last month of the calendar quarter as a proxy for its quarterly average level. In a robustness test, Allen and Saunders (1992) make a trend-cycle adjustment.

  11. We match Schedule HC-K items with their corresponding quarter-end values by following the “Line Item Instructions for Quarterly Averages: Schedule HC-K” in the Y-9C instructions file available at http://www.federalreserve.gov/reportforms/forms/FR_Y-9C20110331_i.pdf.

  12. Federal funds are unsecured loans among depository institutions of their excess reserve balances at Federal Reserve Banks and are referred to as federal funds purchased (sold) for the borrowing (lending) bank.

  13. In untabulated robustness tests, we confirm that all key inferences of the study are unaltered if we use the simple difference between the quarter-end value and the quarterly average (scaled by average total assets).

  14. To ensure that β 1 is not merely picking up a mechanical relation, in untabulated analysis we replace RepoToTotalLiab QA with RepoLiab (quarter-end repo liabilities), and inferences are not materially affected.

  15. We employ two-way clustered standard errors along the firm and calendar quarter-year dimensions in all regression analyses (Petersen 2009).

  16. This timing is further supported by documentation on the Fed’s National Information Center website, and by evidence in Badertscher et al. (2015). To the extent some Y-9C filings are made public before or after our estimated publication window, our ability to find announcement period stock reactions to our window-dressing measure is diminished.

  17. We consider six different measures of expected return in our abnormal return calculation: both value- and equally weighted market return, CRSP size decile return, expected return from both a value-weighted and equally weighted market model, and expected return from a Fama–French three-factor model (Fama and French 1993). Likewise, we test robustness of our findings to alternative 3- and 7-day announcement-window periods. Subsequently reported inferences are insensitive to these choices.

  18. We obtained the matching file at http://www.newyorkfed.org/research/banking_research/datasets.html.

  19. Y-9C data are from http://www.chicagofed.org/webpages/banking/financial_institution_reports/bhc_data.cfm. Banks may restate previously filed Y-9Cs, where the Federal Reserve replaces the original Y-9C with the revised Y-9C. Therefore a data entry in the dataset can reflect a restatement instead of the original submission. However, the likelihood that the repo liability quarter-end balance or quarterly average is revised for a given bank-quarter is small. Moreover, to the extent it occurs, it works against our finding significant market reactions around the initial public release date to the repo deviation measure.

  20. Data are available at ftp://ftp.bls.gov/pub/special.requests/cpi/cpiai.txt.

  21. The sum of banks across these subsamples exceeds the 573 distinct bank holding companies in our overall sample because we reclassify banks into the top 50 subsample each quarter. Inferences throughout are unaltered if we remove primary dealer observations from the sample (159 bank-year observations, where we obtain identities of primary dealers from http://www.ny.frb.org/markets/pridealers_current.html).

  22. To be well capitalized, an institution must maintain a tier 1 risk-based capital ratio of at least 6 %.

  23. Subsequently reported inferences are intact if we remove sample observations where federal funds are more prominent, e.g., accounting for more than 25 % of the sum of repo and fed funds liabilities.

  24. Holding all independent variables at their mean values, movement across the interquartile range in RepoToTotLiab QA results in a 230 % decrease, from 0.11 to −0.15 % of total assets, in RepoLiabDEV.

  25. Untabulated analyses confirm that both the drop in rates before quarter end and the increase in rates immediately after quarter end are statistically significant.

  26. Untabulated analyses reveal that the market response is significant only when the current quarter-end deviation is downward. Stated differently, if the current quarter-end deviation is upward, there is no significant market response to the change in quarter-end deviation.

  27. In a simple untabulated high-low test of difference using the Satterthwaite method for unequal variances, we find marginal significance (one-tailed) that mean AVG_PE is lower in sample observations with above median  %RepoLiabBigDownDev relative to below median (20.954 vs. 24.500; one-tailed p value 0.096). We likewise find that mean AVG_MTB is higher in sample observations with above median  %RepoLiabBigDownDev relative to below median (1.68 vs. 1.46; two-tailed p value < 0.001).

  28. In a simple untabulated high-low test of difference using the Satterthwaite method for unequal variances, we find that mean AVG_TermStruc is significantly higher in sample observations with above median  %RepoLiabDownDev relative to below median (0.514 vs. 0.448; two-tailed p value 0.073).

  29. We exclude First Union Corporation (ranked #4), National City Corporation (#10), and Commerce Bancorp (#23) because all three banks were acquired around 2008 and as a result did not make a 10-K filing for 2008. We also exclude Comerica Inc. (#19) and Unionbancal Corp. (#20) because both report relatively low repo balances in the Y-9C for Q1 2008 ($374,000 in the case of Comerica and $647 million for UnionBancal), which can make repo-related disclosures less of a concern for these banks. This leaves us with 20 banks for which we report hand-collected data in the table below. As shown in the table, all 20 banks report 2008 Q1 repo balance of more than $1 billion.

  30. A total of eight banks disclose repo (or repo and federal funds) average balances in their Q1 2008 10-Q filings. However, in four out of the eight cases, the 10-Q filing window overlaps with the Y-9C publication window. As noted in Sect. 2.3, the Y-9C filing deadline for the first three quarters is 40 days after quarter end. We use an estimated publication date of 42 days after quarter end and construct a five-day return window centered on day 42, i.e., [day 40, day 45]. Two banks’ 10-Q filings (JPMorgan Chase and PNC Financial) are on day 42, which is within our Y-9C filing window. Another two banks’ 10-Q filings are on day 39 (State Street Corporation and Zions Bancorp), for which the market reactions to the 10-Q filings overlap with those to the Y-9C.

References

  • Afonso, G., Kovner, A., & Schoar, A. (2011). Stressed not frozen: The fed funds market in the financial crisis. Journal of Finance, 66, 1109–1139.

    Article  Google Scholar 

  • Allen, L., Peristiani, S., & Saunders, A. (1989). Bank size, collateral, and net purchase behavior in the federal funds market: empirical evidence. Journal of Business, 62, 501–515.

    Article  Google Scholar 

  • Allen, L., & Saunders, A. (1992). Bank window dressing: Theory and evidence. Journal of Banking & Finance, 16, 585–623.

    Article  Google Scholar 

  • Arora, N., Richardson, S., & Tuna, I. (2014). Asset reliability and security prices: evidence from credit markets. Review of Accounting Studies, 19, 363–395.

    Article  Google Scholar 

  • Badertscher, B., Burks, J., & Easton, P. (2015). Day 30: The tacit quarterly information event in the banking industry. Working paper.

  • Barclay, M., & Smith, C. (1995). The maturity structure of corporate debt. Journal of Finance, 50, 609–631.

    Article  Google Scholar 

  • Beatty, A., Chamberlain, S., & Magliolo, J. (1995). Managing financial reports of commercial banks: the influence of taxes, regulatory capital, and earnings. Journal of Accounting Research, 33, 231–261.

    Article  Google Scholar 

  • Benmelech, E., & Dvir, E. (2013). Does short-term debt increase vulnerability to crisis? Evidence from the East Asian financial crisis. Journal of International Economics, 89, 485–494.

    Article  Google Scholar 

  • Board of Governors of the Federal Reserve System, 100th Annual Report. (2013). www.federalreserve.gov/publications/annual-report/files/2013-annual-report.pdf.

  • Brunnermeier, M., & Pedersen, L. (2009). Market liquidity and funding liquidity. Review of Financial Studies, 22, 2201–2238.

    Article  Google Scholar 

  • Bushman, R., & Williams, C. (2012). Accounting discretion, loan loss provisioning, and discipline of banks’ risk taking. Journal of Accounting and Economics, 54, 1–18.

    Article  Google Scholar 

  • Choudhry, M. (2010). The repo handbook (2nd ed.). Oxford: Butterworth-Heinemann.

    Google Scholar 

  • Dechow, P., & Shakespeare, C. (2009). Do managers time securitization transactions to obtain accounting benefits? The Accounting Review, 84, 99–132.

    Article  Google Scholar 

  • Duffie, D., & Lando, D. (2001). Term-structures of credit spreads with incomplete accounting information. Econometrica, 69, 633–664.

    Article  Google Scholar 

  • Fama, E., & French, K. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33, 3–56.

    Article  Google Scholar 

  • Feltham, G., & Xie, J. (1994). Performance measure congruity and diversity in multi-task principal/agent relations. The Accounting Review, 69, 429–453.

    Google Scholar 

  • Furfine, C. (2000). Interbank payments and the daily federal funds rate. Journal of Monetary Economics, 46, 535–553.

    Article  Google Scholar 

  • Gallemore, J. (2013). Does bank opacity enable regulatory forbearance? Working paper.

  • Gorton, G., & Metrick, A. (2012). Securitized banking and the run on repo. Journal of Financial Economics, 104(3), 425–451.

    Article  Google Scholar 

  • He, Z., & Xiong, W. (2012). Rollover risk and credit risk. Journal of Finance, 67, 391–429.

    Article  Google Scholar 

  • Hördahl, P., & King, M. (2008). Developments in repo markets during the financial turmoil. BIS quarterly review. Bank for International Settlements. December 2008, 37–53.

  • Huizinga, H., & Laeven, L. (2012). Bank valuation and accounting discretion during a financial crisis. Journal of Financial Economics, 106, 614–634.

    Article  Google Scholar 

  • Jorion, P., & Zhang, G. (2007). Good and bad credit contagion: Evidence from credit default swaps. Journal of Financial Economics, 84, 860–883.

    Article  Google Scholar 

  • Kotomin, V., & Winters, D. (2006). Quarter-end effects in banks: Preferred habitat or window dressing? Journal of Financial Services Research, 29, 61–82.

    Article  Google Scholar 

  • Krishnamurthy, A., Nagel, S., & Orlov, D. (2014). Sizing up repo. Journal of Finance, 69, 2381–2417.

    Article  Google Scholar 

  • Lakonishok, J., Shleifer, A., Thaler, R., & Vishny, R. (1991). Window dressing by pension fund managers. American Economic Review (Papers and Proceedings), 81, 227–231.

    Google Scholar 

  • Lee, L. (2012). Incentives to inflate reported cash from operations using classification and timing. The Accounting Review, 87, 1–33.

    Article  Google Scholar 

  • Messier, W., Glover, S., & Prawitt, D. (2012). Auditing and assurance services: A systematic approach (8th ed.). New York: McGraw-Hill.

    Google Scholar 

  • Ng, L., & Wang, Q. (2004). Institutional trading and the turn-of-the-year effect. Journal of Financial Economics, 74, 343–366.

    Article  Google Scholar 

  • Ohlson, J., & Gao, Z. (2006). Earnings, earnings growth and value. Foundation and Trends in Accounting, 1, 1–70.

    Article  Google Scholar 

  • Petersen, M. (2009). Estimating standard errors in finance panel data sets: Comparing approaches. Review of Financial Studies, 22, 435–480.

    Article  Google Scholar 

  • Roychowdhury, S. (2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42, 335–370.

    Article  Google Scholar 

  • Securities and Exchange Commission. (2010). Short-term borrowing disclosure. Release Nos. 33-9143; 34-62932; File No. S7-22-10.

  • Shleifer, A., & Vishny, R. (2010). Unstable banking. Journal of Financial Economics, 97, 306–318.

    Article  Google Scholar 

  • Stigum, M. (2007). Stigum’s money market. New York: McGraw-Hill.

    Google Scholar 

  • Valukas, A. (2010, March). Examiner of United States Bankruptcy Court Southern district of New York. “In re Lehman Brothers Holdings Inc., et al., Debtors.” Chapter 11 Case No. 08-13555 (JMP).

  • Whittington, R., & Pany, K. (2010). Principles of auditing & other assurance services (17th ed.). New York: McGraw-Hill.

    Google Scholar 

Download references

Acknowledgments

We thank Anne Beatty, Jose Berrospide, Robert Bushman, Elizabeth Chuk, Dan Collins, Anya Kleymenova (discussant), Anzhela Knyazeva, Phil Picariello, Scott Richardson (editor), Stephen Ryan, William Schwert, Robert Storch, Jerry Zimmerman, two anonymous reviewers, and workshop participants at George Washington University, London Business School, New York University, Rice University, University of Rochester, the University of Minnesota 2011 Empirical Research Conference, the Fifth Annual Toronto Accounting Research Conference, the 2012 Utah Winter Accounting Conference, and the 2014 Review of Accounting Studies Conference for helpful comments and suggestions.

Author information

Authors and Affiliations

  1. Goizueta Business School, Emory University, Atlanta, GA, USA

    Edward L. Owens

  2. Simon Business School, University of Rochester, Rochester, NY, USA

    Joanna Shuang Wu

Authors
  1. Edward L. Owens
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Joanna Shuang Wu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Joanna Shuang Wu.

Appendices

Appendix 1: An illustration of a repo transaction

The economic nature of a repo transaction is a lending arrangement. It entails an accounting liability for the borrower of cash. Securities serving as collateral are not removed from the borrower’s balance sheet. The transaction is called a reverse repo from the lender’s perspective.

Appendix 2: Disclosure requirements for within-quarter information

See Table 9.

Table 9 Disclosure requirements for within-quarter information on balance sheet liability accounts
Full size table

Appendix 3: Bank disclosure of repo liability average balances in SEC filings and earnings announcements

This analysis helps assess the possibility that repo average balances disclosed in the Y-9C Schedule HC-K are preempted by disclosures in the banks’ 10-Q/10-K filings, their earnings announcements, or both. This is important for the interpretation of our short-window market reaction tests. Because this analysis involves hand collection of information from firms’ 10-Q, 10-K, and earnings announcements from Factiva, we limit our analysis to the top 25 banks in our sample based on their December 31, 2007, total assets.Footnote 29 We examine the repo average balance disclosures from the banks’ 2008 10-K and first quarter 10-Q filings and the corresponding annual and quarterly earnings announcements and collect information to answer the following questions:

  1. 1.

    Does the bank disclose its repo liability (or repo and federal funds) average balance in its 10-Q filing for Q1 2008? If so, does the 10-Q filing window precede the [−2, +2] day return window around the estimated Y-9C publication date of day 42 after the quarter end?

  2. 2.

    Does the bank disclose its repo liability (or repo and federal funds) average balance in its earnings announcement for Q1 2008?

  3. 3.

    Does the bank disclose its repo liability (or repo and federal funds) average balance in its 10-K filing for 2008? And if so, does the 10-K filing window precede the [−2, +2] day return window around the estimated Y-9C publication date of day 47 after the 4th quarter end?

  4. 4.

    Does the bank disclose its repo liability (or repo and federal funds) average balance in its annual earnings announcement for 2008?

The following summarizes our findings:

For Q1 2008, five out of 20 banks (25 %) disclosed information on repo (or repo and federal funds) average balances in either their 10-Q filings or earnings announcements that appear before and does not overlap with the 5-day return window around the estimated Y-9C publication date. These cases are italicized in the Table 10.Footnote 30 Therefore, for 75 % (15) of the 20 banks, our Y-9C return window likely captures the first public disclosure of repo (or repo and federal funds) average balances for Q1 of 2008. For Q4 2008, all but three banks disclosed average repo (or repo and federal funds) average balances in their 10-K filings. The high frequency is not surprising because such disclosure is required by the SEC (see Appendix 2). However, for all 20 banks, the 10-K filing dates (which vary between 54 and 61 days after year-end) come after the estimated Y-9C publication date of 47 days after the 4th quarter end. We do find that two out of the 20 banks disclosed the average repo (or repo and federal funds) average balance in their annual earnings announcement, both of which preempted the Y-9C. These cases are again italicized in Table 10. Therefore, for 18 of the 20 banks (90 %), the Y-9C appears to be the first public disclosure of repo (or repo and federal funds) average balances for Q4 of 2008.

Table 10 Banks analyzed
Full size table

Appendix 4: Variable definitions

Italicized variable names beginning with “BH” in the descriptions refer to the mnemonic data identifiers of raw data items obtained from the Federal Reserve Bank Holding Company data set at http://www.chicagofed.org/webpages/banking/financial_institution_reports/bhc_data.cfm. Referenced quarterly “Schedules” are from Form Y-9C. Schedule HC is the “Consolidated Balance Sheet.” Schedule HC-B is “Securities.” Schedule HC-E is “Deposit Liabilities.” Schedule HC-K is “Quarterly Averages.” Schedule HC-R is “Regulatory Capital.” Schedule HI is the “Consolidated Income Statement.” Subscripts i, t, and y reference bank, quarter, and year, respectively (Table 11).

Table 11 Variable definitions
Full size table

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Owens, E.L., Wu, J.S. Quarter-end repo borrowing dynamics and bank risk opacity. Rev Account Stud 20, 1164–1209 (2015). https://doi.org/10.1007/s11142-015-9330-2

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11142-015-9330-2

Keywords

JEL Classification

Search

Navigation