Financial booms have become a chronic feature of the global financial system. When these booms end in crises, the impact on economic conditions can be severe. Carmen M. Reinhart and Kenneth S. Rogoff of Harvard pointed out that banking crises have been associated with deep downturns in output and employment, which is certainly consistent with the experience of the advanced economies in the aftermath of the global crisis. But the aftereffects of the booms may be even deeper and more long-lasting than thought.
Gary Gorton of Yale and Guillermo Ordoñez of the University of Pennsylvania have released a study of “good booms” and “bad booms,” where the latter end in a crisis and the former do not. In their model, all credit booms start with an increase in productivity that allows firms to finance projects using collateralized debt. During this initial period, lenders can assess the quality of the collateral, but are not likely to do so as the projects are productive. Over time, however, as more and more projects are financed, productivity falls as does the quality of the investment projects. Once the incentive to acquire information about the projects rises, lenders begin to examine the collateral that has been posted. Firms with inadequate collateral can no longer obtain financing, and the result is a crisis. But if new technology continues to improve, then there need not be a cutoff of credit, and the boom will end without a crisis. Their empirical analysis shows that credit booms are not uncommon, last ten years on average, and are less likely to end in a crisis when there is larger productivity growth during the boom.
Claudio Borio, Enisse Kharroubi, Christian Upper and Fabrizio Zampolli of the Bank for International Settlements also look at the dynamics of credit booms and productivity, with data from advanced economies over the period of 1979-2009. They find that credit booms induce a reallocation of labor towards sectors with lower productivity growth, particularly the construction sector. A financial crisis amplifies the negative impact of the previous misallocation on productivity. They conclude that the slow recovery from the global crisis may be due to the misallocation of resources that occurred before the crisis.
How do international capital flows fit into these accounts? Gianluca Benigno of the London School of Economics, Nathan Converse of the Federal Reserve Board and Luca Forno of Universitat Pompeu Fabra write about capital inflows and economic performance. They identify 155 episodes of exceptionally large capital inflows in middle- and high-income countries over the last 35 years. They report that larger inflows are associated with economic booms. The expansions are accompanied by rises in total factor productivity (TFP) and an increase in employment, which end when the inflows cease.
Moreover, during the boom there is also a reallocation of resources. The sectoral share of tradable goods in advanced economies, particularly manufacturing, falls during the periods of capital inflows. A reallocation of investment out of manufacturing occurs, including a reallocation of employment if a government refrains from accumulating foreign assets during the episodes of large capital inflows, as well as during periods of abundant international liquidity. The capital inflows also raise the probability of a sudden stop. Economic performance after the crisis is adversely affected by the pre-crisis capital inflows, as well as the reallocation of employment away from manufacturing that took place in the earlier period.
Alessandra Bonfiglioli of Universitat Pompeu Fabra looked at the issue of financial integration and productivity (working paper here). In a sample of 70 countries between 1975 and 1999, she found that de jure measures of financial integration, such as that provided by the IMF, have a positive relationship with total factor productivity (TFP). This occurred despite the post-financial liberalization increase in the probability of banking crises in developed countries that adversely affects productivity. De facto liberalization, as measured by the sum of external assets and liabilities scaled by GDP, was productivity enhancing in developed countries but not in developing countries.
Ayhan Kose of the World Bank, Eswar S. Prasad of Cornell and Marco E. Terrones of the IMF also investigated this issue (working paper here) using data from the period of 1966-2005 for 67industrial and developing countries. Like Bonfiglioli, they reported that de jure capital openness has a positive effect on growth in total factor productivity (TFP). But when they looked at the composition of the actual flows and stocks, they found that while equity liabilities (foreign direct investment and portfolio equity) boost TFP growth, debt liabilities have the opposite impact.
The relationship of capital flows on economic activity, therefore, is complex. Capital inflows contribute to economic booms and may increase TFP, but can end in crises that include “sudden stops” and banking failures. They can also distort the allocation of resources, which affects performance after the crisis. These effects can depend on the types of external liabilities that countries incur. Debt, which exacerbates a crisis, may also adversely divert resources away from sectors with high productivity. Policymakers in emerging markets who think about the long-term consequences of current activities need to look carefully at the debt that private firms in their countries have been incurring.