Impact of Government InterventionOn Banks’ Loan Provision and Quality

Impact of Government InterventionOn Banks’ Loan Provision and Quality

Author: Putu Nila Kusuma Devi

[email protected] 

Coach: Teng Wang [email protected] Co-reader: Dr. Ben Wempe

[email protected] 

Agenda

• Introduction & Research Questions• Literature Review• Research Methodology• Research Findings• Conclusion • Limitation

Introduction (1)

•Contagion effect of bank runs

•The rise of the subprime mortgage

•Interest rate rose from 1% - 5.25% during 2003 to 2006

•Failing big investment banks

•Fallen of the U.S. stock market

Introduction (2)• Government interventions:

– Reduced fed rate 5.25% to 0.25%– October 2008, the government launched

Troubled Assets Relief Program (TARP)Capital Purchase Program (CPP).

• CPP– Purpose: Increase confidence & lending– $205 billion to 709 firms

Research Questions

• “How does the impact of CPP on banks’ loan provision (loan growth and credit growth)?”

• “How does the impact of CPP on banks’ loan quality (delta non-performing loans ratio and delta loan-loss reserves ratio)?”

Literature Review (1)

• Bailout– Critics:

• Stiglitz (2010) : banks bailout has almost no multiplier, need of regulation

• Lewis (2010): the incentives on Wall Street are all wrong

– Pro:• Paulson (2010): systemic risk

Literature Review (2)

• Banks’ Loan Provision– Li (2011) : TARP banks significantly increase

their loan supply by 6.43% of total assets annually.

– Bayazitova & Shivdasani (2010) find that TARP recipient banks increased their outstanding loans by approximately 6.2%

Literature Review (3)

• Banks’ Loan Provision– Taliaferro (2009) shows limited effect of the

CPP on participants’ lending compared to non-participants.

– Duchin and Sosyura (2010) find no evidence of greater credit origination by TARP participants

Literature Review (4)

• Banks’ Loan Quality– Li (2011) finds there is no difference in the

loan quality between TARP banks and non-TARP banks.

– Duchin & Sosyura (2010) find CPP recipients give riskier loans and invest in risky securities

Research Methodology (1)

• Research Subject: – bank holding companies– publicly traded– total consolidated assets of $500 million– CPP and Non-CPP banks

Research Methodology (2)

• Data Analysis Methods– Independent Samples T-test– Hausman Test of Endogeneity– Ordinary Least Squares (OLS)

Y = β0 + β1XC + β2XE + β3CPP + u

Findings (1)

– Generally, CPP banks performed worse in the two

and a half years horizon after receiving CPP infusion.– Non-CPP banks have better financial strength

  CPP Non-CPP All Sig.(2-tailed)

Total asset 11063098 9022150 10171884 0.666

Loaan growth -0.014 0.012 -0.003 0.37Credit growth -0.021 0.021 -0.003 0.281Non-performing loans ratio

0.027 0.019 0.024 0.027**

Loan-loss reserves ratio 0.011 0.008 0.01 0.056*Tier 1 ratio 10.147 12.105 11.002 0.000***Cash to asset 0.024 0.027 0.025 0.017**Price to book 1.167 1.369 1.255 0.006***Uninsured deposit 0.186 0.161 0.175 0.004***No. of banks 169 131 300  

• Independent Sample T-Test

Findings (2)

• With a negative GDP growth and a high unemployment rate, it is clear that funds are needed to enable the society to boost the economy.

  Log population GDP growth Unemployment

Minimum 5.7931 -2.0000 3.2000

Maximum 7.5627 7.3000 8.4000

Mean 6.904780 .681333 5.811667

Std. Deviation .3693710 1.1320143 1.1077851

Findings (3)

Dependent Variable Probability ValueLoan Growth 0.387Credit Growth 0.2236Non-Performing Loans Ratio 0.158Loan-Loss Reserves Ratio 0.5058

• The Hausman test results show evidence of no endogeneity from CPP to the dependent variable for all models.

Findings (4) Loan Growth Credit Growth

Independent Variable Coefficient Prob. Coefficient Prob.

Tier 1 ratio0.002 0.531 0.005 0.264

Troubled asset -0.454 0.000*** -0.588 0.000***Age 0.000 0.062* 0.000 0.182ROA-ytd 0.844 0.196 0.909 0.184Cash to asset 0.878 0.104 0.788 0.280Loans to deposit 0.012 0.066* 0.024 0.028**Log asset -0.025 0.165 -0.052 0.083*Delta Tier 1 0.007 0.058* 0.004 0.380Price to book 0.038 0.009*** 0.058 0.012**Uninsured deposit 0.058 0.623 0.005 0.977Log population 0.029 0.192 0.040 0.209GDP growth -0.018 0.079* -0.038 0.058*Unemployment -0.032 0.001*** -0.055 0.003***CPP 0.010 0.595 0.026 0.405C 0.111 0.581 0.300 0.321R2

0.335 0.279

Number of observation 298 299

Jarque-Bera 110510.3 0.000*** 83469.09 0.000***

White Test F Statistic 1.807 0.000*** 1.817 0.000***

Findings (5)Non-performing Loans Loan-loss Reserves

Independent Variable Coefficient Prob. Coefficient Prob.

Tier 1 ratio-0.001 0.070* 0.000 0.288

Troubled asset -0.018 0.187 0.009 0.045**Age 0.000 0.039** 0.000 0.419ROA-ytd -0.065 0.608 -0.061 0.058*Cash to asset -0.053 0.711 -0.078 0.021**Loans to deposit 0.001 0.841 -0.001 0.082*Log asset -0.001 0.693 0.002 0.031**Delta Tier 1 -0.001 0.296 0.000 0.618Price to book -0.003 0.353 -0.001 0.536Uninsured deposit 0.089 0.000*** 0.027 0.003***Log population -0.003 0.538 0.000 0.936GDP growth -0.004 0.076* -0.002 0.079*Unemployment -0.001 0.729 -0.001 0.244CPP 0.003 0.463 0.000 0.843C 0.069 0.102 0.005 0.746R2 0.122 0.149

Number of observation 300 300

Jarque-Bera 901.071 0.000*** 603.952 0.000***

White Test F Statistic 0.852 0.826 1.248 0.090*

Conclusion

• The CPP was given to banks with less financial strength compared to non-recipients.

• CPP has no impact on loan provision • CPP has no impact on loan quality

Limitation

• Absence of sufficient data• Cross-section data, relatively small sample

size.• Short time horizon

Thank You

Findings (4)

• OLS on Loan Provision– CPP has no significant influence in both loan

growth and credit growth. – Banks with troubled assets will have

difficulties in increasing loan provision– Banks with less likelihood of financial distress

have better ability to provide loans– GDP growth & unemployment rate influence

loan provision negatively

Findings (5)

• OLS on Loan Quality– CPP has no significant influence in both non-

performing loans ratio and loan-loss reserves ratio.

– Uninsured deposit decreases loan quality significantly.

– GDP growth positively affects loan quality