In emerging economies, periods of rapid growth (booms) and large capital inflows can be followed by sudden stops and financial crises (busts). Recoveries can feature low GDP growth and aggregate credit. Research summarized here argues that, at a theoretical level, this pattern can be reproduced with a simple modification of a neo-classical growth model, in the presence of financial markets imperfections giving rise to endogenous borrowing constraints.
Economists colloquially use the expression “boom-bust cycle” to denote a prolonged surge in the GDP growth, followed by a sudden and sharp recession. Boom-bust cycles in emerging economies have received particular attention in the literature. In these economies, high growth can be followed by sudden stops in capital inflows and financial crises. After the bust, the GDP growth and investment rates can remain depressed for a long time (IMF, 2009), and in some instances, they may remain permanently below the pre-crisis level.
Often, during boom-bust episodes, the path of credit expansion and sudden contraction mirrors the cycle in the GDP growth, as shown for instance in Mendoza and Terrones (2008). Moreover, recoveries from the crisis can be “creditless,” meaning that they are characterized by a particularly weak growth in aggregate credit. Te parallels between the evolution of GDP growth and of credit aggregates, as documented in Claessens, Kose, and Terrones (2009), have sparked a vivid line of empirical research. Researchers have strived to identify the source of the positive correlation between GDP and credit dynamics during booms, busts, and “creditless” recoveries.
Is the correlation because credit is not needed when growth is low, or is it because there are times when credit is needed but the financial sector cannot provide it, and this depresses economic activity? Te latter situation would point to imperfections and inefficiencies in the financial market as a source of aggregate fluctuations in GDP growth rates. For instance, if the banking sector were to be suddenly impaired in its ability to intermediate funds, firms would be forced to curtail their investment plans and GDP would fall. A similar argument would apply if a sudden fall in the value of firms’ collateral would constrain firms’ ability to issue new debt.
Does the empirical evidence support the view that credit constraints arising from financial imperfections can account at least in part for GDP dynamics during booms, busts, and recoveries? Even though this question is still very much open, there is suggestive evidence that the answer might be positive, at least in situations when financial markets impairments are due to banking crises, as studied by Abiad, Dell’Ariccia, and Li (2011), and Dell’Ariccia, Detragiache, and Rajan (2007).
Te theoretical literature has long recognized that financial markets imperfection, acting through borrowing constraints, can be a source of aggregate fluctuations. For instance, Mendoza (2008) shows how borrowing constraints scan amplify fundamental shocks to a small open economy and generate deep recessions and credit contractions. In this literature, the fundamental shock to the economy is usually a total factor productivity (TFP) shock. In Piazza (2010) I take a different theoretical approach. In particular, I construct a neoclassical growth model for a small open economy and show how a growth process characterized by a boom-bust cycle can be caused by financial imperfections even in the absence of TFP shocks. Te model is based on two cornerstones, namely, uncertainty on the path of (decreasing) marginal returns to capital and a financial market imperfection.
To fix the ideas on the first cornerstone, it is useful to start from a simple example. Consider a closed economy inhabited by two types of agents: entrepreneurs and workers. Entrepreneurs operate in the industrial sector, and produce according to a neoclassical and constant returns to scale production function with capital and labor as inputs. Workers can live in the countryside, where they are farmers and gain a (low) constant subsistence wage, or they can move to the city and be employed in the entrepreneurial industrial sector. It is clear that, as long as a strictly positive amount of farmers remains in the countryside, the equilibrium wage in the industrial sector must be constant, and equal to the subsistence wage of farmers. Te marginal return to capital in the industrial sector is then constant too. Instead, when the pool of farmers that can be attracted to the industrial sector is exhausted, the scarcity of workers will raise the industrial wage above the subsistence level, thus lowering the equilibrium marginal return to capital.
“Does the empirical evidence support the view that credit constraints arising from financial imperfections can account at least in part for GDP dynamics during booms, busts, and recoveries? Even though this question is still very much open, there is suggestive evidence that the answer might be positive.”
The model described above has the following simple equilibrium property: marginal returns to capital in the industrial sector are initially constant, and start to decrease only when the economy has reached a certain “turning point,” identified as the moment when the pool of farmers in the countryside is exhausted. If we assume that the size of the pool of farmers is not known with certainty, then the timing of the turning point becomes uncertain too. Clearly, the arrival of the random turning point brings “bad news” to the industrial sector, since it signals that marginal returns are starting to decrease. By making proper assumptions on the production function, we can easily construct cases where marginal returns to capital decrease at an arbitrarily small pace after the turning point. In these cases, we can say that the turning point brings arbitrarily “small” negative news to the economy.
Te second cornerstone in my analysis is a financial market imperfection. I assume that entrepreneurs in the urban industrial sector can accumulate capital either by retained earnings, or by borrowing from international investors. Debt contracts suffer from limited enforceability, so that entrepreneurs can choose to renege on their obligations and default on their firm’s foreign debt. Individual default is associated with a punishment, in the form of a temporary productivity loss and restriction from further accessing the financial market. Naturally, a firm which is large, as measured by the size of its installed capital, suffers a large total output loss from the punishment of a reduced productivity. A more painful punishment for default implies that a large firm can commit to higher levels of debt before entrepreneurial default incentives are triggered. A larger installed capital acts then as a commitment device, or “collateral,” that allows entrepreneurs to relax their endogenous borrowing constraints and increase their debt.
Te core result of Piazza (2010) is to show that arbitrarily “small” negative news, revealed at the turning point, concerning the decreasing path of marginal returns can cause booms and busts in credit and GDP growth. In particular, before the turning point, endogenous borrowing constraints are large, growth is high, and the economy is booming. At the turning point, borrowing constraints are suddenly tightened, creating a credit crunch and forcing individual borrowers to default. Productivity costs associated with default cause a bust. After the turning point, borrowing constraints are permanently tighter, generating a recovery with low growth and low aggregate credit. Te linchpin for this result is what I call the equilibrium self-reinforcing property of growth and borrowing constraints: high growth endogenously relaxes borrowing constraints, which in turn foster higher growth and even larger borrowing constraints.
The reason for this property is quite intuitive. If the economy is growing fast, entrepreneurs’ capital stock (the “collateral”) is growing fast too. Hence, entrepreneurs’ borrowing constraints are expanding rapidly. With a rapidly expanding debt, entrepreneurs not only are able to maintain high investment rates, but can also keep rolling over most of their debt obligations. Since old debt is mostly repaid by issuing new, and thus without having to reduce consumption, entrepreneurs have little incentives to default. This allows lenders to further relax borrowing constraints on entrepreneurs, who can then increase investment, boosting the growth rate of capital and of the economy even more.
The conclusion of Piazza (2010) is that, because of the self-reinforcing property, endogenous borrowing constraints turn out to be extremely sensitive to news about growth prospects which, within the neoclassical growth framework I adopt, are linked to news on the path of marginal returns to capital. The sensitivity is so extreme that even arbitrary “small” negative news, revealed at the turning point, can be greatly amplified by the self-reinforcing property, leading to a sudden and permanent credit crunch and thus to boom-busts cycles in credit and GDP growth.
AbiadAbdulGiovanniDell’Ariccia and BinLi2011 “Creditless Recoveries” IMF Working Paper 11/58 (Washington: International Monetary Fund).
ClaessensStijn M. AyhanKose and Marco E.Terrones2009 “What Happens during Recessions, Crunches and Busts?” Economic Policy Vol. 60 (October) pp. 653–700.
Dell’AricciaGiovanniEnricaDetragiache and RaghuramRajan2008 “The Real Effect of Banking Crises” Journal of Financial Intermediation Vol. 17 (January) pp. 89–112.
International Monetary Fund2009 “Chapter 4. What’s the Damage? Medium-Term Output Dynamics after Financial Crises” in World Economic Outlook October 2009: Sustaining the Recovery (Washington: International Monetary Fund).
MendozaEnrique2008 “Sudden Stops, Financial Crises and Leverage: A Fisherian Deflation of Tobin’s Q” NBER Working Paper No. 14444 (Cambridge, MA: National Bureau of Economic Research).
MendozaEnrique and MarcoTerrones2008 “An Anatomy of Credit Booms: Evidence from Macro Aggregates and Micro Data” IMF Working Paper 08/226(Washington: International Monetary Fund).