روابط اعتباری پویا در تعادل عمومی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|28645||2006||31 صفحه PDF||سفارش دهید||15455 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Monetary Economics, Volume 53, Issue 4, May 2006, Pages 847–877
We construct a general equilibrium model with private information in which borrowers and lenders enter into long-term dynamic credit relationships. Each new generation of ex ante identical individuals is divided in equilibrium into workers and entrepreneurs. Workers save through financial intermediaries in the form of interest-bearing deposits and supply labor to entrepreneurs in a competitive labor market. Entrepreneurs borrow from financial intermediaries to finance projects which produce privately observed sequences of random returns. Each financial intermediary holds deposits from a large number of workers and operates a portfolio of dynamic contracts with different credit positions. We calibrate the model to the U.S. economy and find that dynamic contracting is very effective at mitigating the effects of private information. Moreover, restricting borrowers and lenders to use static (one-period) contracts with a costly monitoring technology has adverse effects both on the level of aggregate economic activity and on individual welfare unless monitoring costs are very small. Finally, the optimal provision of intertemporal incentives leads to increasing consumption inequality over time within generational cohorts as in U.S. data.
The last 20 years have seen many efforts at modelling explicitly private information and the financial contracting process in the benchmark neoclassical growth model. In particular, a substantial part of the literature addresses the question of how the presence of private information contributes to the propagation of aggregate economic uncertainties. Bernanke and Gertler (1989), for example, argue that with private information, swings in the firm's balance sheet are a potential source of persistent output dynamics. In Williamson (1987b), monitoring costs of financial intermediaries are important for the propagation of aggregate disturbances. Kiyotaki and Moore (1997) also study business cycle dynamics propagated through the financial contracting process between entrepreneurs and investors, although their story is based on limited commitment instead of private information. Carlstrom and Fuerst (1997) develop a computable general equilibrium model based on Bernanke and Gertler (1989) to address quantitatively the importance of agency costs for the propagation of aggregate shocks. Cooley and Nam (1998) incorporate a problem of debt contracting with asymmetric information into a quantitative monetary business cycle model to generate a persistent liquidity effect induced by monetary disturbances. An obvious yet serious limitation of most of the existing literature on financial contracting and business cycles is that the financial borrowing and lending process is modelled as a one-shot game.1 In practice, financial intermediaries often engage in long-term relationships, rather than interact only once with their borrowers. Modelling the financial lending process as a one-period contract may severely restrict the contracting parties’ ability to achieve risk-sharing, and hence may have important implications for how successfully the contract can be used in a dynamic macroeconomic setting as an explicit description of the financial lending process. In addition, from a technical perspective, in a standard dynamic general equilibrium macroeconomic model, borrowers and lenders are all infinitely lived agents, and very special assumptions have to be made in order to fit the static contracting relationship into the rest of the economy (which is fully dynamic).2 The above limitation of the literature was first pointed out by Gertler (1992), who developed a model in which lenders and borrowers can enter into long-term but finite contractual relationships and used his model to show that shifts in aggregate economic fundamentals can be amplified through the process of long-term contracting. Yet, as Gertler himself pointed out, “a major limitation of this model is that it lies well short of a fully dynamic framework that can be matched to data. While allowing for multi-period contracts …… is a helpful step in this important direction, there is still a long way to go” (p. 470). What this paper attempts to undertake can be viewed as just another step in the long way that Gertler (1992) pointed out. Instead of being ambitious in providing a theory that explains business cycles dynamics, this paper constructs a quantitative dynamic general equilibrium model with no aggregate uncertainty but in which long-lived economic agents can enter into fully dynamic financial lending contracts. We find that, relative to static contracts, dynamic contracts generate equilibrium allocations with higher output and higher welfare. In our model, the economy is populated by a sequence of overlapping generations of agents who potentially can live forever. Each new generation of ex ante identical individuals is divided, in equilibrium, into workers and entrepreneurs. Workers save through financial intermediaries in the form of interest-bearing deposits and supply labor to entrepreneurs in a competitive labor market. Entrepreneurs borrow from financial intermediaries to finance projects which produce idiosyncratic sequences of random returns. The entrepreneur's returns are not observed by any other parties. Financial intermediaries arise as institutions to facilitate financial borrowing and lending by providing risk-sharing for entrepreneurs and workers. Financial intermediaries are competitive, each holding deposits from a large number of workers and operating a portfolio of dynamic contracts with different credit positions. A notable feature of our model is that we model occupational choice as an equilibrium phenomenon. In the model, ex ante identical economic agents choose to become entrepreneurs (borrowers) and workers (depositors). Occupational choice plays a central role in our model as the link between the labor and asset markets and the determination of aggregate output. Our model provides a vehicle for evaluating both qualitatively and quantitatively the implications of dynamic contracting and financial intermediation for the operation of the macroeconomy.3 How does dynamic contracting operate in general equilibrium? What are the implications of dynamic contracting for the equilibrium dynamics of the balance sheets of financial intermediaries? What role does private information play in the determination of the general equilibrium levels of aggregate quantities such as output and the capital stock? What are the implications of dynamic contracting for individual welfare (relative to static contracting under private information and full risk-sharing under complete information)? How does dynamic contracting affect inequality through the provision of intertemporal incentives? How important is costly monitoring for the performance of the macroeconomy? A distinctive feature of our model is that, in equilibrium, financial intermediaries hold positive amounts of assets. These assets arise in our model even though we assume a competitive financial intermediation industry, so that financial intermediaries earn zero profits. The existence of the assets held by financial intermediaries in our model is an equilibrium outcome associated with the dynamic lending process.4 In the model here, the equilibrium provision of intertemporal incentives in the dynamic contract implies that the entrepreneurs make more repayments in earlier stages rather than later stages during the credit contracting process. The assets held by the financial intermediaries are essentially savings of the entrepreneurs. As we will show in the paper, among other factors, the equilibrium size of the assets held by financial intermediaries depends critically on the rate of interest. A key quantitative finding of the paper is that dynamic contracting is very effective at mitigating the effects of private information: even when informational asymmetries are large, the economy can achieve close to a first-best outcome. One implication of this finding is that costly state verification (CSV), which is a key feature of many static models of the borrowing and lending process, is not important quantitatively in the presence of fully dynamic credit relationships. The model of CSV, originally developed by Townsend (1979), has been used widely as a vehicle for studying the relation between financial intermediation and macroeconomic aggregates (Bernanke and Gertler, 1989, Williamson, 1987b and Boyd and Smith, 1997). We also demonstrate the importance of dynamic contracting for the determination of macroeconomic aggregates in our model by forcing borrowers and lenders to use static (one-period) contracts with a costly monitoring technology rather than dynamic contracts. We show that moving from dynamic to static contracting (i.e., restricting the ability of contracts to use intertemporal incentives) has adverse effects both on the level of aggregate economic activity and on the welfare of individuals unless monitoring costs are very small. In other words, this paper offers strong quantitative evidence supporting the argument (Gertler, 1992) that modelling the financial lending relationship as a one-shot game can be very misleading in explaining macroeconomic activities. We also show that the quantitative framework we construct can partially account for the increasing consumption inequality over time within generational cohorts that Deaton and Paxson (1994) document in U.S. data. That private information and incomplete insurance provide a potential explanation for consumption inequality has been the theme of several recent studies. These studies, including Green (1987), Atkeson and Lucas, 1992 and Atkeson and Lucas, 1995, and Banerjee and Newman (1991), are all theoretical investigations and have not addressed the relationship between incentives and consumption inequality using a quantitative growth model that features dynamic contracting such as the one that we develop in this paper.5 Of course, our finding that private information and incentives can account only partially for the observed pattern of consumption inequality is well anticipated. The literature on inequality and its determinants shows that other factors, including ex ante heterogeneity in ability, uninsurable idiosyncratic shocks, and borrowing constraints, also play important roles in determining consumption inequality (see, e.g., Becker and Tomes, 1979; Deaton and Paxson, 1994; Storesletten et al., 2004). Finally, we illustrate how our model economy can provide a vehicle for conducting various types of policy experiments. In particular, we use our model to analyze the effects of requiring financial intermediaries to hold a fraction of their deposits as reserves. We find that the cost of holding 10%10% of total deposits as required reserves is roughly 1%1% of aggregate output. In equilibrium the economy holds more capital in total in response to an increase in required reserves. Section 2 presents the model, Section 3 discusses computation of the model, Section 4 explains how we calibrate the model, Section 5 presents the results, and Section 6 concludes.
نتیجه گیری انگلیسی
This paper develops and implements a general equilibrium model with private information in which borrowers and lenders enter into long-term credit relationships. We calibrate the model to U.S. aggregate data and use the model to analyze the impact of informational frictions on the behavior of the macroeconomic aggregates and on the welfare of individuals. A key assumption in our model is that there are no aggregate risks, so that the macroeconomic aggregates do not vary over time. An important extension to our model would be to introduce aggregate uncertainty in the form of common shocks to the productive opportunities of entrepreneurs. This extension would allow us to examine how financial intermediation affects the propagation of aggregate shocks. Introducing aggregate uncertainty is a challenging computational problem since agents in the economy then need to keep track of the dynamic behavior of the distribution of asset holdings across workers and of the distribution of promised utilities across entrepreneurs. This is the subject of ongoing research.21