Abstract

Agencies observed that their most important decision resulting from the risk assessment process was whether to provide export credit cover to a country or not. Every agency was off cover altogether for some countries where the stance of policies, the track record, and the external environment appeared to offer little prospect that the country would be in a position to service new loans on commercial terms. However, almost every agency covered in this study was on cover for political reasons for some countries for which cover would not be made available if the outcome of the risk assessment process had been the only determinant of cover policy.

Links Between Risk Assessment and Cover Policy

Agencies observed that their most important decision resulting from the risk assessment process was whether to provide export credit cover to a country or not. Every agency was off cover altogether for some countries where the stance of policies, the track record, and the external environment appeared to offer little prospect that the country would be in a position to service new loans on commercial terms. However, almost every agency covered in this study was on cover for political reasons for some countries for which cover would not be made available if the outcome of the risk assessment process had been the only determinant of cover policy.

The trend toward more differentiated and nuanced cover policies has become more pronounced in recent years, and most agencies attempt to provide cover, even if very limited, for as large a number of countries as possible.25 If cover is made available, risk assessment is used by agencies to determine the type of cover (short-term or medium- and long-term) and the pricing of cover. Recent years have also seen an evolution in the thinking of agencies on the effectiveness of their cover policy instruments. In the late 1980s, differentiation of premia charged had been seen as a way to make cover available to most borrowing countries, provided that the price accurately reflected the risk entailed. More recently, however, almost all agencies have come to the view that demand for export credits in high-risk markets is insufficiently price-sensitive, and the risks resulting from asymmetric information about creditworthiness too large for premia to be the main means of allocating cover. While the trend toward differentiated premia for borrowers has continued, not least because of the continued weak financial performance of most agencies, the use of other cover policy instruments is generally also regarded as necessary in managing risk portfolios. These include quantitative ceilings on the total annual level of commitments to countries and/or on the size of individual transactions that can be covered, and security requirements, such as the guarantee of banks in the form of irrevocable letters of credit.26 Instruments of cover policy are discussed in more detail in Box 7. An example of how one major agency (Hermes of Germany) sets its premia is given in Box 8.

General Stance of Agencies’ Cover Policies

Tables 2 and 3 summarize the stance of cover policies at end-1993 for 21 developing countries, distinguishing between short-term cover and medium- and long-term cover.27 The tables show clearly that most agencies impose some restrictions on their cover, with a requirement that security devices be associated with cover being the most frequently used restriction on short-term cover, and with ceilings on the size of transactions or the total volume of new cover being the most frequently used restrictions on medium- and long-term cover.28

Chart 8 presents the evolution of agencies’ cover policy stance in their main markets over the period 1989 to 1993 in the form of country-specific indices. The indices reflect the ease and cost of obtaining cover, with a decline in the indices representing a tightening of cover policy. It is notable that for most of the countries shown, agencies have tightened their cover policy over the period studied. This is consistent with the increased focus on risk assessment. However, as noted above, it has not prevented an increase in the volume of export credits over the same period. Rather, the trend can be seen as evidence of the strength of demand for export credits: as agencies have tightened cover policy and raised the price of cover, exporters and borrowing countries have been able and willing to meet the new conditions and pay higher premia.

Chart 8.
Chart 8.

Cover Policy Indices1

Sources: Export credit agencies; Berne Union; and IMF staff estimates.1The index of cover policy is based on reports from 12 major agencies to the Berne Union in the final quarter of each year. If an agency is off cover for a country, a value of zero is assigned to that agency: less restrictive cover stances are given higher values, up to a value of 8 for an agency that is open for cover without restrictions. The value assigned to each agency is weighted to reflect the relative importance of that agency for the country concerned.

Cover Policy for Major Export Credit Markets

The recent experience with the five markets where export credit exposure was highest at end-1993 (Russia and the former U.S.S.R. taken together, China, Brazil, Mexico, and Indonesia) illustrates how agencies’ policies have shaped export credit developments in recent years. Chart 9 compares the developments in total exposure (including arrears and unrecovered claims) in these top five markets. Agencies’ experience with Russia is discussed in the section on cover policy for the economies in transition. Agencies’ experience in other major markets is discussed below.

Chart 9.
Chart 9.

Total Exposure in Top Five Markets

(In billions of U.S. dollars)

Sources: Berne Union: and IMF staff estimates.

China

The rise in export credit agencies’ exposure to China, as shown in Chart 10, is striking. It has been driven in large part by increases in demand associated with China’s high growth rate and increases in the investment ratio. The increase in exposure is entirely a reflection of new commitments. China was by far the most important market for export credit agencies in 1993, receiving more than twice as much in new commitments as any other single country, and this trend has continued in 1994. By end-September 1994, export credit agencies’ exposure to China had risen to over $35 billion, twice the level reported two years earlier. Competition remains intense among agencies for business with China, and subsidization of exports in the form of tied-aid credits remains an important factor.

Chart 10.
Chart 10.

Export Credit Exposure: China, Brazil, and Mexico1

(In billions of U.S. dollars)

Sources: Berne Union; and IMF staff estimates.1For data definitions, see Chart 4.

Instruments of Cover Policy

Premium Rates

Realistic pricing of risks is seen by all agencies as an important objective, and all agencies have now established premium structures for export credits that provide a direct link between the rate of premium and the perceived degree of country risk. Most agencies have been devoting considerable resources to analyses of both the degree of differentiation and the overall level of premia. Premia have an obvious direct impact on the financial positions of agencies. Many agencies had raised their premia, in some cases for the first time in 30 years, in the wake of the first series of reschedulings and cash flow deficits in the early 1980s, but then made no further general adjustments to their rates. However, as cash flow deficits persisted during the late 1980s, many agencies raised premia again. The unexpected severity of more recent deficits has led most agencies to adjust overall premium rates still further, while increasing the degree of differentiation. Premia vary with the maturity of the export credit, the type of borrower, and the borrowing country.

Agencies also saw differentiated premium rates as a way to help dampen demand in riskier markets, thus rationing cover in an economically efficient manner, and to encourage a shift in the direction of trade toward stronger markets. However, agencies observed that premium differentiation had clear practical limits. Some believed that even the highest levels of premia that they charge fail to compensate them adequately for some risks that they take.1 Moreover, almost all agencies were skeptical about the adequacy of premia as an instrument for limiting demand and thus for managing their risk portfolios. They thought that demand for cover was highly inelastic with respect to premium levels in the more risky countries. Therefore, almost all found it necessary to supplement high premia with other instruments, particularly quantitative ceilings, for countries seen as high risks.

Discussions on premium policy among agencies have intensified in recent years. One reason for this increased focus on premia is their significance as an instrument of competition among agencies, and between agencies and private sector insurers. Though some agencies believe that different premium levels have little impact on the competitiveness of export transactions except in markets with the lowest risk, the levels and structure of premia on medium- and long-term export credits have recently become the subject of intense debate both in the context of moves towards harmonization in the EU, and in the OECD export credit group.2

The increased attention being given to premia may also reflect the diminished relative importance of interest rate subsidies as an instrument of competition among agencies, as the result of the decline in international interest rates from the higher levels of the 1980s and the strengthening of the provisions of the OECD Consensus on the use of interest subsidies. The issue of premium rates has thus become more closely connected with the questions on export subsidization (see also Appendix III).

Quantitative Limits and Other Instruments

Limits on total exposure continue to be a feature of most agencies’ cover policies toward at least some developing countries. Such ceilings are also seen as an important means of ensuring portfolio diversification. However, most agencies reported that they used these ceilings flexibly, and, when exposure reached the ceiling, a thorough review would be conducted. Typically, the ceiling would be raised, but this would normally be accompanied by other measures such as increases in premium rates or other restrictions.

Other instruments, such as reducing the percentage of cover or extending the period the exporter must wait before filing a claim, continue to be used, but less than in the past. A number of agencies reported that limiting the percentage of cover was a very effective device for curbing demand in cases where they had reached the upper limit of what they considered reasonable premia. Others thought that it was precisely in high-risk cases where demand was inelastic that exporters would increase contract prices to offset their participation in the risk, and this would have a perverse effect of increasing the total amount of the export credit. It is for this reason that one agency covers typically 100 percent of the export credit, and most others have increased the share covered to between 90 and 95 percent. However, most agencies consider it an important principle that the exporter or financier should continue to bear a part of the risk against the failure of an export transaction or project as a safeguard against badly designed or unviable transactions.

Most agencies do not use the claims-waiting period as an active instrument of cover policy but rather in cases where payments tend to be late, but can be counted on to arrive eventually. A longer waiting period in these cases prevents a proliferation of claims payments that are quickly followed by recovery. Other agencies commented that in their experience extensions only served to lengthen delays still further as debtor countries became aware of changes in agencies’ policies, and that the lengthening of claims-waiting periods could well lead agencies to remain on cover longer than warranted by underlying conditions, as shown by their recent experience with Iran.

1Several agencies commented that there was a maximum rate of premium that could reasonably be charged, and that premia above this level would only be paid by exporters in cases where they expected the borrowers to default. Some agencies also noted that in some cases where they were on cover for national interest reasons, premia were kept at levels below those that would have been implied by their risk assessment systems because the latter would be prohibitive.2Discussions have centered on the level of premia that would allow agencies to “break even” and thus prevent indirect export subsidies. The criterion of “break-even points” is, of course, crucially dependent not just on premium levels but also on the timeframe over which agencies would be expected to break even. In this context, most agencies argue that premium levels cannot (and should not) be set at rates that would cover cash flow deficits on old business but rather at levels that compensate for expected deficits on more recent business activity only. Discussions continue, and the Participants to the Arrangement have recently agreed to try to establish guiding principles for setting premia and related conditions.

Brazil

Experience with Brazil has been very different (Chart 10). While some export credit agencies have continued to make new commitments, these have remained modest. Agencies’ total exposure to Brazil has, however, fallen only marginally from the levels of the late 1980s, because of Brazil’s continued recourse to Paris Club reschedulings (through August 1993) which is clearly reflected in the rising proportion of agencies’ portfolios in the form of arrears and unrecovered claims. Given the close relationship between agencies’ cover policies, and hence new commitments, and borrowing countries’ payments performance, the relatively low level of new commitments to Brazil can be seen as a direct consequence of Brazil’s inconsistent payments record under reschedulings and on post-cutoff date credits over the last several years. Moreover, commitments that were made were increasingly channeled to Brazil’s private sector.

Mexico

The increase in commitments to Mexico in recent years reflects agencies’ confidence in the country’s adjustment efforts and in the debt subordination strategy. Export credit agencies provided early and substantial financing for Mexico’s adjustment programs and thus helped to accelerate access to capital markets following the debt- and debt-service-reduction agreement with commercial banks (Chart 10).29 New commitments have remained at high levels since then. The result has been a steady increase in the exposure of export credit agencies with a declining share of unrecovered claims, as Mexico graduated from the Paris Club and is now making substantial repayments on rescheduled claims.

Iran

Starting in 1989, Iran was seen as a major new market. Exporters competed intensely for new business but were reluctant to extend credit without official support. Most, but not all, agencies quite readily provided export credit cover, first on a short-term basis and then with longer maturities, and this was reflected in a very rapid rise in export credit commitments. This was followed by an abrupt decline in export credit cover in 1993 as doubts arose about Iran’s ability to continue debt servicing and arrears began to emerge on a large scale. Despite the sharp reduction in new commitments, agencies’ exposure to Iran remained broadly unchanged through end-1993, but with a striking increase in the share of arrears and unrecovered claims (Chart 11).30

Chart 11.
Chart 11.

Export Credit Exposure: Iran and Algeria1

(In billions of U.S. dollars)

Sources: Berne Union; and IMF staff estimates.1For data definitions, see Chart 4.

Most agencies agreed that their cover policies had led to a rapid buildup of debt and contributed to the emergence of difficulties, particularly through a rapid increase in the share of export credits on relatively short maturities in the period before payments difficulties became acute. To some extent this reflected demand factors, in particular demand for consumer goods and a strong bias in Iran’s external debt management toward short-term credits even for investment goods.

Determination of Premium Levels

All agencies visited employ highly structured premium systems. However, the average level and steepness of the premium curve varies substantially across agencies, as do the degree of differentiation among recipient countries and the methods of calculating premia. These differences reflect the variety of views on the relative importance of various risks and on the appropriate method of pricing each of them, as well as the fact that premium levels are used in conjunction with varying constellations of other cover policy instruments, and that the terms of the cover agreements vary widely.

The premium system1 used by Hermes, the German export credit insurance agency, utilizes five country risk categories, namely:

Category 1: OECD countries;

Category 2: Countries with well-established payments records and no expected payment difficulties;

Category 3: Developing countries with the typical level of developing country risk;

Category 4: Rescheduling countries and those with imminent payment problems; and

Category 5: High-risk countries for which Hermes is off cover cither entirely or for medium-and long-term loans.

A country where Hermes has a very high concentration of risk may be assigned to a higher country category. Premium charges are due at the time the risk begins, with very limited exceptions. The new system differentiates between public and private buyers both for suppliers’ and buyers’ credits. In the case of public buyers the basic premium is determined only by the country risk. For Category 3 countries—the category taken as the basis for determining premia—1 percent of the covered amount is charged. Premium levels for other categories are adjusted by a factor reflecting the difference in country risk relative to Category 3. Thus for Category 1, the basic premium is 0.33 percent; for Category 2, 0.67 percent; and for Categories 4 and 5, 1.5 percent and 2 percent, respectively. In addition, a further, time-dependent premium component is also charged. The annual level of this component is 0.72 percent of the outstanding covered amount for Category 3, and this level is adjusted by the same factors as the basic premium for each other category.

In the case of private buyers an additional flat premium component is added, amounting to 35 basis points for the basic premium portion and an additional 25 basis points per annum. This surcharge is reduced if a commercial bank guarantee is available,2 or an international financial institution is involved, and in the case of cofinancing with bilateral aid.

1Described in AGA Report (a newsletter issued by Hermes), No. 49 (April 1994).2From a bank acceptable to Hermes as the sole debtor for the amount involved.

Algeria

Algeria had long been a major market for export credit agencies for both consumer goods and large-scale investments in the hydrocarbon sector. Algeria’s centralized import allocation system facilitated a heavy dependence on export credits, and it managed to remain current despite comparatively high debt-service ratios, largely arising from heavy amortization payments on export credits. However, the debt-service situation, together with increasing doubts about the sustainability of Algeria’s policy stance, led agencies to shift cover to exports with shorter maturities.31 New commitments declined after 1991, though some large agencies continued to make new commitments to Algeria notwithstanding the mounting economic and financial difficulties (Chart 11). Algeria’s debt-service obligations reached over 70 percent of exports of goods and services in 1992 and then rose still further with the fall in the oil price in 1993.32

Other Major Markets

Chart 12 shows developments in export credit agencies’ exposure to Indonesia, Turkey, and Venezuela. All have experienced increases in commitments, and, within this trend, an increase in short-term commitments. Based on the experience of export credit agencies with Iran, it might appear that a trend of rapidly rising export credit commitments together with a rising share of commitments on relatively short maturities is a strong leading indicator of future payments difficulties. However, this is not always the case, particularly for countries that have access to diversified sources of financing.33

Chart 12.
Chart 12.

Export Credit Exposure: Indonesia, Turkey, and Venezuela1

(In billions of U.S. dollars)

Sources: Berne Union; and IMF staff estimates.1For data definitions, see Chart 4.

More generally, cross-country comparisons of trends in export credits can be misleading. Indonesia. Turkey, and Venezuela all had similar profiles to that of Algeria in the early 1990s, but the economic situations of these three countries vary considerably. An increase in the proportion of short-term exposure can sometimes be a sign of improvement, as in the case of India (Chart 13), where the increase in short-term exposure of export credit agencies in 1992 reflected renewed commitments to India following the successful implementation of an adjustment program. Similarly, the recent increase in short-term commitments to the Philippines (Chart 13) took place in the context of a broadly satisfactory macroeconomic framework and also represents a shift of credits to the private sector. However, uncertainties about whether the Philippines would return to the Paris Club after a period of full debt-service payments may also have played a role in making agencies more cautious on the maturity profile.34

Chart 13.
Chart 13.

Export Credit Exposure: India, Philippines, and Nigeria1

(In billions of U.S. dollars)

Sources: Berne Union; and IMF staff estimates.1For data definitions, see Chart 4.

Thus, as illustrated by all of these cases, developments in overall export credit agency exposure and data on new commitments need to be interpreted with caution. Shifts in the volume of overall new commitments, or in the proportion of short-term exposure, may reflect shifts in the demand for export credits because of shifts in import demand in the borrowing country, shifts by exporters toward official support, or policy decisions by export credit agencies.35 Moreover, there may be changes in policies by export credit agencies that do not directly affect the level of new commitments but are nevertheless important, for example, increases in premia charged on new commitments. Finally, it should be noted that exposure of export credit agencies reflects actions taken in the past as well as actions taken in the present. A good example is the case of Nigeria, which is the eighth largest country in terms of export credit agency exposure, but for which new commitments recently have been minimal because of its poor performance (Chart 13).

Cover Policies for Eastern Europe and Countries of the Former Soviet Union

Tables 4 and 5 summarize the stance of cover policies on short-term and on medium- and long-term credits towards central and eastern European and selected countries of the former Soviet Union at end-1993. As noted in Chapter III, agencies have found it difficult to assess the creditworthiness of economies in transition. However, with the experience gained over the past few years, and in comparing transition economies, a hierarchy of creditworthiness as perceived by agencies has clearly emerged.

Table 4.

Summary of Short-Term Cover Policies of Export Credit Agencies Toward Selected Economies in Transition, End-1993

article image
Sources: Agencies visited; Berne Union; and IMF staff estimates.

One agency can have several types of restrictions.

Including cases when restriction is only on private sector; agencies whose stance is not yet determined but are open on a case-by-case basis are included here.

Including agencies off cover for new business.

Including guarantees of payment or transfer, commercial bank guarantees, or irrevocable letters of credit.

Including limits on individual transactions, on total commitments, and on annual new business.

Table 5.

Summary of Medium- and Long-Term Cover Policies of Export Credit Agencies Toward Selected Economies in Transition, End-1993

article image
Sources: Agencies visited; Berne Union; and IMF staff estimates.

One agency can have several types of restrictions.

Including cases when restriction is only on private sector or when agency is open for medium- but not for long-term cover; agencies whose stance is not yet determined but are open on a case-by-case basis are included here.

Including agencies off cover for new business.

Including guarantees of payment or transfer, commercial bank guarantees, or irrevocable letters of credit.

Including limits on individual transactions, on total commitments, and on annual new business.

Agencies view the Czech Republic and Hungary as better risks than most other countries: cover to these countries is basically demand determined. Most agencies are on cover for all maturities without significant restrictions in Poland as well, although certain agencies’ approach toward Poland is significantly more restrictive because of the debt and debt-service reduction agreement with the Paris Club in 1991. Slovenia is also considered as a good credit risk because of its good payments record.

The second tier of countries includes the Slovak Republic, the Baltic states, and Romania. In these countries, agencies have typically extended only relatively shorter-term credits and monitor payments records closely, but have a basically positive attitude toward providing new cover. Cover policy of most agencies toward Bulgaria is more restrictive due to its past delays in servicing post-cutoff date debt and in reaching bilateral agreements under Paris Club reschedulings. Croatia has been affected by war risk and uncertainties about its share of Yugoslav debt. Albania and the former Yugoslav Republic of Macedonia are considered to require mostly concessional financing rather than export credits on commercial terms.

The impact of these varying perceptions of creditworthiness on the part of export credit agencies can be seen in Chart 14, which shows export credit agencies’ exposure and new commitments to Hungary and Bulgaria. In the case of Hungary, which, as indicated above, has established a very good payments record, both total exposure and new commitments have risen steadily from 1988 to 1991; since then they have been broadly stable. In the case of Bulgaria, on the other hand, new commitments have declined since the late 1980s, although exposure has increased, due to a sharp increase in arrears and unrecovered claims.

Chart 14.
Chart 14.

Export Credit Exposure: Hungary and Bulgaria

(In billions of U.S. dollars)

Sources: Berne Union; and IMF staff estimates.1For data definitions, see Chart 4.

As noted earlier, Russia was viewed by most agencies as a unique case. Having assumed responsibility for the debt of the former Soviet Union, it already has substantial obligations to export credit agencies and its payments record has been inconsistent. This has made export credit agencies reluctant to extend new cover without security arrangements. On the other hand, governments have been conscious of the need to support Russia’s reform efforts and support through agencies has been their most important instrument for meeting this need. As a result, the largest agencies in particular have extended substantial new credits to Russia; such credits are sometimes treated explicitly in agencies’ accounts as national interest lending by governments (where such distinctions are made).

Chart 15 shows developments in the exposure of export credit agencies to Russia and the former Soviet Union. As Russia has taken on responsibility for the debt of the U.S.S.R., the chart shows the combined exposure.36 From 1992 onward, new export credit commitments were made to the Russian Federation, though disbursements continued to be made from commitments made prior to 1992 to the U.S.S.R. The overall exposure of export credit agencies has increased sharply since 1989, reflecting in large part the strong political support for financing. Since 1991, when arrears started to accumulate, followed by comprehensive reschedulings in 1993 and 1994, exposure in the form of unrecovered claims and arrears has also increased significantly, though only on commitments originally made to the U.S.S.R.: commitments to the Russian Federation have not been rescheduled. While new commitments have declined somewhat during the last two years, they remain substantial. The extent of export credit agencies’ support for Russia can also be seen in the change in their relative exposure to Russia compared with exposure to other countries. At the end of 1987, the U.S.S.R. ranked fourth in terms of export credit agency exposure, behind Brazil, Algeria, and Poland. At the end of 1993, export credit agency exposure to Russia was 50 percent higher than that to any other country, and amounted to over 13 percent of the total exposure of export credit agencies to non-OECD countries (Chart 2).

Chart 15.
Chart 15.

Export Credit Exposure: Russia and the former U.S.S.R.1

(In billions of U.S. dollars)

Sources: Berne Union; and IMF staff estimates.1For data definitions, see Chart 4.

In the countries of the former Soviet Union other than the Baltic states and Russia, agencies have adopted varied approaches. Some were open for cover in the three larger economies (Belarus, Kazakhstan, and Ukraine); some were closed for all except the Baltic states. Most agencies are cautiously opening cover for the resource-rich Asian states, Kazakhstan, Uzbekistan, and Turkmenistan, or are contemplating such a move. Ukraine. Belarus, and Moldova were considered more risky markets, and, despite their potential, most agencies were not on cover for them in the absence of a decisive improvement in the policy environment. The remaining states of the former Soviet Union are either considered by most agencies to be too poor to receive export credits on commercial terms (for example, the Kyrgyz Republic), or have war risk, precluding cover altogether (for example, Armenia, Azerbaijan, Georgia, and Tajikistan).

Developments and Prospects
  • View in gallery

    Cover Policy Indices1

  • View in gallery

    Total Exposure in Top Five Markets

    (In billions of U.S. dollars)

  • View in gallery

    Export Credit Exposure: China, Brazil, and Mexico1

    (In billions of U.S. dollars)

  • View in gallery

    Export Credit Exposure: Iran and Algeria1

    (In billions of U.S. dollars)

  • View in gallery

    Export Credit Exposure: Indonesia, Turkey, and Venezuela1

    (In billions of U.S. dollars)

  • View in gallery

    Export Credit Exposure: India, Philippines, and Nigeria1

    (In billions of U.S. dollars)

  • View in gallery

    Export Credit Exposure: Hungary and Bulgaria

    (In billions of U.S. dollars)

  • View in gallery

    Export Credit Exposure: Russia and the former U.S.S.R.1

    (In billions of U.S. dollars)