Skip to main content
A better test for sick banks

Now that poorly managed banks have been exposed as a central cause of the current Asian financial crisis, governments and investors are asking why the problems weren't spotted earlier. More specifically, they are taking a fresh look at techniques used to diagnose banks' health and anticipate problems. Such "early-warning" indicators are used by government regulators to prevent financial crises and by credit rating agencies to alert customers to the risks of investing in a particular country.

In both industrialized and developing nations, three key criteria have traditionally been used to measure banks' health: capital adequacy (measured as the risk-weighted capital-to-asset ratio), asset quality (an assessment of the quality of a bank's loans), and liquidity (a bank's holding of cash and ability to sell assets).

These criteria have proven quite reliable in assessing the mature financial systems of the industrialized world. But according to Liliana Rojas-Suárez, former IDB economist and present chief economist for Latin America and the Caribbean at Deutsche Morgan Grenfell, they may not be the best way to evaluate banks in developing countries. In a recent paper Rojas-Suárez shows that prior to the Mexican peso crisis of 1994 1995, customary "early-warning" criteria failed to predict almost all of the subsequent crises at individual banks. In fact, many of the banks that got into trouble actually satisfied some of these criteria.

Rojas-Suárez argues that these errors are largely due to the inadequacy of accounting standards and financial reporting requirements in many developing countries. These weaknesses allow banks to publish figures that give the appearance of greater financial health than may actually be true. The highly concentrated ownership of real and financial assets in many developing countries also subverts the ability of regulators to discern the true financial condition of banks, she says. That is because wealthy investors can shift assets and lend to each other through unregulated "nonfinancial" entities.

Rojas-Suárez proposes an alternative set of criteria based on the interest rates that banks in developing countries charge borrowers and offer depositors. According to her analysis, weak banks often try to gain market share by offering high rates to depositors and comparatively low rates to borrowers. She believes looking at the "spread" between these two rates can give regulators a better reading of a bank's health than traditional means. Rojas-Suárez says that this approach would have predicted between 70 and 80 incidents of bank problems during the Mexican crisis.

Jump back to top