ARE LOAN GUARANTEE PROGRAMS EFFECTIVE?

1
ARE LOAN GUARANTEE PROGRAMS EFFECTIVE?

• by
• Dale W Adams
• The Ohio State University

2
Overview

• Three types retail, portfolio, and wholesale
• Reasons for loan guarantee programs
• Why SMEs lack access to formal loans
• Evaluating guarantee programs
• The benefit-cost analysis
• Strengths and weakness of three types of programs
• Discussion Topics

3
Types of Programs

• SME Retail Lender Bank
• Retail Portfolio Wholesale

4
--Retail Type--

• Third party is heavily involved
• Large increase in transaction costs
• Sometimes duplicate loan analysis
• Effectiveness is difficult to measure
• Moral hazard adverse selection
• Opportunities for financial substitution in
  borrowers portfolio, in lenders portfolio,
  between lenders, between formal and informal

5
--Portfolio Type--

• Fewer transaction costs
• Third party less involved
• Some adverse selection problems
• Easier to measure effectiveness
• Substitution problems among lenders
• Heterogeneous loans
• Similar to wholesale loan guarantees

6
--Wholesale Type--

• Even less transaction costs
• Third party has little to do
• Appropriate when specialized SME lenders are
  involved
• Only useful when retail lenders have too few
  funds to lend to all of their creditworthy SMEs
• Not appropriate when borrower and lender are
  deposit-taking organizations

7
Program Justifications

• Stimulate lending to groups who lack access to
  formal loans SMEs
• Overcome imperfections in financial markets.
• Loan guarantee is a partial collateral substitute
  that reduces the lenders risk of not recovering
  the loan.
• Help lenders learn about new clients and market

8
Why do SMEs lack access to formal loans?

• Risk for lender lack acceptable collateral, high
  rate of failure among new SMEs, judicial system
  is defective, civil unrest and economic turmoil,
  etc.
• Transaction costs for lender and borrower SME
  doesnt know banks, lender doesnt know this
  segment of market, distance, paperwork,
  inappropriate lending technology, number of
  visits, etc.

9
Evaluating Guarantee Programs

• Only the number of loans guaranteed?
• Outreach
• Sustainability subsidy dependence
• Benefits and costs the additionality problem,
  the substitution problem, and the attribution
  problem

10
Costs of programs

• Costs are relatively easy to measure
• Set-up costs
• Initial funding and later refunding
• Opportunity costs of funding
• Programs cause transaction costs for lenders,
  for borrowers, the guarantee agency, and donor or
  government

11
Measuring Benefits

• Benefits are more difficult to measure than costs
• Number of loans guaranteed is not a reliable
  measure of benefits
• Measuring additionality is main problem
• If no additionality, no benefits from program
• Two additionality cases,
• A substitution case, and
• A case illustrating attribution problem

12
Case A

• Objective increase loans to SMEs
• Bank X made 100 loans to this group before
  program, total value US 100,000
• With guarantee program same bank made 100 new
  guaranteed loans 100 regular loans to SMEs,
  total value US 200,000
• 100 additionality in number and value

13
Case B

• Same objective as Case A
• Before program Bank X makes loans to 100 SMEs for
  total value of US 100,000.
• With guarantee program, Bank X shifts 50 of its
  riskiest SME loans to guarantee, and makes 50 SME
  loans without guarantee. No change in total
  value of SME lending.
• Zero additionality for number of loans and value

14
Case B continued

• Bank shifts the most risky SME borrowers to
  guarantee in order to capture risk subsidy
  (adverse selection).
• Most evaluations of loan guarantee programs
  ignore the additionality problem and assume that
  all loans guaranteed additional loans made
  because of the program. This results in
  substantial overestimates of the benefits of
  these programs.

15
Substitution ProblemsCase C

• Same objective as Cases A and B
• Bank X makes 100 SME loans worth US 100,000
  before guarantee program
• Bank Y makes no SME loans before guarantee
  program
• After subsidized guarantee program for Bank Y
  makes 100 SME loans for US 100,000, on more
  favorable terms than Bank X, Bank Y takes all of
  Bank Xs SME clients
• No additionality in SME lending, although 100
  loans are guaranteed
• Ignoring substitution results in overestimate of
  benefits

16
Attribution ProblemsCase D

• What would the lender have done over time without
  the loan guarantee program?
• Isolating the effect of the guarantee program on
  lender behavior from the effects of other changes
  in the economy over time is difficult.

17
Case D continued

• Same as Case A, except no loan guarantee program.
  
• Economic reforms increased the profitability of
  economic activities of SMEs and lenders
  voluntarily decide it is good business to expand
  lending to this group.
• Some of the additional lending associated with
  the guarantee program may be due to other factors
  

18
Case D continued

• Attributing all increases (over time) in lending
  to SMEs to a loan guarantee program, also
  overestimates the benefits of the guarantee
  program. At least some of the changes in lending
  might have occurred without the guarantee.

19
Other Measures of Program Performance

• Additionality alone is not sufficient
• Is additionality sustained?
• Outreach number of loans guaranteed compared to
  total number in target group
• Sustainability of guarantee program subsidy
  dependence

20
Conclusions

• Unclear if retail programs are effective
• Risk only one of a number of problems that limit
  SME access to formal loans
• Transactions costs may be more important
• Costs may be greater than benefits
• Benefits difficult to document and are often
  overestimated
• Increase overall transactions costs in system
• Other policy options might be more effective

21
Discussion Topics

• Is there evidence showing retail loan guarantees
  anywhere result in additional SME lending?
• Are transaction costs a major problem in SME
  lending for both lender and borrower?
• Do retail loan guarantee and reinsurance programs
  increase transaction costs?
• Are portfolio or wholesale loan guarantees
  appropriate where large amounts of deposits are
  mobilized by potential SME lenders?
• Importance of microlending in Asia

22
Discussion Topics continued

• Do loan guarantee programs eliminate most
  problems that block creditworthy SMEs from
  accessing loans?
• Would a supervising agency have the skills and
  information to second-guess the decisions of
  guarantee agencies, who, in turn, second guess
  the lending decisions made by the ultimate
  lender?
• Instead of loan guarantee programs, why not
  charge higher interest rates on loans? This is
  what many successful NGOs and informal lenders
  do.

23
References

• Graham Bannock and Partners Ltd. Credit
  Guarantee Schemes for Small Business Lending A
  Global Perspective, unpublished report prepared
  for ODA, London, England, April 1997.
• Vogel, Robert C. and Dale W Adams, Costs and
  Benefits of Loan Guarantee Programs, The
  Financier, 4(May 1997) 22-29.

24
References continued

• Meyer, Richard L. and Geetha Nagarajan, Credit
  Guarantee Schemes for Developing Countries
  Theory, Design and Evaluations, unpublished
  report prepared for the African Bureau, U.S.
  Agency for International Development, Washington,
  D.C. April 15, 1996.
• Department of International Development, Do
  Credit Guarantees Lead to Improved Access to
  Financial Services? Recent Evidence from Chile,
  Egypt, India, and Poland. Policy Division
  Working Paper, Department for International
  Development, London, February 2005.
• Orbeta, A.C., C.G. Lopez and Dale W Adams, An
  Assessment of Loan Guarantee Programs for
  Small-scale Borrowers in the Philippines,
  Working Paper No. 12, Credit Policy Improvement
  Program, Secretary of Finance, Manila,
  Philippines, September 1998.