International Bank Lending to Developing Countries: An Empirical Analysis of the Impact of Country Risk

Article excerpt

In light of the resumption of international bank lending after the 1980's crisis and recent indicators of financial and economic instability, we estimate the impact of country risk on bank lending to thirty three developing countries. Based on a model of sovereign risk with cross section data for the period 1992-1994, the results indicate that transfer risk, defined as the ability of borrowers to honor foreign debt obligations, is a significant factor. By estimating directly the impact of country risk on bank loans, this model improves the earlier methodology in which external debt ratios have been used to estimate debt servicing capacity and the probability of default. The results have useful implications for strategies on the pricing of international loans and financial policies of debtor countries to maintain creditworthiness.


Over the past decade political changes resulting from the demise of communism and the implementation market oriented economic and financial reforms have resulted in a large increase in international trade and an enormous amount of external capital flows into the emerging markets of Eastern Europe, Latin America, Asia, and Africa. Most of these flows comprised international bank loans. Recent data indicated that lending by banks to the developing countries rose by $61 billion in the first half of 1996, with more than half of these new loans going to Asia. Three countries, South Korea, Thailand, and Indonesia accounted for $24 billion of the new loans. Euromoney (July 1997) reported that the past three-year boom in syndicated lending was due to intense competition among lenders and unusual advantageous conditions for borrowers. Pavey (1997) noted that the spread or margin over the London Inter Bank Offer Rate (LIBOR), the interest on Eurocurrency loan to non-bank borrowers, has fallen to a record low. The nations with significant bank exposure to Southeast Asia by mid 1997 were U.S.A. ($32.3 billion), Germany ($47.2 billion), France ($40.4 billion), and U.K. ($29.7 billion). Also, Japanese bank loans outstanding to Indonesia totaled $23 billion during this period.

It is imperative to investigate the determinants of this lending spree in light of the pattern of lending that partially precipitated the 1980s debt crisis. Since some debtor countries are still affected by the economic effects of the debt overhang problem and are still negotiating debt forgiveness and rescheduling then the determinants of the current lending behavior of banks warrants further investigation. With the present global financial market and economic instability further research on banks' exposure and default risk is also justified. The recent financial crisis in Southeast Asia and Russia with significant contagion effects further highlights the importance of research on assessing the impact of country risk on international lending. One of the lessons from the 1980 debt crisis was the necessity for bankers to implement more methodical procedures for assessing sovereign risk.

The earlier studies on international lending focused on problems dealing with the debt servicing capacity, the probability of default and rescheduling, and the optimal level of lending in terms of credit ceiling etc. External debt ratios and macroeconomic data as explanatory variables were essential in the different models, while the dependent variable depended on the statistical procedure employed. Frank and Cline (1971) used discriminant analysis to test eight external debt ratios, associated with debt servicing difficulties, on a binary value dependent variable that consists of rescheduling and non-rescheduling cases. This approach was improved by Feder and Just ( 1977) by the application of logit analysis. The application of logit and probit techniques has been popular in some recent studies, for example, Rivoli and Brewer ( 1997), and Da Haan, Siermann and Van Lubek (1997), combined political risk variables with economic variables to predict debt servicing difficulties. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.