Presentation on theme: "Competition or Collaboration? The Reciprocity Effect in Loan Syndication Jian Cai Washington University in St. Louis The 45 th Annual Conference on Bank."— Presentation transcript:
Competition or Collaboration? The Reciprocity Effect in Loan Syndication Jian Cai Washington University in St. Louis The 45 th Annual Conference on Bank Structure and Competition May 6-8, 2009 Chicago
The Syndicated Loan Market Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 2 In the U.S. Increased 6 times (CAGR = 15%) Outside the U.S. Increased 30 times (CAGR = 28%) The syndicated loan market has experienced tremendous growth and become an increasingly important source of corporate finance since the early 1990s
An Inherent Agency Problem A syndicated loan is a credit facility two or more lending institutions jointly agree to provide to a borrowing firm Two types of syndicate members: lead arranger(s), participant Lenders Bank screening and monitoring is a key economic function of banks in relationship lending according to contemporary financial intermediation theories; these responsibilities are mainly delegated to lead arrangers Associated with loan syndication, we observe the following: “Informed” lead arrangers vs. “uninformed” participant lenders Costly but often unobservable due diligence and monitoring effort Diluted incentive for lead arrangers to monitor their borrowers Syndication introduces an agency problem – possibility of opportunistic behavior by lead arrangers Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 3
Research Question Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 4 However, there has been little empirical evidence of such opportunistic behavior A larger portion of “quality” loans are syndicated [Simons (1993)] Agency problems do not prevail in loan syndications [Panyagometh and Roberts (2002)] Default rates in the syndicated loan market are quite low [Sufi (2007)] Question: How does the syndicated loan market overcome the obvious agency problem?
Two Existing Explanations Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 5 1. The Incentive Effect The lead arranger retains a larger share of the loan that presents more severe info asymmetry and requires more intense monitoring and due diligence 2. The Reputation Effect Reputation concerns of the lead arranger mitigate the agency problem in loan syndication; a more reputable lender is more likely to syndicate loans, etc. Opaque Borrowers Less Reputable lead arrangers
Lead Arrangers = Participant Lenders? Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 6 In the U.S. market, 77% of lead arrangers also participate in loans The largest lead arrangers are indeed the largest participant lenders Top 10 Lead Arrangers: 1. JPMorganChase 2. Bank of America 3. Citigroup 4. Wachovia 5. Deutsche Bank 6. Credit Suisse 7. Wells Fargo 8. GE Capital 9. UBS 10. ABN AMRO The U.S. Syndicated Loan Market, 2004-2006 Top 10 Participant Lenders: 1. Bank of America 2. Wachovia 3. JPMorganChase 4. ABN AMRO 5. Wells Fargo 6. U.S. Bancorp 7. GE Capital 8. Citigroup 9. National City Corp. 10. Royal Bank of Scotland
Who Participate in Whose Loans? Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 7 Lenders often maintain stable relationships with certain other lenders and rotate their roles between leading and participating The U.S. Syndicated Loan Market, 2004-2006 JPMorganChase Bank of AmericaCitigroup 28% 39% 52% 30% 47% 17%
A Novel View: The Reciprocity Effect Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 8 Syndicate arrangements are reciprocal I show that such reciprocal arrangements serve as an effective mechanism to mitigate agency conflicts in loan syndication Reciprocal arrangements → Reciprocity shared among lead arrangers → The reciprocity effect The key to the cooperative equilibrium is the punishment/threat of not inviting lead arrangers whose loans previously failed Opportunistic Behavior Loan Default Unable to Participate Due Diligence & Monitoring Loan Default Able to Participate
Empirical Prediction Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 9 Reciprocal participation gives the lead arranger additional incentive to act in the interest of the entire syndicate and hence induces the lead arranger to exert the desired monitoring effort Prediction: The moral hazard problem is reduced among loans whose lead arrangers share reciprocity with one or more participant lenders; this reduced moral hazard results in: (i) a smaller share of the loan retained by the lead arranger, (ii) a lower interest rate charged to the borrower, and (iii) a lower probability of loan default.
Empirical Evidence Bank Structure Conference, May 2009 10 The Reciprocity Effect in Loan Syndication Controlling for borrower, lead arranger, and loan characteristics, I show that for loans with reciprocity: Lead arrangers retain on average 4.3% less of the loan The average interest spread over LIBOR on drawn funds is 11 basis points lower The default probability is 4.5% lower These results are both statistically and economically significant and robust to various specifications The reciprocity effect also exists for (i) informationally-opaque borrowers, (ii) smaller borrowers, (iii) smaller loans, (iv) less reputable lead arrangers, and (v) less reputable borrowers
Definition of Reciprocity Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 11 Focus here is on current reciprocity; results also hold for other forms Total reciprocity vs. reciprocity at origination (ex post vs. ex ante) Total reciprocity: reciprocity over the entire sample period Reciprocity at origination: reciprocity in existence at loan origination used in all empirical analyses for testing the reciprocity effect Loan A Bank A (Lead) Bank B (Participant) Loan B Bank B (Lead) Bank A (Participant) Reciprocity Current Reciprocity Past & Future Reciprocity
Measures of Reciprocity Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 12 MeasureDefinitionAt Origination * Existence1 if the loan shares reciprocity with another loan, and 0 otherwise 0.71 BreadthFraction of reciprocal participant lenders among all participant lenders 0.45 DepthAverage fraction of reciprocal loans taken by the lead arranger 0.085 LengthAverage fraction of the loan period overlapped with reciprocal loans 0.46 * All statistics are means and for current reciprocity at origination only. Data sources: DealScan, Compustat, and bankruptcies data Sample: 46,448 syndicated loan facilities originated for non-financial U.S. firms from 1992 to April 2007
Regression Specification Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 13 Lead share Interest spread Loan default Existence Breadth Depth Length Robust standard errors allowing for clustering within borrowers/leads/borrower-lead groups
14 Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication The Reciprocity Effect Lead Share Spread on Drawn Funds Loan Default Reciprocity Existence -4.31*** (1.008) -11.13*** (3.088) -0.045*** (0.014) Reciprocity Breadth -6.69*** (1.250) -16.47*** (3.784) -0.047*** (0.015) Reciprocity Depth -13.13*** (4.318) -23.40 (15.919) -0.197*** (0.060) Reciprocity Length -4.68*** (1.085) -15.86*** (3.799) -0.041*** (0.016) * significantly different from zero at the 10% level, ** at the 5% level, and *** at the 1% level.
Economic Significance Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 15 Average lead share = 29.5% Average loan amount = $217 million Average maturity = 50 months Average interest spread = 221 basis points Default rate (1992-2001) = 9.1% 4.3% less of the loan: A reduction of 15% $9 million savings per loan 11 basis points lower: A reduction of 5% $238,700 savings per year $994,583 savings in total 4.5% lower chance of default: A reduction by half
Conclusion Bank Structure Conference, May 2009 The Reciprocity Effect in Loan Syndication 16 Examined reciprocity in loan syndication and its effect in resolving agency conflicts between lead arrangers and participant lenders Uncovered strong and consistent empirical evidence that lead arrangers’ moral hazard is mitigated by the prevalent existence of reciprocity in syndicated loans important implications: To lending institutions, smaller shares retained of loans they lead, thus less capital tied to individual loans and better risk diversification To borrowing firms, lower borrowing costs, which may explain why the syndicated loan market has grown so fast in recent years To regulators, lower default rate, which may improve social welfare