«by MICHAEL GUDGER FAO AGRICULTURAL SERVICES BULLETIN 129 The designations employed and the presentation of material in this publication do not imply ...»
Issuing guarantees, critics point out, is costly. Administrative overheads are very high and consume the capital of the guarantee fund if it does not quickly establish a large enough revenue stream to spread these costs over a very large number of operations. Guarantee companies have generally failed to generate a large volume of low unit cost operations.
Pricing of Guarantees
Issuing a guarantee requires the guarantor to accurately assess the riskiness of the loan and to develop a price that reflects that risk. If it cannot do so effectively, the guarantee company’s losses will exceed its income, reducing its capital. To date, no SME and micro-borrower guarantee company has consistently priced its guarantees at levels that permit it to maintain its capital, let alone increase its capital through retained profits. In fact, the management of a guarantee company is caught in a dilemma. It can charge a premium systematically related to the cost of guarantees, which the guarantee unattractive because of the price. Alternatively, management can price the guarantee at a low enough price to sell it but at the cost of losing capital because the price does not cover costs and losses. These losses often are offset either by donor funds or continuing subsidies.
Guarantee Funds have no Comparative Advantage in Credit Assessment Bankers, critics of guarantees argue, are better at evaluating risk than are guarantee funds.
Bankers both have more experience and know their clients better than guarantee funds.
“Like a bank, the fund must base its decision on the quality of the proposals presented to it, and on the history, management style and other qualifications of the concerned entrepreneurs. The fund, in other words, must use the same methods of credit analysis as banks in determining whether or not to issue a guarantee. Moreover, the fund in all likelihood will not have better or more experienced credit analysts than banks. As a result the fund is likely to guarantee projects with about the same reluctance or willingness as a bank is to loan to a given undertaking.
If banks often reject loans because the applicant lacks collateral, and the guarantee fund is supposed to provide this collateral, then the fund must base its credit decisions on fewer criteria than those used by the bank. In short, the fund’s range of credit analysis is narrower than that of the banks.”2
Banks Will Engage in Opportunistic Behavior with Guarantees
If lenders are offered third party credit guarantees, they will select against the guarantor and guarantee only the loans that are at high risk of going bad or have already gone bad. This anti
-selection process will give the guarantee fund a lower quality credit portfolio that will in turn be very, and probably destructively, costly. This point is made convincingly by Prof. Claudio Gonzalez Vega in numerous postings on the Development Finance Network. My paper on the Caja Social in Colombia presented at an Inter -American Bank Symposium sets out in some detail the dynamics of this interaction between lender and guarantee fund.3 Krzysztof Herbst, “Local Guarantee Funds in Poland” in Antony Levitas and Grazyna Gesicka, editors, Local Guarantee Funds, Friedrich Ebert Stiftung, Warsaw, 1995, pp. 59-60.
See M. Gudger, Sustentabilidad de los Sistemas de Garantia de Credito” in Sistemas, 1996.
Banks Prepared to Lend to SME and Micro Borrowers Can establish Their Own Guarantees There is no a priori reason that a bank would want a third party guarantee. A bank could simply increase the interest rate instead of buying a guarantee and set up a loan loss reserve which is retained instead of being paid to a guarantee fund. In fact, this is exactly what some successful SME and micro lenders do. They know better than do guarantee funds the likely default rate, and they establish provisions to meet these losses without any additional administrative overheads. Again, the Caja Social in Colombia provides an example. The Caja was established in 1901 and now has 134 offices in 42 cities. In 1995 it had assets of $580 million of which $423 million were loans. The average loan size is around $2,000.
Since the late 1980s, the Caja has operated profitably by meeting its three types of financial intermediation costs: the cost of funds, the costs of administration, and the cost of loan losses.
The loan losses in the 1985-95 period were about 2% per year, considerably lower than the fees charged by guarantee funds operating in Colombia in the same period.
SME and Micro Lending Requires an “Appropriate” Technology, not Guarantees
SME and Micro-lenders do not require guarantee but instead a different loan appraisal and loan supervision technology that is substantially different from the norms of commercial banking. From several studies of the Cajas Municipales in Peru the elements of this successful strategy can be identified.
• A target market clearly defined by terms of socio-economic and geographical characteristics.
• Products suited to the situation of the target market instead of standard loan products that do not meet the needs of the targeted market segment.
• An appropriate lending technology for a sector characterized by small loan sizes and high administrative costs, which generates a sufficient volume of good loans to keep expenses at manageable levels and not impair profitability.
• A cost-effective strategy of acquiring small and micro clients through a community outreach programme and organization of groups under the guidance of community-based loan officers.
• An inexpensive loan delivery system.
• An effective loan monitoring system to gather information continuously and permit intervention before a loan becomes non-performing.
• An enforceable system for classifying and identifying borrowers who are able and willing to honour repayment commitments and procedures that channel loans preferentially to those that do so over time.
• A strategy and a reputation for vigorously pursuing defaulters, as a lender thought to be “soft” on loan collections will face increased defaults.
• A realistic loan pricing policy so that interest rates are adequate to cover costs of funds, operating costs and loan defaults.
From this perspective, SME and micro-lending is less a matter of collateral and third party guarantees than it is the development of a coherent strategy and technology for reaching these groups and an operational methodology that will permit the lender to survive and grow.
Guarantees would be of little value to a lender who had such a strategy. Guarantees would do little to induce a lender without such a strategy to lend to SME and micro-borrowers, and would do nothing to insure the survival, let alone growth, of a lender without such a strategy.
Guarantees Do Not Produce Additionality
Proponents of guarantees cite additionality as the major justification for the creation of the guarantee funds with public or donor funds. More credit on more favourable terms flows to the target groups with guarantees than without guarantees. Prof. Dale Adams has directly challenged this argument. He argues that there is no evidence of additionality. None of the proponents, in his view, have produced evidence that guarantees result in additionality. He further argues that as guarantee funds are part of a larger credit-targeting programme, the usual result is that subsidized credit flows to politically important groups largely irrespective of their creditworthiness.
CHAPTER TWO CONTEMPORARY CREDIT GUARANTEE PRACTICE
The arguments set out in the previous chapter are theoretical and make only passing reference to operating credit guarantee systems. Of these there are a large number both in Europe and the U.S. with many examples also in developing countries4. This chapter reviews some of the experience in Europe, Asia and the non-Asian developing countries in addition to multilateral guarantee programmes sponsored by donors and NGOs. Chapter Three is devoted to the U.S.
financial guarantee industry, which has been neglected as a source of information as to how to operate guarantees profitably and produce massive additionality.
EUROPEAN GUARANTEE SYSTEMS: WESTERN EUROPEAN COUNTRIES:
Advocates of guarantees generally hold up European programmes as models of successful guarantee programmes and argue that the long history of these programmes serves as a guide for other countries. Much of the discourse assumes the benefits for SME borrowers and focuses intensely upon the various administrative structures.5 Almost all West European countries have guarantee systems. In some countries, these are quite large and sophisticated systems based on multi-level mutual models (France and Italy) and other countries, they have developed into guarantee banks (Germany and Austria). Despite the size, sophistication and long history of these guarantee schemes there is surprising little analysis of them. Published materials tend largely to be self-serving and/or descriptive, but seldom delve into the issues of costs, subsidy-dependence, additionality and sustainability. This material usually argues benefits but seldom engages in a rigorous analysis of what precisely these are or how they are generated, or perhaps most important of all. Whether the benefits could have been produced, for less cost, in another way.
An excellent discussion of the various structures of guarantee schemes with particular reference to developing countries can be found in Richard L. Meyer and Geetha Nagarajan, Credit Guarantee Schemes for Developing Countries: Theory, Design and Evaluation. USAID, Africa Bureau, Washington, 1996. A companion publication is the Annotated Bibliography on Agricultural Credit and Rural Savings. Volume XVII, A special Issue on Credit Guarantee Schemes for Agriculture and Small, Medium and Micro Enterprises, Rural Finance Program, The Ohio State University, 1996. This is an excellent source of additional bibliographical information.
See Juan Luis Llorens, “Los Sistemas de Garantias de Credito para las PYMES en Europa” in Sistemas, 1996 for a description of the European models.
The first analytic work to appear on European guarantee funds does not provide a very positive appraisal. 6 It suggests that the volumes of guarantees are very low and that the cost per guarantee is quite high. In the 12 European countries studied, the total number of guarantees issued in 1993 (the last year for which data are available) was about 43,000 with an average loan size of $98,000. Nearly half of the total number of loans (18,000) were originated in Greece where the average loan size was about $23,000. The largest country in the study, Germany, originated only 6,400 guaranteed loans with an average guarantee commitment of $270,000. Spain originated about 6,500 with the average guarantee commitment of $48,000; France reported 2,700 loans whose average guaranteed amount was about $262,000. In Italy only 956 new guaranteed loans were issued with an average guarantee commitment of $287,000. The U.K. issued about 3,900 guaranteed loans with an average value of $47,000. In the 12 countries, the volume of new guaranteed loans was estimated at about 2-3% of total annual bank lending.
The German scheme requires budget support of about US$27 million per year or about $4,200 per loan guaranteed. The cost to the government budget per loan is about 1.5% of the loan value guaranteed. Italy has a budget outlay estimated at US$38 million or about US$40,000 per loan guaranteed. The relationship between the cost to the budget per loan and the loan size is 14%. In the U.K., the budget outlay was at least US$24 million or just over $6,000 per loan. This cost to the budget is about 13% of the loan size. In France, the cost data is not available as France provides recapitalizations of the guarantee system instead of an administrative subsidy.
It is hardly surprising in the light of the low volume of operations and the high cost per operation that the German Bundesbank’s official view is unsupportive. It sees the German and European schemes as an interest rate subsidy to small and medium size enterprises for political purposes and, thus, not a “natural component of a market economy.”7 Securitization of Guaranteed SME Loans in Europe and Finance for Innovation. Graham Bannock & Partners study for the European Innovation Monitoring System (DGXIII-D), Brussels, 1995.
Austria A recent study of the Austrian guarantee scheme seems to support these findings. A total of 5,242 guarantees were issued in 1994 for a total amount of 2.4 billion Austrian Schillings. At a current exchange rate of 13.065 to the dollar, the total value of guarantees issued was about US$ 184 million for an average of about US$ $35,000 per guarantee. Total subsidies in 1994 were AS 377 million or about US $28.9 million. This implies a cost of about US$5,500 per new guarantees issued, or roughly 15% of the average value of a guarantee issued in that year.8
Some older data from the very large Spanish scheme sheds some additional light on the issues of costs and the relationship of these costs the value of guarantees issued. These data suggest that the administrative costs of issuing guarantees average about 6% of the value of the loans guaranteed, of which only about 50% are derived from the fees charged. The rest derives from the investment income on the capital, which apparently came principally from the budget.9 The charge of 6% to issue the guarantees does not include the costs of calls on the guarantees.10 Thus, the Spanish scheme, in terms of cost, falls between the very efficient German programme and the more costly schemes in the U.K. and Italy. The data provided are silent on subsidy support for the Spanish programme.
Kurt Leutgeb, “The Experience of Burges Forderungbank Regarding Loan Guarantee Institutions in Central and Eastern Europe,” Burges Forderungsbank, Vienna, 1995.
Data supplied to the author by the Sistema Espanol de Garantias Reciprocas.
The charge actually may be higher as the loan amounts appear to be the total value, not the portion guaranteed. Typically 50%-75% of a loan is guaranteed.
EASTERN EUROPEAN COUNTRIES