Chapter 8. Recent Argentine and Brazilian Crises and Effective Multilateral Development Bank Interventions

Jian-Ye Wang, and Márcio Ronci
Published Date:
February 2006
  • ShareShare
Show Summary Details
Xavier A. Jordan44

Like several other recent emerging-market financial crises, the crisis in Argentina starting in late 2001 as well as in Brazil from late 2002 until early 2003 were accompanied by major cutbacks in pre- and postshipment, export-oriented trade finance extended by international lenders to domestic borrowers.

Heightened global risk aversion, concerns over public sector debt dynamics and private sector international debt amortizations, as well as political uncertainties due to upcoming presidential elections all played a key role in the significant withdrawal of international trade credit to Brazilian borrowers from late 2002 until early 2003. Nonetheless, Brazil’s trade finance cutback was somewhat surprising for four reasons. First, it was greater than trade line curtailments in prior, and arguably more severe, Brazilian financial crises, such as the debt restructuring episodes of the 1980s, the Tequila crisis spillover in 1994–95, the Asian crisis in 1998, and the shock from the devaluation of the real in 1999. Second, a substantial international support package to address the financial crisis was rapidly put in place on the basis of Brazil’s reasonably credible macroeconomic policies. The foreign exchange resources made available by the IMF program were more than sufficient to cover debt service obligations of the sovereign well into the new administration’s first year in office. Third, the country’s domestic financial system was on a sound footing, and was not experiencing any systemic distress. Fourth, trade finance reductions in Brazil occurred in spite of the fact that the pre- and postshipment export finance facilities being withdrawn were usually short term and self-liquidating in nature. In other words, they funded transactions that generated foreign exchange that could be used to repay the relevant trade finance operations, and therefore had considerably less performance as well as credit risk than other types of Brazilian cross-border credit exposure.

There are several possible explanations for this relatively unprecedented cutback in trade finance during the recent Brazilian crisis. Many of the conventional causes of financial crises in emerging markets (public sector deficits, uncoordinated monetary policy, dysfunctional financial sectors, etc.) either did not exist or were being adequately addressed in Brazil in mid-2002, especially after the country’s new $30 billion IMF program was announced in August. Nevertheless, trade finance lines continued to be withdrawn by international banks. Brazil’s experience in mid- to late 2002, therefore, illustrates how international trade finance cutbacks can continue, even after implementation of an IMF or similar international financial support package. Moreover, it starkly shows how such trade finance reductions—particularly in an environment of volatile and skittish markets—can play a central role in exacerbating macroeconomic instability already being experienced by a country during a financial crisis.

The Brazilian crisis also highlights the importance of at least one form of targeted intervention by multilateral development banks (MDBs) that could be undertaken more systematically in similar situations going forward to address the vulnerability of emerging markets to sharp cutbacks in international trade finance during financial crises. However, as discussed below, the intervention that proved particularly successful in Brazil during the period from late 2002 to early 2003 would not in all likelihood have been possible without a well-functioning and solvent domestic banking system—in other words, one that could continue to credibly intermediate international trade finance flows—as well as a sovereign that had not defaulted on its international debt obligations.

On the other hand, the Argentine situation offers some lessons on alternative actions as well as transactional structures that could be taken or encouraged by MDBs to help restore trade finance liquidity during a financial crisis when a country’s domestic banking system has stopped playing its role as an intermediary for domestic as well as international credit flows, and if a sovereign has defaulted on its international debt obligations. In this alternative scenario, rather than mobilizing international credit flows through B loan/preferred creditor status transactional structures, MDBs can help channel locally-sourced liquidity to exporters by supporting securitizations of trade finance assets as well as placement of the resultant debt instruments with domestic institutional investors.45

The Brazilian Situation in Late 2002 and Early 2003

Unlike past financial crises, when international trade finance lines were maintained in spite of cuts in other categories of cross-border debt extended to Brazil, there was a significant decline in this type of credit exposure to the country starting in mid-2002 (Figure 8.1). Banco Central do Brasil has estimated that trade finance lines declined by approximately 30 to 40 percent in 2002 from their levels in 2001, with foreign banks’ credit lines to Brazilian banks (a major part of which are comprised of trade finance exposure) declining from a high point of $22 billion in late 2001 to $16 billion in October 2002. The impact on exporters of this decline in trade finance was magnified by a shortening of tenors on trade facilities left outstanding by international banks, with offered maturities falling from 360 days to as short as 30 days. Spreads charged on lines left in place, if at all, for high-quality borrowers also increased from their conventional levels to more than 6 percent per annum over LIBOR during July-October 2002. The cutback in trade finance was accompanied by a withdrawal of incremental credit insurance and monoline wrap capacity for holders of Brazilian financial assets, including trade finance. This occurred because the international insurance and reinsurance industry had already “maxed out” on Brazil, and could not exceed country limits on Brazilian exposure as a result of its perceived financial crisis. Therefore, insurers would not make new cover capacity available to potential credit risk hedgers.46

Figure 8.1.Brazil: Trade Finance Flows, 2001-02

(In billions of U.S. dollars)

Source: Central Bank of Brazil

The shortage of trade finance not only inflicted a liquidity squeeze on the country’s export sector, but also impacted the pace of devaluation of the real and therefore conditions in Brazilian local financial markets. For instance, during the July-October 2002 period, the shortage fueled greater-than-normal corporate demand for U.S. dollars and put downward pressure on the real, as exporters were forced to repay lines that were not being rolled over. It also contributed to a decline in U.S. dollar spot trading activity to 10 percent of its normal volumes, which made the country’s foreign exchange markets less liquid—and therefore more volatile. Accelerating devaluation tendencies of the real, in turn, negatively affected investor perceptions of public sector debt dynamics, since a significant portion of domestic federal government securities in Brazil is indexed to the U. S. dollar.

In response to this situation and as a complement to its recently announced IMF program, the government initiated several measures in the third quarter of 2002 to provide temporary, short-tenor liquidity to trade finance markets. One example was a special provision of $560 million of 180-day funds for 330 export operations from the Fundo de Amparo do Trabalhador of country’s development bank (BNDES). As a result of increased reserve usage flexibility under its $30 billion IMF program, the Banco Central do Brasil also made available to domestic banks, through periodic commercial auctions, approximately $900 million of the $2 billion of available foreign exchange resources for 180-day trade finance onlending.

The above-mentioned government efforts to inject temporary, short-tenor liquidity into the country’s trade finance markets were complemented by an International Finance Corporation (IFC) initiative undertaken between September 2002 and March 2003 to rapidly provide international funding to leading Brazilian banks that are major players in the country’s trade finance sector. Before this intervention was concluded in March 2003 as a result of improved market sentiment and trade finance liquidity in Brazil, IFC trade finance facilities (TFFs) were arranged for four domestic private sector banks that are leading intermediaries in Brazil’s trade finance markets (Unibanco, Itau, BBA, and Bradesco). The four separate TFF A/B Loan (i.e., preferred creditor status) transactions undertaken for these banks raised an aggregate amount of $740 million of international funding for the country. This amount included $110 million of IFC 2-year A loans, a $50 million IDB 2-year A loan, and $580 million of 1-year B loan funding mobilized from 35 international commercial banks (domiciled in Asia, Europe, and the United States).

The B loan/preferred creditor status feature of the TFFs was fundamentally successful in addressing international banks’ political as well as transfer and convertibility risk concerns involving a relatively low-credit-risk asset class such as trade finance. In that regard, it helped to (i) convince international banks with existing credit-line authorizations to the relevant banks to re-extend trade funding that had been withdrawn; (ii) allow Brazilian banks to lock in 12-month funding from commercial sources that otherwise would not have been available, due to the above-mentioned shortening of tenors on trade finance lines left outstanding by international lenders; and (iii) mobilize several “new money” commitments from international banks that did not have existing credit line authorizations to the relevant Brazilian banks.

Since they were all immediately committed and disbursed, the TFFs made trade finance resources quickly available to a broad cross-section of Brazil’s exporter community, in terms of (i) the number of exporters that benefited from onlendings undertaken under the program by the four Brazilian banks; (ii) the amount of onlendings to such exporters; and the (iii) tenors provided to borrowers—which were considerably longer than maturities offered at the time by either government programs or commercial lenders under conventional bilateral lines. The relatively large size of the TFFs, alongside that of the government’s own trade finance interventions, had a noteworthy circuit-breaker effect on the negative exchange rate dynamics referenced above, especially given the extremely illiquid conditions and demand/supply imbalances in spot foreign exchange markets in Brazil at the time.

As risk aversion toward Brazil decreased by mid- to late 2003, normalcy returned to Brazilian financial markets, and conventional trade finance liquidity was restored to banks as well as exporters. All of the above-mentioned TFFs were prepaid by the four borrowing banks that had undertaken transactions under the program.

The Argentine Crisis

One immediate consequence of the Argentine crisis that started in late 2001 was the withdrawal of credit facilities by most international banks for any new financial operations involving the country, as well as the curtailment of existing credit exposure to Argentine borrowers. In this environment, even trade finance activities were not immune to the cutbacks by international lenders of credit being offered, both to local bank operations as well as domestic exporters, given the perception of heightened political as well as transfer and convertibility risk in Argentina as a result of its sovereign default.

An analysis of the effects of Argentina’s crisis on the country’s banking system is beyond the scope of this chapter. Suffice it to say, however, that the crisis led to a virtual collapse of financial and credit intermediation via the country’s banking system. Even by mid-2004, in spite of some positive developments in the banking sector resulting from stabilization of many economic variables and sustained deposit growth since mid-2002, Argentina’s banks had still not been able to resume normal operations. Banking sector normalization continues to be hindered by the government’s failure to define its stance on certain aspects of bank compensation for losses incurred as a result of various gaps and mismatches occasioned by governmental actions during the early phases of the crisis. And the current scenario is likely to push the banking system’s recovery and, therefore, its ability to play an active role in providing the credit necessary for sustainable growth in the real economy, out further into the future.

Moreover, many leading Argentina banks from early 2002 until mid-2004 were technically insolvent as a result of the crisis (in spite of the fact that most of them soon returned to a position of relative liquidity, or continued to be characterized by reasonably sound operating capacity). They also were in the process of restructuring their international liabilities. As a result of these two conditions, the de facto ability of the banks to act as conduits for trade finance lines would already have been diminished, even if international credit lines offered to them had not been cut as a result of the crisis.

Nevertheless, since mid-2002, as retail deposits returned, banks have started to accumulate significant amounts of domestically sourced Argentine peso and foreign currency (mostly U.S. dollar) liquidity. In spite of this trend, however, they have been slow to resume credit operations (i.e., onlend funds from their growing deposit base). New credit lines that have been made available recently by banks (for instance, peso-denominated, medium-term mortgages at fixed rates and some U.S. dollar-denominated trade finance facilities) remain small or limited, and do not yet indicate a clear trend of improving credit conditions. Instead, banks have remained focused on securing government compensation for losses arising from the crisis, and addressing continuing net worth concerns. The latter point is important, since about half of Argentine bank assets are public sector securities and compensation instruments yet to be issued, whose ultimate value is not certain (on the other hand, private sector nonperforming loans on their balance sheets are now below 3 percent of GDP). Moreover, bank balance sheets are structurally mismatched, mostly as a result of various government actions/regulations (e.g., asymmetric “pesification”). For all these reasons, in the absence of fundamental reform, the country’s banking sector as a whole can be deemed to have negative net worth, and its solvency will continue to be very much linked to the government’s own credit profile.

As mentioned above, most Argentine banks have recently started to provide some level of trade finance that is funded by U.S. dollar deposit inflows they have captured in the last 12 to 18 months. This has occasioned since 2003 a gradual reduction of U.S. dollar interest rates for trade finance that is offered to exporters. However, with lenders offering few international credit lines to Argentine banks, spreads on trade finance currently made available by local financial institutions to domestic customers remain extremely high by historical standards, basically because of the uncertain stability and short (rarely beyond 30-day) tenor of banks’ U.S. dollar, retail deposit funding base. Therefore, banks have been reluctant to offer funds to borrowers with longer tenors, and if they do so, such funding must necessarily be priced at extremely expensive rates (over 10 percent) to compensate for the banks’ potential asset/liability tenor mismatches, if borrowers actually take money at these rates at all. It is also important to note that large banks, which are primarily foreign-owned, are only lending to top-tier exporters in limited sectors, and that smaller banks have not traditionally been major trade finance players. Moreover, since the U.S. dollar deposit base of the latter is unlikely to grow beyond current levels, their role in providing incremental export finance going forward is likely to be limited.47

Argentina’s top 20 (so-called “first-tier”) exporters (of mostly grain and petroleum-based products) account for approximately 50 percent of the country’s export sales. With the exception of the period immediately after the onset of the crisis, these firms have had relatively little difficulty in obtaining pre- and postshipment export funding from the country’s traditional providers of trade finance (local and international commercial banks, as well as global commodity trading firms). However, second-tier exporters have not had such favorable access to trade finance since the onset of the country’s crisis in late 2001. This second tier of exporters has responded to cutbacks in trade funding from commercial banks by not offering deferred payment terms to purchasers of their products (which reduces their international competitiveness, compared with exporters of the same products in other countries that do so), or by self-financing—from their own net worth—export activity. The latter response, after initial adjustments in 2002 involving both their production cycles and working capital utilization profiles, has allowed them to take advantage of more favorable Argentine export prospects as a result of peso devaluation, in spite of the dearth of reliable and reasonably priced trade finance. However, it is not necessarily optimal from a long-term perspective, since the scarce equity resources of Argentine firms—which should normally be used to finance fixed assets and increases in productive capacity—are instead being used to fund a portion of day-to-day working capital needs.

The scarcity of reliable and reasonably priced trade finance for second-tier exporters has been largely due to the above-mentioned breakdown in conventional trade finance markets, and has occurred as a result of (i) the inability of Argentine banks (both locally-and foreign-owned) to obtain ample bilateral credit lines from international banks to fund the export activities of their clients; and (ii) the relatively unstable and short-tenor retail deposit funding base of banks, which forces them to price at expensive rates any loans that might be made with maturities longer than those of their deposit liabilities. Moreover, the uncertain availability of conventional trade finance credit facilities from banks has made companies that formerly used such funding facilities—but had them pulled after the advent of the crisis—reluctant to reaccess them even when these credit lines may now be made available, as they have little assurance of their medium- to longterm availability.

In the absence of sufficient, reasonably priced trade finance from conventional banking system conduits, exporters have at least one alternative funding source (aside from self-financing). This entails securitization of their pre- and postshipment trade finance receivables, and placement of the resultant securities with domestic institutional investors.

For several reasons, there are good prospects in the short to medium term for such trade finance securitization activity. As far as supply of such securities is concerned, second-tier exporters may have to turn with increasing frequency to nontraditional sources of funding, given ongoing difficulties in the country’s banking sector. Although this trend may be less pronounced now than it was in the immediate aftermath of the country’s crisis, second-tier companies are still not able to access on a reliable basis reasonably priced trade finance through conventional bank channels.

In terms of demand for securitized trade finance assets, for several reasons there is likely to be considerable appetite from domestic investors for this type of credit exposure. First, domestic institutional investors must hold the bulk of their investments/assets in the form of local securities/credit exposure. Second, domestic institutional investors, such as pension funds that hold increasing amounts of liquidity, are actively looking for new investment alternatives (to complement their large holdings of government securities), given the dearth of securities issued by the private sector in Argentina since mid-2001. Third, unlike Argentine banks during the country’s recent crisis, institutional investors such as pension funds have relatively long-tenor liabilities (i.e., the retirement savings of their contributors); thus, they are more willing to fund relatively long-tenor assets. Therefore, pension funds as well as other “buy-and-hold” institutional investors are likely to be more willing than bank depositors are with their short-term deposits to roll over securitized trade finance asset exposure, as they are actively searching for longer duration assets to match their long duration liabilities.

Such domestic institutional investor appetite for trade finance securitizations has already been demonstrated in oversubscriptions of trade finance securitization offerings brought to market so far in Argentina (for example, excess investor demand of approximately $30 million for a $20 million Export Securitization Program transaction undertaken in late 2003—the largest domestic Argentine securitization undertaken last year).48

In a country such as Argentina, where a sovereign default has occurred and the domestic banking system has essentially jack-knifed, it is probably going to be relatively difficult for MDBs—even with the use of B loan/preferred creditor status structures—to rapidly mobilize international funding from commercial banks to help address the country’s trade finance shortfalls. Thus, the approach used by the IFC during Brazil’s recent financial crisis would not likely be effective in a scenario like that of Argentina, where the state of the domestic banking system, since it is so closely associated with the credit of a defaulted sovereign, is perceived by international lenders to be unsound and not fundamentally creditworthy.

The ability of MDBs to commit additional credit to a country in crisis, however, could be effectively used to help mobilize domestic savings or liquidity from institutional investors to help fill trade finance shortfalls. In the context of securitization programs targeted toward domestic institutional investors of the kind described above, such credit commitments could take three forms. First, MDB credit lines could be provided to domestic financial entities to allow them to originate trade finance assets and warehouse them for eventual securitization and placement with domestic institutional investors. Second, MDB backstop facilities could be made available to domestic financial entities to provide assurance to local institutional investors that a particular securitization placement will not be undersubscribed. MDB firm-purchase commitments for a portion of the relevant securitization offerings could play an important role in keeping such undersubscription scenarios from occurring.49 Third, MDBs could offer partial credit guarantees for domestic securitizations of trade finance receivables.50


The comparative description in this chapter of the effects of Argentina’s and Brazil’s recent financial crises on trade finance availability highlights the importance of well-functioning domestic banking systems in maintaining trade finance flows to local exporters. Moreover, during an emerging-market financial crisis, when a country’s banking system is relatively sound/solvent and the sovereign has not defaulted on its international debt obligations, it is possible for targeted interventions by MDBs to play a role in mobilizing international trade finance liquidity from commercial lenders (and perhaps insurers) through B loan/preferred creditor status structures that address transfer and convertibility concerns of the latter.

However, when a sovereign has defaulted on its international debt obligations and a country’s domestic banking system is dysfunctional or insolvent, B loan/preferred creditor status interventions are not likely to be successful in mobilizing incremental trade finance liquidity from international markets for the emerging market in question.

In these alternative scenarios, nevertheless, there may still be scope for MDBs to play a role in supporting securitizations and domestic placements of trade finance receivables. While more time-consuming and costly to structure than plain B loan/preferred creditor status transactions,51 domestically-oriented trade finance securitizations as well as accompanying MDB interventions meant to support them could still play an important role in mobilizing alternative domestic sources of trade finance liquidity in emerging markets such as Argentina that are experiencing protracted financial crises.

    Other Resources Citing This Publication