Lending booms are the cornerstone of numerous recent theories on financial and banking crises.1 The precise origins of lending booms are diverse. They may arise following a possibly poorly regulated financial liberalization, a surge in capital inflow driven by external factors, or a terms-of-trade shock (or other types of supply shock) that boosts domestic investment or consumption or both. They may also come as a consequence of a macroeconomic stabilization program: it has long been noted that exchange rate-based stabilization programs are often associated with ultimately unsustainable booms in consumption, output, and credit.
During a lending boom, the typical story goes, credit to the private sector rises quickly. Leverage increases, and financing is extended to projects with low—possibly even negative—net present value, either because monitoring becomes more difficult when the volume of lending increases rapidly, which increases the likelihood of fraud (including looting, self-lending, and evergreening), or because domestic borrowers' net worth increases. As lending expands, the quality of funded projects goes from bad to worse, exposure increases, and the banking sector becomes more vulnerable.
Some scholars emphasize the aggravating effect of expected public bailouts in the event of a generalized bankruptcy. Bailout guaranties, whether explicit or implicit, induce private borrowers and lenders to develop [End Page 47] and carry over riskier projects than may be socially efficient. Entrepreneurs and lenders price new projects under the best possible scenario, taking into account the government intervention in the worst states of nature, and the quality of new loans worsens.2 The story usually ends in tears: the private sector gets scared or the projects fail to deliver, the bailout guaranties are called in, and the whole edifice comes tumbling down.
Others focus on the importance of the credit channel or financial accelerator.3 The mechanics are relatively straightforward: during a boom, asset prices increase, which increases borrowers' net worth, facilitates new lending, fuels higher asset demand and even higher asset prices, and so on. During the bust, the opposite happens: a proportion of actors are not able to repay their loans, and banks call in the collateral at firesale prices. The banks become more vulnerable as the asset side of their balance sheet shrinks. New loans are curtailed, and investment collapses together with asset prices. As a result of this increased vulnerability, a mild correction in asset prices may trigger a full-blown banking crisis.4
Consumption booms serve as the basis for many other explanations of the boom-bust cycle. For instance, the cycle may stem from an unsustainable consumption boom rooted in a less-than-perfectly-credible exchange rate stabilization program. Calvo and Végh provide an extensive review of these theories.5
There is ample empirical evidence that credit overexpansion and banking crises are related. Demirgüç-Kunt and Detragiache, for instance, show that after controlling for the existence of deposit insurance, the ratio of private credit to gross domestic product (GDP) and the (lagged) real growth of private credit are significant determinants of banking crises.6 Honohan considers credit growth one of the leading variables for diagnosing and predicting banking crises.7
As scholars of the recent financial crises note, countries that rely on foreign capital inflows may experience a nastier variety of financial debacle that combines a banking crisis and a balance-of-payments collapse. Chile [End Page 48] in 1982, Argentina in 1979, Mexico in 1994, and Thailand in 1997 are notorious examples. Several studies examine how the fiscal burden of a banking crisis can generate a balance-of-payments crisis.8 Conversely, a weak financial sector may prevent financial authorities from defending their currency, effectively hastening its demise. Goldfajn and Valdés, as well as Goldstein, study the direct link between an intermediation boom and the likelihood of banking and balance-of-payments crises occurring as a result of capital flows.9 Kaminsky, Lizondo, and Reinhart report that five out of seven studies analyzing credit growth as a determinant of currency crises find statistically significant results.10 In their own currency-crisis warning system, these authors consider that the...