The Financial Crisis and Foreign Bank Participation in Developing Countries

John Bonin, Wesleyan University, US
Robert Cull, DEC-FP
Violeta Gutkowski, Constultant, Argentina
Maria Soledad Martinez Peria, DEC-FP
Natalia Teplitz, Consultant, Argentina

Research Objective

1. Summary of Aims, Objectives and Significant Achievements

Objective: 

The objective of this research was to study how the recent 2007-2008 crisis affected the behavior of foreign bank lending to developing countries. We also sought to examine how the behavior of foreign bank lending compared to that of other banks in developing countries.

Significant achievements: 

- Put together a comprehensive country-level bilateral dataset (from lender country to borrower country) on foreign bank claims across developing countries before and during the 2007-2008 crisis.
- Wrote a paper quantifying the importance, composition, and growth of foreign bank claims across developing countries before the crisis and investigating the changes in foreign bank claims during and immediately after the crisis.
- Assembled a bank-level dataset on bank lending, bank ownership, and bank performance in Latin America and Eastern Europe.
- In the process of writing a paper exploring the behavior of bank lending in Latin America and Eastern Europe during the crisis, in particular, exploring the role of bank ownership.

2. Nominated Outputs

• Paper quantifying the importance, composition, and growth of foreign bank claims across developing countries before the crisis and investigating the changes in foreign bank claims during and immediately after the crisis. Completed.
  
• Paper exploring the behavior of bank lending in Latin America and Eastern Europe during the crisis. In particular, the paper compares lending by foreign banks, private domestic banks, and government owned banks.
  

3. Major Difficulties

Some of the Bank for International Settlements (BIS) data we wanted to use in the first paper were confidential and there were delays in getting the approval of some central banks in order for the BIS to release the data to us. Luckily, most of the data we needed was eventually made available to us.

Putting together the bank ownership data was also quite cumbersome and took longer than we expected since the information in Bankscope was often incomplete. In order to put together the ownership database, in many instances we had to complement Bankscope data with data from other sources such as Banker’s Almanac, central bank and individual bank websites.

4. Surprises

None.

5. Full Completion Report

• Background and motivation

Since the 1990s, many developing countries witnessed a sharp increase in the participation (direct or cross-border) of foreign banks in their financial sector. In certain regions like Latin America and Eastern Europe, foreign banks came to dominate the local banking sector. Most of these banks came from developed countries that have recently undergone a deep financial crisis.

• Original objectives of the research, along with any changes and their justification; ; note their consistency with Bank, DECRG, and KCP objectives

The objective of this research was to study how the recent 2007-2008 crisis affected the behavior of foreign bank lending to developing countries. We also were interested in understanding how lending by other banks was affected in countries with high foreign bank presence and compare the behavior of foreign bank lending to that by domestic-private and government-owned banks.

• Assessment of the extent to which the objectives have been met, with explanations of any shortfall.

Overall, we have met our objectives. We have completed a first paper describing the importance, composition, and growth of foreign bank claims across developing countries before the crisis and investigating the changes in foreign bank claims during and immediately after the crisis. Also, we are in the process of completing a second paper comparing the behavior of foreign, domestic and government-owned bank lending during the crisis in Latin America and Eastern Europe, two of the regions with some of the highest levels of foreign bank participation. 

• Methodology and data

The first paper is largely descriptive: we gather data from the BIS on total, cross border and local claims to developing countries from the main lender countries and we describe the importance, composition, and growth of foreign bank claims across developing countries before, during and after the crisis. We use data from the BIS. We also conduct some econometric estimations to try to explain differences in the growth of foreign bank lending across countries. In particular, we explore the role of lender liquidity, the impact of the extent of local presence, and of the importance of lending in local currency.
In the second paper, we assemble bank-level data on bank lending, bank ownership and bank performance in Latin America and Eastern Europe and we run regressions to examine the behavior of foreign relative to domestic and government-owned banks during the crisis. We use data largely obtained from Bankscope.

• Results: What have we learnt so far?

From the first paper, we find that, during the period 2005-2008, total foreign claims to developing countries grew by 81 percent in real terms and the growth rates for cross-border and local claims were very similar. However, there were sizeable differences in the growth rate of foreign bank claims across regions. Foreign bank claims growth was fastest in South Asia and Eastern Europe, where they grew by 131 and 110 percent, respectively. In contrast, foreign claims grew by 80 percent in East Asia, 50 percent in Latin America, 36 percent in the Middle East, and 78 percent in Sub-Saharan Africa.

The BIS data also make it clear that foreign bank lending to developing countries was curtailed as a result of the recent global financial turmoil. Overall, foreign claims fell by almost 5 percent between June 2008 and March 2009. But, there were notable differences in the patterns for cross-border and local claims: while cross-border claims fell by 9 percent in real terms, local claims in all currencies fell by only about 1.5 percent. Across regions, we also observe large differences in the size of the contraction in claims during the crisis: claims fell by approximately 16 percent in the case of East Asia, 8 percent in Eastern Europe, 9 percent in the Middle East and North Africa, 3 percent in Sub-Saharan Africa and South Asia, and, finally, by only 1 percent in Latin America.

The aftermath of the crisis was also very different across types of claims and regions. Overall, total foreign claims grew by 1.7 percent in the aftermath of the crisis through December 2009. The recovery in the growth of claims was driven by the increase in local claims, which grew by 2.3 percent, while cross border claims grew by only 0.8 percent. Across regions, East Asia exhibited a significant recovery with claims growing by almost 14 percent. Claims to South Asia grew by 8 percent and those to Latin America and the Middle East grew by almost 7 percent. On the other hand, claims to Eastern Europe and Africa continued to decline at 2 and 3 percent, respectively, in the aftermath of the crisis.

Overall, we conclude in the first paper that heavy reliance on direct, cross-border lending resulted in steeper contractions in total foreign claims during the crisis. In addition, contractions were more severe and recovery was slower in regions that relied heavily on lending in foreign currency.

We are still working on the second paper but so far we are finding that in Latin America foreign bank lending did not contract significantly during the crisis and government bank lending expanded. The main drop in lending in Latin America came from domestic banks. On the other hand, foreign bank lending collapsed during the crisis in Eastern Europe, while government bank lending did not change significantly. These results are still preliminary and more analysis is required.

• Dissemination details (including future plans)

We will disseminate our research in various ways: (1) we will put out our papers as World Bank Working Papers, (2) we will present them in academic and policy conferences, (3) we will submit them to journals, and (4) we also plan to produce blogs summarizing the main results from the studies. Finally, we expect that many of the findings of this research will also be featured in the upcoming Global Financial Development Report.

• Impact—within and outside the Bank

It is still early to measure the impact of our research but we expect that our findings will help inform bank clients about the behavior of foreign banks during crises and about the type of foreign bank engagement (local vs cross-border, foreign currency or local currency lending, etc) that seems less prone to retrenchment during periods of financial turmoil.

• Implications for future research

In the future, it will be interesting to research the extent that foreign bank lending recovers from the crisis and how it compares with lending by domestic banks.

• Capacity Building

This research will help the Bank understand the implications of foreign bank participation in the banking sector.

• Additional resources leveraged

John Bonin, an expert in banking in Eastern Europe, contributed his time to this project at no cost to the Bank.

• Operationalization and development policy implications

The findings from our research can help produce policy notes to guide Bank clients views about the implications of foreign bank participation in developing countries.

Annex 1: Reference list and background papers

Cull, R., Martinez Peria, M., and Teplitz, N. “Foreign Bank Claims to Developing Countries Before, During, and After the 2007-2008 Crisis”. World Bank. Mimeo.

Bonin, J., Cull, R., and Martinez Peria, M. “Bank credit and bank ownership during the 2007-2008 crisis.” World Bank. Mimeo