«by MICHAEL GUDGER FAO AGRICULTURAL SERVICES BULLETIN 129 The designations employed and the presentation of material in this publication do not imply ...»
Several Eastern European countries have experimented with or have established guarantee funds. Poland has the greatest number of guarantee schemes, both in the public sector and in the private sector. Several state-owned banks (including the Bank for Food Economy and the National Economy Bank) issue bank guarantees for various purposes as traditional bank guarantees. No information can be gathered as to the volume and the results of these bank guarantees except that the Bank for Food Economy as of 1996 carried an item on its balance sheet showing a loss of about U.S. $600 million for export guarantees.11 Specialized state agencies grant credit guarantees. These agencies are the Agricultural Marketing Agency which issues credit guarantees for purchase and storage of food products, the State Treasury Agricultural Property Agency guarantees up to 70% of the price paid by private parties to purchase agricultural land from the state, and the Agricultural Restructuring and Modernization Agency issues guarantees to help finance investment in the agricultural sector. These agencies either do not release data or do not disaggregate it from their overall budget. Interviews however suggest that the total volumes of these guarantees are quite small.
Poland has numerous very small-scale private guarantee funds, many of which are contained within a development foundation and thus are very difficult to evaluate in any systematic manner. These regional funds include the Nida Enterprise Fund, the Loan-Investment Fund (Olysztyn), the Lublin Enterprise Fund, The Bialystok Enterprise Fund, and the Bilgorajska Regional Development Fund. As of late 1996, additional funds were in various stages of formation.
Additionally, USAID and the Polish-American Bank have some operations covered by the Loan Portfolio Guarantee Programme. The British Know How Fund has contributed to the capital of the Polish British Enterprise foundations in Lublin and Bialystok about $20 million.
The STRUDER Programme, with EU support, is in the process of setting up an additional guarantee fund with capital of about ECU 5 million.
This loss however is probably due more to the general financial crisis of the banking industry than to the The Government of Poland owns KUKE, a new export insurance company that issues guarantees to Polish exporters. While not yet profitable, KUKE does have a declining loss ratio and a solid balance sheet characterized by steadily growing capital. Insurance operations continue to produce net negative results for the present while investment activities make the company profitable. Almost all of the operations are with large industrial and commercial companies. No exports from the SME sector have been covered by guarantees.
In late 1996, the guarantee funds met in Nidzica and formed the National Association of Guarantee Funds both to share experiences and to serve as an apex organization and the lobbying arm of the movement. In fact, lobbying was defined as their most important activity for 1997.
General conclusions about the Polish guarantee fund movement are difficult given the structure of most of the funds within regional development foundations. Several important
aspects that should be noted:
• The volume of operations is very small, with usually only a dozen or so guarantees being issued by each of the funds.
• The leverage (or capital multiplier) of the guarantee fund ranges from 0.3 to 1 up to 2 to 1. Some have yet to issue guarantees for more than about 1/3 of their capital while others have issued guarantees outstanding that are twice as large as their capital. Most of the foundations have to make a term deposit in the lending bank for the amount of the loan guaranteed. In the best cases, they have to deposit 50% of the amount guaranteed.
• Guarantee fees charged are very low and do not appear systematically related to the expected rates of default or the rate of default that characterizes SME lending in general.
• Operating expenses are quite high. They generally far exceed guarantee fee income.
• Losses to date are very low as most of the loans were recently guaranteed and many are still disbursing.
fact that guarantees were the instrument of the loss.
The Hungarian guarantee programme arose from the PHARE aid programme of the European Union, which was designed to help the country restructure as a market driven economy. In 1990, PHARE supplied ECU 10 million to establish a rural credit guarantee fund. Later an additional ECU 10 million was granted by the EU, and with local contributions reached approximately ECU 22 million.
Between 1991 and 1996, the Rural Credit Guarantee Fund (RCGF) issued a total of 1,985 guarantees, an average of about 330 guarantees per year. The total value of these loans was reported to be 17,871 million Hungarian Forints (HUF). Inflation was quite high, making this figure is difficult to interpret. However, at 1991 prices, the value of the loans were 8,610 million HUF, or about 4.3 million HUF per loan. This is approximately U.S.$ 25,000 per guaranteed loan. The RCGF issued guarantees for about 87% of the amount of these loans.
Operating expenses were about HUF 73.2 million at 1991 prices and the guarantees paid (net of recoveries) were HUF 283.2 million for combined expenses of HUF 356.4 million.
The operating expense ratio of the RCGF was a surprisingly low 0.85% of the amount of guarantees issued. Defaults however were about 4.1%. Thus, for the period 1991-1996, the RCGF required about 5% of the loans guaranteed and about 5.7% of the amount guaranteed (5%/87%) to meet the operating and loss costs of the guarantees.
This are surprisingly low numbers in an economy in transition. They should be taken with some caution for two reasons. First, during a period of high inflation, loans are often paid off as their real value has shrunk to a fraction of the initial value. This often enables the borrower to obtain another loan. Second, there is a clear trend toward higher operating costs and increasing defaults as the portfolio of guarantees matures. Operating costs rose from HUF 4 million to HUF 17.4 million between 1991 and 1994 and then stabilized at about HUF 16.5 million. In 1996, administrative costs were about 1.5% of the loans guaranteed and about 1.7% of the loans guaranteed. The defaults had risen to about 5.9% of the loans guaranteed and about 6.8% of the amount guaranteed, thus producing a combined operating cost of about 8.5% of the amount of guarantees issued. The balance sheets of the RCGF do not show fee income separately but it was reported to range from 2-4%. If this were the case, fee income would offset between a quarter and a half of these costs. The rest would have to be met by investment income on the donated capital.
Romania, apparently following the Hungarian example, set up two funds with PHARE support in 1994. One fund guarantees loans to SMEs while the other focuses on agriculture.
As of mid-1997, these funds had issued 118 guarantees for about US$27 million. No losses have yet been taken although in the agriculture fund, it is reported that 21 of 67 loans are rated as “sub-standard to bad.” Both funds depend heavily upon earnings on investments. When the Subsidy Dependence Index (SDI) is applied to 1995 balance sheet data, the SDI is 35% for the SME fund and 95% for the agriculture fund. However, SDI numbers at such an early stage should be seen only as a very early indication that income does not meet expenses.
More telling is a calculation of the cost of administration plus the provisions in the SME fund as a percentage of loans, which is about 9.4%. Stated differently, it costs $9.40 to guarantee $100, which can support a loan size of $140-150. If total costs (variable and fixed) are included the cost rises to about 16%. Premium incomes cover about 21% of the $9.40 or $1.97. The rest comes from investments of capital supplied at no opportunity cost. In the agriculture fund, it costs about $19.84 in administration and loss provisions to issue a guarantee for $100 supporting a loan size of $166. Of this amount, premiums and fees cover only 15.7% ($2.11); the rest derives from investment income. Thus, if the Romanian funds were to fully meet costs by transferring them to the users of guarantees, they would have to charge about 7.4% of the value of the guarantees for the SME fund and excess of 17% for the agriculture fund.12
Data supplied by Geetha Nagarajan.
SUMMARY OF THE EUROPEAN EXPERIENCE:
In Western Europe, there are no data as to the losses suffered by the guarantee funds and the amount of these losses offset by guarantee income. This important missing information would provide us with a clearer picture of these schemes. Without this information, we can only focus upon the administrative cost of issuing guarantees. With the exception of the relatively efficient German operation where guarantee fund operating costs are only about 1.5% of the amount of the guaranteed loan, the rest of the countries for which data are available, Britain, Italy and Austria, have administrative costs of 13%, 14% and 15% of the amounts guaranteed respectively. The large Spanish system’s data, although not complete, suggests a charge of about 6% is necessary to cover administration. Available data suggests that, Germany excepted, it is unlikely that borrowers would tolerate the increased cost of borrowing if they were charged a fee that covered the administrative costs alone. To this would have to be added the default costs, which are not readily available to the public.
Administrative costs appear lower in Eastern Europe, but default costs bring total costs to levels that would discourage many reasonable borrowers. In Poland, administrative costs exceed guarantee incomes in most, if not all, the small programmes. In Hungary, the administrative charges appear to be around 5.7% and the losses amount to about 6.8% in the last year for which data is available, for a total cost of 12.5%. In Romania the cost of administration and provisions for bad debt net of guarantee fees appear to be around 7.4% for the SME fund and about 17% for the agriculture fund.
The West European guarantee schemes are the oldest and among the largest in the world. It seems doubtful that they could operate without subsidy if they were required to charge the full cost of their operations. Heavily dependent on subsidies, they have low volumes of operations and high operating costs that make them unlikely models for developing countries considering using guarantee funds to promote lending to SMEs and micro borrowers.
However, none of this data bears upon the arguments concerning additionality, but, if there is additionality, it is very costly to achieve.
CANADA’S SMALL BUSINESS LOANS ADMINISTRATION
Canada’s Small Business Loans Acts was designed to assist SMEs gain access to term financing by providing government backed guarantees issued by the Small Business Loans Administration (SBLA). Its Business Improvement Loan (BIL) programme covers businesses having gross revenues of up C$5.0 million. A bank, a credit union, a trust company, a caisse populaire, or an insurance company may issue guaranteed loans.13 Banks are allowed to charge prime +3% for floating rate loans and for fixed rate loans, 3% over the residential mortgage rate for the applicable term. The maximum term of a guaranteed loan is 10 years. Currently, lenders pay to the government a one-time fee of 2% of the value of the loan (which may be financed by the lender). The lenders pay an annual administration fee of 1.25% of the yearly average end of month balance outstanding on the BIL’s, which can be passed on to the borrower only through the interest rate.
The guarantee, called the Crown’s Contingent Liability to Lenders, after January 1996 is 90% of the losses sustained on of the first C$250,000 of BIL’s registered loan amount, 50% of the second C$125,000 and 10% of any amount over C$375,000. The guaranteed loans may be used for purchase of land, premises, equipment and payment of the programme fees. Only farming, covered under a separate programme, is excluded.
The financial performances of the BILs are disclosed in the annual reports of the SBLA.14 The total value of loans made from programme’s beginning in January 1961 through March 1996 was $18.2 billion. However, $17.7 billion was made during the 1977-1996 period.
Since 1961, the government has paid indemnities net of recoveries C$503 million. The fee income during the period was C$225.2 million, producing a net loss of C$278 million. Net losses have been 2.76% of the total amount of BIL loans. Of this, about 1.24% was collected in fees, implying that government absorbed the difference of about 1.52-%.
Interestingly, the SBLA calculates its loss reimbursements differently, which produces a larger percentage loss. Of C $17.7 billion of BILs registered between 1977 and 1996, borrowers repaid C$10.8 billion and the losses reimbursed by the Crown were C$515 million, indicating total BIL recovery by banks of C$11,3 billion. Guarantee payments are thus about 4.5% on the total reflows to banks, with an outstanding balance of C $ 6.3 billion still carried All currency amounts in the section are cited in Canadian dollars. The exchange rate at the time of writing this section was C$ 1.41 = U.S.$ 1.
The Annual Reports can be found at http://strategis.ic.gc.ca/SSG/la01036r.html.
on lenders’ books. The 1996 Annual Report notes that the government’s contingent liability on this amount is about C$1.4 billion which will produce about C$500 million of additional claims losses, thus increasing the cumulative loss rate as calculated by the SBLA to around 5.75%.
From 1994-1996, the Government has moved the loan guarantees toward a fully self-funding basis. In this period, claims paid were about C$117.6 million while fees collected were C$179 million. Up front loan registration fees were doubled from 1% to 2% while loan volume increased markedly from C$1.3 billion to C$9.2 billion in the 1994-1996 period. A considerable increase in claims is expected during the third and fifth years following this upsurge in coverage, justifying the increase in fees. The SBLA’s objective is to ensure that claims costs will be fully offset by revenues.