Determinants of the Venezuelan banking crisis of the mid-1990s: an event history analysis.

Extract

---

Determinants of the Venezuelan banking crisis of the mid-1990s: an event history analysis.

Resumen: Este trabajo utiliza un análisis de historia de eventos para evaluar empíricamente qué variables macroeconómicas y bancarias pueden explicar las quiebras bancarias que se produjeron durante la crisis de la segunda mitad de la década de 1990 en Venezuela. Una baja rentabilidad, medida por el margen financiero neto, y el escaso crecimiento económico resultan significativos en aumentar la probabilidad de quiebra bancaria. Otros indicadores útiles, para algunas especificaciones del modelo, son el porcentaje de activos impagados y el de activos improductivos respecto a los recursos propios, que aumentan la probabilidad de crisis. Por el contrario, una proporción elevada de activos líquidos reduce la probabilidad de quiebra en algunos casos, mientras que tipos de interés altos sobre depósitos la aumentan. Aunque este último resultado podría hacer pensar que la política monetaria en Venezuela fue excesivamente restrictiva antes de la crisis, no existe evidencia que lo confirme en el caso de otras variables clave como los tipos de interés reales sobre préstamos y, más aún, el crecimiento de la masa monetaria en términos reales.

Palabras clave: Venezuela, crisis bancarias, indicadores adelantados.

Abstract: This paper uses event history analysis to test the significance of several macro-economic and bank-specific variables in explaining bank failures during the Venezuelan banking crisis of the mid-1990s. Poor bank profitability, proxied by a low net interest margin, and low GDP growth are found significant in increasing the probability of bank failure. Other useful indicators, for some model specifications, are the share of non-performing loans and that of non productive assets to banks' own funds, which raise the likelihood of crisis. A large amount of bank liquid assets, in turn, reduces the likelihood of failure for some model specifications. The opposite is true for high real deposit rates. Although it could be interpreted, at first sight, as a too restrictive monetary policy, this is not supported by the lack of significance of the real lending rate and, even more so, real money growth, a more direct indicator of the monetary policy stance.

Keywords: Venezuela, banking crisis, early indicators.

Introduction

Recurrent events of banking crises around the world have generated an extensive literature from causation to prevention. However, there is still no consensus on which are the main determinants of bank failure, so that an early warning systems constructed to prevent such failures, or at least minimize their impact. This explains the efforts made by the International Monetary Fund and national authorities from different countries to identify a set of financial soundness indicators to be used as a surveillance tool of potential crises (Sundararajan et al., 2002). However, the validity of a specific set of indicators is still not firmly confirmed by empirical analysis, particularly for emerging countries.

In the mid-90s, Venezuela experienced a systemic banking crisis, with a large amount of bank failures and a fiscal cost estimated at 17% of the GDP. Notwithstanding the severity of the crisis, there is no empirical analysis of the determinants of these bank failures. Such analysis could contribute to building early indicators of bank unsoundness for the Venezuelan banking system, particularly if they do not require much calculation and can be easily tracked.

The most obvious indicators, and also those analyzed first in the literature, are bank specific variables from bank balance sheets or financial statements. More recently, macroeconomic indicators have received large attention, especially in the case of emerging countries. This is because macroeconomic imbalances are generally larger and more frequent in these countries and also tend to have bigger consequences, financial systems being shallower. However, macroeconomic variables alone can hardly explain the failure of a particular institution, which limits policy conclusions to the aggregate level. It seems important, therefore, to include both bank-specific and macroeconomic variables as potential determinants of banking crisis.

Th...

See the full content of this document

---