Ebook Macroeconomic Risk and Banking Crises in Emerging Market Countries: Business Fluctuations with Financial Crashes
Banking crises are not a strange episode in market economies. Indeed, they have been recurrent in emerging market countries during the last 25 years (see Lindgren et al., 1996; Caprio and Klingebiel, 1997, 1999; and Bordo et al., 2001). As long as banks are major players in modern economies, a banking crisis sparks off multiple adverse consequences including output losses, monetary instability, and other nonmonetary effects associated with information losses. The sequels of widespread banking failures are worse in developing countries because their banks own a larger share of total financial assets than the share owned by banks in industrial countries. Furthermore, the costs of restructuring a bankrupt banking system are inversely related to the degree of poverty of the nations and can be as large as one fourth of the GDP (Goldstein and Turner, 1996; Hoggart et al., 2001).
Notwithstanding being costly, to some extent crises are inevitable because every financial system is risky. An examination of the way banks do business reveals that under a reason ably high level of leverage, banks cannot guarantee the claims of investors if large adverse macroeconomic shocks occur (Dewatripont and Tirole, 1994, chap. 2). The value of the banks’ portfolios may fall because of a weak economic performance of the banks’ borrowers, specially when some credit risk cannot be diversified. A slowdown of the economy bankrupts a higher proportion of borrowers compared to “normal times”, and it is conceivable that the downturn will be severe enough that the banks themselves are in distress and cannot fully repay their creditors. This reasoning along with the greater volatility of emerging market economies can account for the higher vulnerability of these economies to waves of banking failures.
The causality between macroeconomic conditions and financial instability also goes the other way around because declines in the value of banks’ portfolios can weaken the economy. In a world of forward looking economic agents where everybody incorporates the eventual negative effects of a financial crash into his economic decisions, not only financial crises, but also their likelihood of occurrence affects the economy. Consider for example the effect of a negative term of trade shock; bank creditors, will notice that across the board firm bankrupt cies are more likely to happen than before and that this will ultimately raise the riskiness of banking business. Bank creditors will take portfolio decisions and will demand a country-risk premium that are consistent with the new degree of financial fragility of the economy. The country-risk premium adds to the interest rate on bank loans to firms, and from this perspective, the riskier financial system depresses the net rate of return of the productive activities. This pushes the demand for inputs down which in turns lowers the economy’s employment and income. In sum, on the one hand macroeconomic conditions are an important determinant of the soundness of the financial system and on the other hand the degree of financial fragility pervades all sectors of the economy.
This study builds a business-cycle model of a SOE where aggregate risk produces sporadic bank failures in a world where banks intermediate inflows of capital. Banks issue debt to international investors, fund a large number of firms, and are subject to a capital-adequacy regulation. Firms borrow from banks because they have to pay factors of production before realizing their sales proceeds. As project risk is not a completely diversifiable risk, economic downturns trigger a large ratio of poor project returns depreciating the value of the banks’ portfolios and, under some circumstances, recessions are severe enough that they produce the insolvency of the banking system. In this sense, crises are driven by fundamentals and not by a change of expectations. Furthermore, the probability of widespread banking failures is known in every period and it is part of the information set of all agents, making economic decisions dependent on the likelihood of a financial crash due to the eventual crisis sequels.
The paper abstracts from monetary and liquidity aspects of financial intermediation and stresses the aggregate risk involved in the credit creation process in a SOE. Panics, understood as circumstances where depositors attempt to withdraw their funds simultaneously, play no role in the model as deposits are represented by single-period contracts. A banking crisis is defined here as the inability of the banks to honor their debts.
Exogenous capital requirements on banking impose a risk on bankers who can lose part of their wealth during a crisis. International investors face a similar risk insofar as a financial crash affects the repayment capacity of banks. As long as household income depends on the state of financial markets and is affected by bank failures, the household sector is not immune from financial turmoils either.
The financial frictions in the economy that justify the existence of banks arise from a combination of two related elements. One is the presence of information asymmetries about both firms’ output and banks’ portfolio returns; the other is the existence of agency costs coming from the employment of a costly-state-verification technology. In this context, financial intermediation resembles delegated-monitoring banking because from time to time international investors have to ‘monitor a monitor’. Following Krasa and Villamil (1992a,b), a two-sided simple-debt contract is the optimal investment mechanism employed to solve the asymmetric information problem.
The contribution of the paper to the literature consists in addressing the following points using a computable general equilibrium framework. First, macroeconomic conditions and not self-fulfilling expectations trigger banking crises. Second, because the likelihood of a crisis is incorporated into the decisions of every economic agent, both macroeconomic risk and financial fragility affect business cycles. Third, the country-specific interest rate is endogenous because international investors evaluate the fundamentals behind their borrowers’ ability to repay.
Following Diamond and Dybvig (1983), most macroeconomic models dealing with banking crises in developing countries emphasize the role of sudden expectation changes in setting off bank runs. However, the empirical evidence on the determinants of banking crises seems to indicate that pure self-fulfilling expectations are one but not the most important determinant of the crashes. For instance, Kaminsky and Reinhart (1999) point out that “in both [currency and banking] crises we find a multitude of weak and deteriorating economic fundamentals that it would be difficult to characterize them as a self-fulfilling crises.” The two prominent fundamentals that would explain the crises are a weak macroeconomic performance and rising interest rates.
The literature on the effects of credit frictions on macroeconomic fluctuations that emphasizes agency costs follows Bernanke and Gertler (1989). Within the real business cycle paradigm, Fuerst (1995) and Carlstrom and Fuerst (1997) incorporate agency costs into the standard model by assuming informational asymmetries in the production of capital goods. Bernanke et al. (1999) synthesize this literature and offer a detailed analysis of the so called “financial accelerator”. Both Carlstrom and Fuerst (1998) and Cooley and Nam (1998) compare the consequences of assuming that financial frictions arise in the production of capital goods, vis-a-vis the case where the frictions arise in the production of final output. Despite dealing with credit and financial issues, this literature has not addressed the role of either bank credit risk or aggregate risk. In Bernanke et al. (1999) banks never default because borrowers offer a state-contingent non-default payment that guarantees the lender a return equal in expected value to the riskless rate. Cooley and Nam (1998) and Carlstrom and Fuerst (1998) take an alternative way out of aggregate risk by assuming that debt contracts are signed after the realization of the stochastic component that explains the macroeconomic uncertainty.
Out of the business-cycle paradigm, several works have studied the relationship between aggregate risk and banking crises by extending the Diamond and Dybvig (1983)’s framework (see Allen and Gale, 1998 and the papers cited therein). While this literature shows how business-cycle risk can lie at the root of run-driven banking crises, it is silent about the specific mechanism through which economic fluctuations impact on the bank-portfolio return. Similarly, explaining the way through which the financial fragility exerts influence on business cycles is not the aim of these works either.
Consistent with the empirical evidence, both declining aggregate productivity and rising interest rates are capable of bringing about a banking crisis in the model of this paper. Particularly, a calibration of the model to emerging market countries shows that the probability of a banking crisis goes from 0.5% during a boom to 1% during a deep recession. However, not every recession, no matter how deep, causes a crisis; it certainly does when the down turn is both deep and unexpected. Other results show that the interest rate on the debt issued by banks in international markets as well as the interest rate on debt issued by firms are counter-cyclical and responsive to aggregate risk. Once a crisis is in place, its ability to produce long-lasting effects depends on the existence of explicit crisis-related costs.
The paper proceeds as follows. The next section describes the model and explains how the probability of a bank failure depends on fundamentals. It also characterizes the financial contracting technology. Section 3 discusses static general equilibrium results, and section 4 discusses the dynamic effects of banking crises and aggregate risk. Section 5 contains concluding remarks and avenues to be explored in future research.
Posted in :
