Abstract
We investigate the effect of the Troubled Asset Relief Program (TARP) on the shares of lead banks in syndicated loans by using quarterly data for the period from 2008 to 2011. We find that TARP capital injections to lead banks have a significantly positive effect on their shares of syndicated loans. Specifically, we estimate that the effect of TARP on the lending of lead banks increases their shares by 6.1% that is equivalent to an additional $50.37 million for an average syndicated loan of $840 million. The significantly positive effect is stronger for a lead bank that is larger, has a greater appetite for risk, and is not under supervisory stress. Overall, our results show that TARP was an effective mechanism for promoting the lending of lead banks. The larger fractional share of lead banks in syndicated loan packages signals both the loan quality and their commitment to mitigating the adverse selection and moral hazard problems. This study sheds new light on the mixed evidence of TARP on lending and contributes to the research and policy debates on the role of stringent supervisory scrutiny in lending.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Participants include banks and institutional investors including sponsors of structured products (Ivashina and Scharfstein 2010). Our sample shows an average loan amount of $840 million and an average 13 lenders take a portion of a syndicated loan. Among the 13 lenders, four are banks that are relatively large with average total assets of $578 billion.
In addition to the typical agency problems in a lending relationship that exist between the borrower and the lender, moral hazard takes place in the syndicated loan market between the lead arranger and the other syndicate members within the syndicate. See Dennis and Mullineaux (2000) for more details about the scope of agency problems associated with syndication.
Note Ivashina (2009) points out that the lead bank is uniquely exposed to idiosyncratic risk, and thus its portfolio is not perfectly diversified.
Adverse selection and moral hazard occur if the lead arranger chooses to syndicate the most risky loan and loosely monitors the borrower after the loan is originated because of their limited stake, respectively.
Similarly, Holmstrom (1982) highlights that either monitoring or rewards that are consistent with responsibilities can promote team collaborations.
FDIC law 32.3 (lending limits) indicates that the maximum size of a loan to a single borrower is 15% of the bank’s capital and surplus. Specifically, “Loan Concentration Limit” mandates that banks limit their exposure from each deal. The loan syndication facilitates portfolio diversification and limits the individual lender’s exposure to the overall credit risk that regulators would always recommend.
Cornett et al. (2013) argue that the bailout funding worked as a relatively cheap resource for its recipient banks and find that many relatively weak banks used TARP funding to make low quality loans.
They employ within-loan estimations to assess the impact of bank capital on lending in the syndicated loan market and find that the capital levels of banks significantly increase the bank allocation shares to the same syndicated loan. To mitigate the endogeneity concerns, the authors use TARP as a quasi-natural experiment and present evidence on the causal effect of bank capital on lending.
The DID analysis assumes a uniform date for the post-TARP dummy, while the actual TARP enrollment dates vary across TARP recipients. We follow the same approach as in Black and Hazelwood (2013) to identify our TARP variables that allows heterogeneity in TARP enrollment dates.
Ross (2010) finds favorable market reactions to the announcement of a dominant bank’s involvement in the syndicated loan. Similarly, Gopalan et al. (2011) study the behaviors of lending partners in the case of borrowers’ bankruptcies. Further, the complex nature of loan syndication requires lenders to be experts in the borrower’s industry (Francois and Missonier-Piera 2007).
The literature uses either the BHC level (Berger and Roman 2015) or commercial bank level (Li 2013) as the unit of measurement. The use of financial information at the consolidated level is consistent with the practice of bank supervision where the bank’s risk exposure is aggregated at the BHC level.
Using TARP funding as a natural experiment to address endogeneity of bank capital, Chu et al. (2019) also perform baseline within-loan analysis to examine the TARP effect on overall bank shares and the shares of participant banks. When examining overall bank shares, their within-loan estimations exclude single-lender loans in the sample that 50% of lead bank observations is missing. When examining participant bank shares, their subsampling designs are not superior than ours. A split sample analysis allows for all the covariates to differ and the coefficients to vary, unlike our analysis using an interaction term to allow only the interacted variable to have different coefficients for the groups by lender type.
Our sample includes 1,424 unique borrowers and 131 unique bank lenders. The clustering choice at the bank level or borrower level accounts for the correlation between multiple loans made by the same bank or borrowed by the same borrowers, respectively. Our results remain the same if we cluster the standard errors at the bank level.
We also use various TARP infusion rates by replacing the denominator with either Tier 1 regulatory capital, total regulatory capital, or total equity, all of which show consistent results.
To provide a perspective of economic magnitude, for an average bank with average total assets of $578 billion in our sample, the TARP injection equals about $1.69 billion ($578.06*0.732 (RWA share) *0.4% (TARP infusion rate)).
They argue all three channels: shocks to borrowers, shocks to bank capital, and variation in investor sentiment will require lead banks to hold larger shares of the loans originated in response to an economic downturn, thus amplifying credit cycles.
Their opposite results are based on within-loan estimations without borrower and loan characteristics, using a smaller and only including participant bank sample. Compared to their sample size of 1,533 observations; 322 for lead banks and 1,201 for participant banks, we use a relatively larger sample of 5,325 observations: 639 and 4,686 for lead banks and participant banks, respectively. Also, our finding of reduced shares of participant banks is in line with our hypothesis because of their fewer responsibilities and no incentive to monitor their borrowers than their peer lead banks in a syndicate. Our unreported analysis also shows a greater number of participant lenders when a TARP recipient participates in the syndicate. These findings imply increased risk-sharing among syndicated members when participant banks received TARP.
In banking, an NPL is a loan after being in default in 90 days. In general, a bank with low level of impaired loans indicates its higher ability to increase lending, or greater risk appetite. Although one may argue that high level of NPLs could in principle to follow a logic of gamble for resurrection, thus inducing banks to lend more, there is scant empirical study in this regard (Accornero et al. 2017).
Starting in May 2009, the banks with total assets greater than $100 billion are assessed in greater detail for their capital needs from a forward-looking perspective under the SCAP. According to Bayazitova and Shivdasani (2012), banks under SCAP face more rigorous regulatory surveillance to certify their well-being to the investing public in order to raise the necessary funding.
Using an alternative measure of size, the sample median of a bank’s total deposits, we find the same results at significance level of 1%.
See Acharya et al. (2018) for a detailed discussion of the potential impacts of stress tests on lending. Consistent with our findings, they find that lead banks subject to stress tests significantly decrease lending, particularly to riskier borrowers, to reduce their credit risk.
To save space, we report only the TARP recipient dummy results, as the TARP infusion rate results are qualitatively similar.
First, Fed director is a dummy variable equal to one if one of the bank's directors was on the board of directors of one of the 12 Federal Reserve Banks or a branch of the FRB in 2008 or 2009. Second, Subcomm_on FI is a dummy variable equal to one if the bank is headquartered in a district of a US House member who served on the Subcommittee on Financial Institutions and Consumer Credit in 2008 or 2009. Third, Democracy is a dummy variable equal to one if the bank's local representative was a democrat in the 2007–2008 election cycle. Last, Local FIRE Donation is the percentage of campaign contributions from local fire industries in total contributions received by a representative in the 2007–2008 election cycle.
Valid instruments should be correlated to the TARP recipient dummy but uncorrelated to the bank share and omitted variables. The research on TARP finds that these political and regulatory connections as instruments used in our model can affect the bank’s probability of receiving TARP funds (Blau et al. 2013). Because Duchin and Sosyura (2014) argue that macro-political representations follow election cycles with high turnovers, it is reasonable to conjecture that these instrument variables are not directly related to an individual bank’s share in the syndicated loan.
The instrument variables are significantly related to the TARP recipient dummy variable. The Chi-square tests indicate the hypothesis that coefficients on the four political and regulatory variables are jointly zero is rejected at the 1% level of significance that confirms the strength of the instrument.
The selection bias includes unobservable and nonrandom TARP applicants (i.e., banks choosing whether to apply and the Treasury choosing whether to invest).
The results for the analyses can be made available by request.
There are eight involuntary participants: Citi group, JP Morgan, Wells Fargo, Morgan Stanley, Goldman Sachs, Bank of New York, Bank of America, and State Street Bank. As an immediate action to stabilize the financial system, TARP mandated that these banks receive funding without their initiatives.
We consider that systemically important banks have a higher chance for bailouts because of the too-big-to-fail factor. We consider that relatively healthy and undercapitalized banks are the banks that were targeted by TARP. To balance the sample, for example, we drop treated TARP recipients with total assets larger than the maximum value of the control group, and control group observations with total assets below the minimum value of the treated group. Similarly, we repeat the same trimming method for the bank’s total deposits and ROA. We change the Tier 1 capital thresholds to the maximum value of the treated TARP group, and the minimum value of the control group to consider of TARP targeting undercapitalized banks.
References
Accornero M, Carpinelli L, Sorrentino AM (2017) Non-performing loans and the supply of bank credit: evidence from Italy. Bank Italy Occas Pap No 374. https://doi.org/10.2139/ssrn.2954995
Acharya VV, Berger AN, Roman RA (2018) Lending implications of U.S. bank stress tests: costs or benefits? J FinancIntermed 34:58–90. https://doi.org/10.1016/j.jfi.2018.01.004
Afonso G, Santos J, Traina J (2014) Do too-big-to-fail banks take on more risk? Fed Res Bank New York Econ Policy Rev 9:7–26
Bassett W, Demiralp S, Lloyd N (2020) Government support of banks and bank lending. J Bank Financ. https://doi.org/10.1016/j.jbankfin.2017.07.010
Bayazitova D, Shivdasani A (2012) Assessing TARP. Rev Financ Stud 25:377–407
Berger AN, Black LK, Bouwman CHS, Dlugosz J (2017) Bank loan supply responses to federal reserve emergency liquidity facilities. J FinancIntermed 32:1–15. https://doi.org/10.1016/j.jfi.2017.02.002
Berger AN, Makaew T, Roman RA (2019) Do business borrowers benefit from bank bailouts?: the effects of TARP on loan contract terms. FinancManag 48:575–639. https://doi.org/10.1111/fima.12222
Berger AN, Roman RA (2015) Did TARP banks get competitive advantages? J Financ Quant Anal 50:1199–1236
Black LK, Hazelwood LN (2013) The effect of TARP on bank risk-taking. J Financ Stab 9:790–803
Blau BM, Brough TJ, Thomas DW (2013) Corporate lobbying, political connections, and the bailout of banks. J Bank Financ 37:3007–3017
Calomiris CW, Khan U (2015) An assessment of TARP assistance to financial institutions. J Econ Perspect 29:53–80
Chava S, Roberts MR (2008) How does financing impact investment? the role of debt covenants. J Finance 63:2085–2121
Chavaz M, Rose AK (2019) Political borders and bank lending in post-crisis America. Rev Financ 23:935–959. https://doi.org/10.1093/rof/rfy027
Chu Y, Zhang D, Zhao Y (2019) Bank capital and lending: evidence from syndicated loans. J Financ Quant Anal 54:667–694. https://doi.org/10.1017/S0022109018000698
Cornett MM, Li L, Tehranian H (2013) The performance of banks around the receipt and repayment of TARP funds: over-achievers versus under-achievers. J Bank Financ 37:730–746
Dennis SA, Mullineaux DJ (2000) Syndicated loans. J FinancIntermed 9:404–426. https://doi.org/10.1006/jfin.2000.0298
Duchin R, Sosyura D (2014) Safer ratios, riskier portfolios: banks’ response to government aid. J Financ Econ 113:1–28
Egly PV, Mollick AV (2013) Did the U.S. treasury’s capital purchase program (CPP) help bank lending and business activity? Rev Quant FinancAcc 40:747–775. https://doi.org/10.1007/s11156-012-0297-9
El Ghoul S, Guedhami O, Pittman JA, Rizeanu S (2016) Cross-country evidence on the importance of auditor choice to corporate debt maturity. ContempAcc Res 33:718–751
Ertan A (2017) Real earnings management through syndicated lending. SSRN Electron J. https://doi.org/10.2139/ssrn.2851402
Feltham GA, Hughes JS, Simunic DA (1991) Empirical assessment of the impact of auditor quality on the valuation of new issues. J Acc Econ 14:375–399
Francois P, Missonier-Piera F (2007) The agency structure of loan syndicates. Financ Rev 42:227–245
Gopalan R, Nanda V, Yerramilli V (2011) Does poor performance damage the reputation of financial intermediaries? evidence from the loan syndication market. J Financ 66:2083–2120
Harris O, Huerta D, Ngo T (2013) The impact of TARP on bank efficiency. J IntFinanc Mark Inst Money 24:85–104
Holmstrom B (1982) Moral hazard in teams. Bell J Econ 13:324–340
Ivashina BV, Scharfstein D (2010) Loan syndication and credit cycles. Am Econ Rev 100:57–61
Ivashina V (2009) Asymmetric information effects on loan spreads. J Financ Econ 92:300–319. https://doi.org/10.1016/j.jfineco.2008.06.003
Kim D, Sohn W (2017) The effect of bank capital on lending: does liquidity matter? J Bank Financ 77:95–107. https://doi.org/10.1016/j.jbankfin.2017.01.011
Leland HE, Pyle DH (1977) Informational asymmetries, financial Structure, and financial intermediation. J Financ 32:371–387
Li L (2013) TARP funds distribution and bank loan supply. J Bank Financ 37:4777–4792. https://doi.org/10.1016/j.jbankfin.2013.08.009
Lin C, Ma Y, Malatesta P, Xuan Y (2012) Corporate ownership structure and bank loan syndicate structure. J Financ Econ 104:1–22
Lin Y, Liu X, Srinivasan A (2017) Unintended consequences of government bailouts: evidence from bank-dependent borrowers of large banks. SSRN Electron J. https://doi.org/10.2139/ssrn.3062484
Nguyen TVH, Ahmed S, Chevapatrakul T, Onali E (2020) Do stress tests affect bank liquidity creation? J Corp Financ 64:101622. https://doi.org/10.1016/j.jcorpfin.2020.101622
Puddu S, Waelchli A, Walchli A (2015) TARP effect on bank lending behaviour: Evidence from the last financial crisis. IRENE working paepr. http://hdl.handle.net/10419/191477
Rosenbaum P, Rubin D (1983) The central role of the propensity score in observational studies for causal effects. Biometrika 70:41–55
Ross DG (2010) The “Dominant Bank Effect:” how high lender reputation affects the information content and terms of bank loans. Rev Financ Stud 23:2730–2756
Sufi A (2007) Information asymmetry and financing arrangements: Evidence from syndicated loans. J Finance 62:629–668
Taliaferro R (2009) How do banks use bailout money? optimal capital structure, new equity, and the TARP. SSRN Electron J. https://doi.org/10.2139/ssrn.1481256
Thornton J, di Tommaso C (2019) Liquidity and capital in bank lending: evidence from European banks. Financ Res Lett 34:101273. https://doi.org/10.1016/j.frl.2019.08.021
Wooldridge JM (2002) Econometric analysis of cross section and panel data. MIT Press
Wu D (2015) The effects of government capital and liquidity support programs on bank lending: evidence from the syndicated corporate credit market. J Financ Stab 21:13–25. https://doi.org/10.1016/j.jfs.2015.09.002
Yingling EL (2008) Testimony of Edward Yingling, President and Chief Executive Officer, American Bankers Association on behalf of the ABA, House of Representatives Committee on Financial Services Hearing on “OVERSIGHT OF IMPLEMENTATION OF THE EMERGENCY ECONOMIC STABILIZATION ACT OF 2008 AND OF GOVERNMENT LENDING AND INSURANCE FACILITIES; IMPACT ON ECONOMY AND CREDIT AVAILABILITY”, Nov 18, 2008
Acknowledgements
The paper was a part of doctoral dissertation for Bolortuya Enkhtaivan. Therefore, authors greatly appreciate valuable advice and support from the late Dr. Siddharth Shankar who was an advisor, and Drs. Anand Jha and George Clarke who were the members of the Dissertation Committee. Also, we thank Dr. Lei Li for the instrumental variables, Dr. Lamont Black and participants of the 2015 Midwest Finance Association and 2016 Financial Management Association meetings, and anonymous reviewers for their valuable feedback and suggestions. We also thank Mr. Jonathan Moore for copyediting our manuscript.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Variable name | Variable definition |
|---|---|
Dependent variables (Source : Dealscan) | |
Bank Share | The weighted average of the bank's contribution to the loan facilities |
TARP variables (Source: US Treasury - CPP reports) | |
TARP dummy | =1 if a bank is a TARP recipient at the time of loan issuance, and 0 otherwise |
TARP infusion rate | Total TARP investment as percent of RWAs |
Bank characteristics (Source: Call Reports) | |
Lead | =1 if the bank is a lead arranger, and 0 otherwise |
Size | Natural log of total assets in $billions |
Tier 1 capital ratio | Tier 1 capital/RWAs |
Deposits | Deposits/total assets |
Cash | Cash/total assets |
Loan allowance rate | Loan allowance/total assets |
ROA | Net income/total assets |
Liquidity | (Cash + available for sale securities)/total assets |
Leverage ratio | Tier 1 capital/total assets |
Industry experience | Individual bank’s total volume of deals in a borrower’s 3-digit SIC-code industry in the past 5 years/total volume of deals in the same industry in the past 5 years |
Past relationship with the borrower | The past relationship equals the total volume of deals with the same borrower in the past five years divided by the total volume of deals of the borrower with all banks in the past five years |
Top 10 bank | The top-10 bank equals one if the bank ranks in the top-10 banks in terms of total volume of deals in USD billions it has issued in the past 5 years, and 0 otherwise |
Borrower characteristics (Source : Compustat) | |
S&P Rating | =1 if the borrower is rated as long-term S&P credit rating in the previous quarter and 0 otherwise |
Tobin’s Q | Market value of total assets/book value of total assets |
R&D rate | Research and development expense/total assets |
Leverage | Debt/total assets |
Profitability | Operating income before depreciation/total assets |
Cash holding | Cash/total assets |
Size | Natural log of the total assets in $m |
Tangibility | Net property, plant, and equipment/total assets |
Cash flow volatility | Standard deviation in cash flows during the previous 4 quarters |
Loan characteristics (Source : DealScan) | |
Maturity | Natural log of loan maturity in days |
Size | Natural log of loan amount in $m in a package |
Spread | Natural log of spread in bps |
Revolver indicator | =1 if the loan is a revolver, and 0 otherwise |
Secured indicator | =1 if at least one loan in a package has collateral and 0 otherwise |
Loan purpose dummies | =1 for corporate purpose, =2 for working capital, =3 for takeover, =4 for debt repayment, =5 for acquisition, =6 for backup, =7 for LBO, =8 for recapitalization and =9 for others |
Refinancing | = 1 if the loan is refinanced and 0 otherwise |
Rights and permissions
About this article
Cite this article
Enkhtaivan, B., Lu, W. The effect of TARP on lending: Evidence from the lead bank’s share in syndicated loans. Rev Quant Finan Acc 57, 1169–1193 (2021). https://doi.org/10.1007/s11156-021-00974-5
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11156-021-00974-5