The Impact of Credit Policy on Bank Loans to SMEs: Focusing on the Aggregate Credit Ceiling System of the BOK

Size: px

Start display at page:

Download "The Impact of Credit Policy on Bank Loans to SMEs: Focusing on the Aggregate Credit Ceiling System of the BOK"

Lenard Booker
6 years ago
Views:

1 BIS 9 th Annual Workshop of the ARN 22 nd March 2016 The Impact of Credit Policy on Bank Loans to SMEs: Focusing on the Aggregate Credit Ceiling System of the BOK Hosung Jung and Hosung Lim Economic Research Institute The Bank of Korea

2 Contents Motivation Contributions Research Questions Data & Summary Statistics Model Specification Empirical Results Robustness Tests Conclusions

3 Motivation Central banks of major countries have expanded the credit policy since the Global Financial Crisis(GFC). Prior to the GFC : Stabilizing prices and securing financial stability by altering the amount of liquidity through interest rate adjustments. After the GFC : Overcoming the financial market friction (i.e. credit crunch), CBs have conducted the credit policy and unconventional monetary policy so as to affect the flow of funds and credit allocation in the private sector. 3

4 Motivation The Bank of Korea (BOK) has made active use of the Aggregate Credit Ceiling System (ACCS) as one of its primary credit policy instruments since 2012 BOK sought to adjust the aggregate credit ceiling (ACC) flexibly to ease the effects of external and internal economic shocks to SME loans (Asian foreign currency crisis in 1997, September 11 in 2001, the global financial crisis in 2008). The Aggregate Credit Ceiling Loan (ACCL) is aimed at providing loans to financial institutions within a ceiling set by the Monetary Policy Committee, based on their particular SME lending performance. Established Special Support Ceiling linked to SME Credit Loans : April, 2012 Incorporated the special support limit for loans to self-employed business owners : October, 2012 Reformed the Aggregate Credit Ceiling System (ACCS) : April,

5 <Figure 1> Changes in Aggregate Credit Ceiling and Its Interest Rate Note: Shaded part is the period used for the analysis. Source: The Bank of Korea 5

6 <Figure 2> Trend of Bank Loans to SMEs Note: Shaded part is the period used for the analysis. Source: Financial Supervisory Service, Korea 6

7 <Table 1> Details of Aggregate Credit Ceiling Loan Programs (as of the end of June 2013) Entitled funds Time of Introduction Quota [Trillion Korean Won] Program for Trade Financing March Program for Credit Loans April Program for Small-Scale Sole Traders October Program for High Technology Start-ups April Program for Local SMEs March Source: The Bank of Korea 7

8 <Table 2> Bank Loans to SMEs and Aggregate Credit Ceiling Loans to Each Bank Minimum Median Unit: trillion Korean Won Average Maximum End of 2009 Total bank loans to SMEs Bank loans eligible for ACCL ACCL by the BOK End of 2010 Total bank loans to SMEs Bank loans eligible for ACCL ACCL by the BOK End of 2011 Total bank loans to SMEs Bank loans eligible for ACCL ACCL by the BOK Note: Calculated based on the individual institutional limit of seven major banks [Woori, Hana, KB, IBK, KEB, SC, and Shinhan] Source: The Bank of Korea 8

9 Motivation The credit policy of the central banks, which affects credit allocation to small and medium-sized enterprises (SMEs) and households, has received much attention since the global financial crisis. We need to examine how Aggregate Credit Ceiling System (ACCS), a major instrument of credit policy of the BOK, affects the credit availability of SMEs. 9

10 Contributions This research empirically analyses availability of credit for SMEs based on bank-firm level panel data, with a particular focus on credit policy. Its findings can serve as a useful reference for implementing credit policy, which is being increasingly adopted by central banks since the global financial crisis. 10

11 Research Questions Whether banks increase their lending to ACCS-eligible SMEs and reduce lending rates Examining how the ACCS helps improve banks intermediation function based on their loan data and the financial statements of banks and enterprises Whether there are different effects of ACCS on SMEs by firms characteristics focusing on the effects on firms credit rating and history 11

12 Research Questions Whether new eligibility into the ACCS and an increase in the ratio of existing ACCL ratio impact on the total loan amount and lending rate Examining advantages of new entry into the ACCS Whether new eligibility into the ACCS and an increase in the ratio of existing ACCL ratio increase or decrease in the number of transacting banks Examining advantages of new entry into the ACCS 12

13 Data loans from seven major banks (Woori, Hana, KB, IBK, KEB, SC, and Shinhan) of Korea to incorporated SMEs: Macroprudential Analysis Dept. (currently Financial Stability Department), The Bank of Korea Enterprises financial statement : Korea Enterprise Data Co. LTD Banks financial statement : FAIRS, The Bank of Korea Period/Frequency: 2009~2011, Annually 13

14 About the dataset (1) Banks (enterprises) do not determine the eligibility for providing (accessing) loans based on ACCS. As ACCL eligibility is determined for each loan type, the possibility of sample selection bias is very low. Thus, the samples used for the analysis are the entire sample of SME loan performance of individual banks. As ACCL eligibility is determined for each loan type, the possibility of sample selection bias is very low. 14

15 About the dataset (2) Using Bank-firm level loan data, We can identify differences on banks between ACCL eligible loans and non-accl eligible loans We also can identify which banks are offered to a company with a high/low percentage of ACCL eligible loan/interest rate That is, eligibility of ACCL is not only determined by enterprises characteristics, but also by relationship between banks and enterprises. 15

16 <Table 3> Number of SMEs and Loan Amount Unit : trillion Korean Won Year ACCL eligible SMEs ACCL ineligible SMEs Total SMEs 2) Number of enterprises Loan amount 3) Number of enterprises Loan amount Number of enterprises Loan amount End of , , , End of , , , End of , , , Notes: 1) Based on the balance of the loans from seven major banks (Woori, Hana, KB, IBK, KEB, SC, and Shinhan) in Korea to incorporated SMEs. SMEs are limited to enterprises having credit ratings and the financial variables necessary for the analysis. 2) SMEs are limited to incorporated SMEs and exclude large enterprises and self-employed business owners. 3) The loan amount includes domestic currency loans, foreign currency loans, advances for customers, and acceptances and guarantees outstanding. Source: The Bank of Korea 16

17 Summary Statistics Dependent variable : log value of the loan amount of individual enterprises and weighted- average loan interest rate The loan amounts : domestic currency loans, foreign currency loans, advances for customers, and acceptances and guarantees outstanding Loan interest rate : {[Loan amount per account of the borrower Loan interest rate per account] / Total loan amount of the borrower} Independent variable : ratio of ACCL Ratio of ACCL = ACCL/Loan amounts Control variables for loans : credit rating, dummy for relationship banking, ratio of mortgage loans Control variables for enterprises : business history, log value of assets, profitability (ROA), leverage ratio, coverage ratio, ratio of total loans to total assets Ratio of total loans to total assets : proxy for loan demand Control variables for individual bank : log value of assets, leverage ratio 17

18 <Figure 3> Bank Loan Distribution of SMEs 18

19 <Figure 4> Interest Rate Distribution of Bank Loans to SMEs 19

20 <Figure 5> Business History Distribution of SMEs 20

21 <Panel A: Loans> Name of variables Minimum P1 Median Average Standard D eviation P99 Maximum Ln(Loans) Loan interest rate[%] Ratio of mortgage loans <Panel B: Enterprises > <Table 4> Summary Statistics Name of variables Minimum P1 Median Average Standard D eviation P99 Maximum Ln( Assets) Business history Number of transacting banks Credit rating Profitability [ROA] Ratio of total loans to total assets Leverage ratio Coverage ratio <Panel C: Banks > Name of variables Minimum P1 Median Average Standard D eviation P99 Maximum Ln (Assets) Profitability [ROA] Leverage ratio Non-performing loan ratio Notes: 1) The loan amount includes domestic currency loans, foreign currency loans, advances for customers, and acceptances and guarantees outstanding. 2) Loan interest rate means weighted average interest calculated as {[Loan amount per account of the borrower Loan interest rate per account] / Total loan amount of the borrower}. 3) Mortgage loans include guaranteed loans. 4) Credit rating is classified from Grade 1 to Grade 10, and Grade 1 is the highest grade. 5) The ratio of total loans to total assets is defined as the ratio of total borrowings to total asset of the enterprise. 6) Enterprise leverage ratio is defined as the ratio of total assets to total capital. 7) Enterprise coverage ratio is defined as the ratio of operating income to interest expense. 8) Bank leverage ratio is defined as the ratio of total asset to total capital. 9) Bank non-performing loan ratio is defined as non-accrual loan ratio. Source: The Bank of Korea (for loan and bank level data), Korea Enterprise Data (for enterprise level data) 21

22 Model specification (1) (2) (3) (4) (5) 22

23 Model specification Panel analysis Adopt Hausman Test to decide whether to use a fixed effect or a random effect in the Panel analysis In addition, it seems that a fixed effect model, which assumes that unobserved individual heterogeneity and other explanatory variables are related, is more realistic A random effect model is also used based on the argument that the model would be more efficient when unobserved individual heterogeneity and other explanatory variables are independent, and in case that panel data have short time series It would be better using random effect model for robustness check 23

24 Empirical Results The effects of ACCL Eligible Loans on Bank Loans to SMEs Focusing on credit rating and business history for enterprises 24

25 <Table 6> The Effects of ACCL Eligible Loans on Bank Loans to SMEs Panel A: Focusing on Credit Rating 25

26 <Table 6> The Effects of ACCL Eligible Loans on Bank Loans to SMEs Panel B: Focusing on Business History 26

27 Empirical Results The Effects of ACCL Eligible Loan on the Interest Rate of Bank loans to SMEs Focusing on credit rating and business history for enterprises 27

28 <Table 7> The Effects of ACCL Eligible Loans on the Interest Rate of Bank Loans to SMEs Panel A: Focusing on Credit Rating 28

29 <Table 7> The Effects of ACCL Eligible Loans on the Interest Rate of Bank Loans to Panel B: Focusing on Business History SMEs 29

30 <Table 8> Transition Matrix between ACCL Eligible and Ineligible Enterprises <Panel A>Transition Matrix Between ACCL Eligible and Ineligible Enterprises by Bank ACCL ineligible enterprises t ACCL eligible enterprises t ACCL ineligible enterprises t % 9.24% ACCL eligible enterprises t % 25.06% <Panel B> Transition Matrix between an Increase in the Ratio of the ACCL Eligible Loans and Decrease of Those by Banks Decrease of ACCL ratio t Increase of ACCL ratio t Decrease of ACCL ratio t % 9.43% Increase of ACCL ratio t % 28.87% 30

31 Empirical Results The Effects of New Eligibility or an Increase in the Ratio of ACCL on Bank Loans and Interest Rates to SMEs Focusing on bank loans and interest rates to SMEs 31

32 <Table 9> The Effects of New Eligibility or an Increase in the ratio of ACCL 32

33 Empirical Results The Effects of New Eligibility or an Increase in the Ratio of ACCL on an Increase or Decrease in Transacting Banks 33

34 <Table 10> The Effects of New Eligibility or an Increase in the Ratio of ACCL on an Increase or Decrease in Transacting Banks 34

35 Robustness Tests 1 The Effect of ACCL Eligible Loan on Bank Loans to SMEs: Bank Loans Excluding the BOK s Support from the Total Loan The loan amount excluding the BOK s support = [Total loans (ACCL eligible loan x BOK s support ratio)] Reflecting that there is a considerable difference between the ACCL for general fund and that for regional SMEs in terms of support ratio, different ratios are respectively applied to general fund loans (Corporate Procurement Loans, Electronically Secured Receivables Loans, and trade finance) and regional SME supporting funds. 35

36 <Table 11> The Effects of ACCL Eligible Loan on Bank Loans to SME :Bank Loans Excluding the BOK s Support from the Total Loan Amount Panel A: Focusing on Credit Rating 36

37 <Table 11> The Effects of ACCL Eligible Loan on Bank Loans to SME :Bank Loans Excluding the BOK s Support from the Total Loan Amount Panel A: Focusing on Business History 37

38 Robustness Tests 2 The Effect of ACCL Eligible Loan on Bank Loans and Interest Rate to SMEs (Random Effects) Industry dummy and bank dummy are also used 38

39 <Table 12> The Effects of ACCL Eligible Loan on Bank Loans and Interest Rate to SMEs : Random Effects 39

40 Conclusions ACCS helps increase bank loans to SMEs and decrease lending rates. These effects are particularly prominent for enterprises with medium-credit scores, startups, and enterprises newly eligible for the ACCS. Enterprises tend to access loans from a greater number of banks after being included in the system It does influence banks decisions by reducing information asymmetry between banks and enterprises 40

Warwick Business School. ABFER Specialty Conference on Financial Regulations: Intermediation, Stability and Productivity, January 2017

Warwick Business School. ABFER Specialty Conference on Financial Regulations: Intermediation, Stability and Productivity, January 2017 ABFER Specialty Conference on Financial Regulations: Intermediation, Stability and Productivity, January 2017 Summary Objective: Examining the role of macroprudential policies to contain crossborder bank