The most recent U.S. recession has revitalized interest in business cycle research, while simultaneously challenging the predictions of existing macroeconomic models that abstract from differences across firms in aspects such as their productivity, capital, inventories and debt. Such rich differences may be important toward reconciling the empirical observation that the mean firm in the U.S. economy does not need to borrow to finance its investment with the widespread view that changes in the availability of credit had an important influence on the 2007 - 2009 recession.

This proposal explores the relation between production heterogeneity, discrete choices and aggregate fluctuations. The project seeks to develop and examine quantitative dynamic stochastic general equilibrium models to better understand the mechanics of economic fluctuations following real and financial shocks, and the role of inventories at various stages of the business cycle. At the heart of both parts of the proposal are macroeconomic models with rich heterogeneity across firms with respect to productivity, capital, inventories and debt. Such heterogeneity implies a multi-dimensional distribution of firms that affects the aggregate state of the economy, and thus economic fluctuations.

The recent real and financial crisis in the U.S. and abroad has driven researchers to try to integrate financial factors into standard models commonly used to study aggregate fluctuations. Before now, there has been little quantitative research on the channels through which changes in the availability of credit influence macroeconomic aggregates like business investment and production in a fully articulated setting. Project 1, Credit shocks and aggregate fluctuations in an economy with production heterogeneity, is unique in its approach to the topic, deriving an endogenous productivity channel through which changes in the allocation of credit can influence aggregate dynamics in an environment where firms differ in their productivities, capital, and debt. Existing research has generally abstracted from the rich heterogeneity across firms evident in microeconomic data. This is an important omission if such heterogeneity affects the evolution of macroeconomic variables. By contrast, this project explicitly includes a real friction hindering capital reallocation across firms in a setting with persistent productivity heterogeneity to reproduce salient micro-level investment patterns. Taken alongside collateralized borrowing arrangements, these elements imply a rich distribution of firms shaping total production, investment, and employment. That distribution evolves slowly in response to aggregate shocks, and itself protracts the economy's overall response to such shocks.

Inventory investment is highly volatile, both in business cycles and at higher frequencies. However, the series behaves quite differently over the two frequency bands. It moves positively with final sales at business cycle frequencies, while it moves negatively with sales over the short term. These conflicting patterns represent a challenge for existing micro-founded models of inventories. Project 2, Inventories, idiosyncratic shocks, and aggregate fluctuations, seeks to overcome the challenge. This project works to develop a single parsimonious DSGE model that can accommodate both sets of regularities when its parameters are disciplined by aggregate and firm-level data and inventory decision rules are endogenously derived as the result of an (S,s) motive.

An increased understanding of the ways in which real and financial frictions shape the distributions of productive inputs in modern developed economies, alongside the ways in which such distributions affect aggregate responses to real and financial shocks, will deepen policymakers' understanding of economic fluctuations. This, in turn, will help them to better anticipate the stages of the business cycle and to more effectively evaluate the need for various policy prescriptions and weigh the relative merits of competing policies. Preliminary findings in Project 1 indicate that the evolution of the distribution of firms following a credit shock can cause a large, gradual decline in economic activity and slow subsequent recovery. The model predicts clear differences in the responses of GDP and other macroeconomic variables following a credit shock versus a nonfinancial shock, and thus may offer policymakers a valuable tool. By distinguishing the responses to shocks, it can help identify the primary source of a recession and thus help them to determine which, if any, policies should be implemented to best promote economic growth and stabilization.

Analysts and policymakers place enormous emphasis on inventory investment as signal of economic conditions. There is no micro-founded quantitative business cycle model simultaneously consistent with the short-term and business cycle behavior of inventories for use in policy analysis as yet. Project 2 is an effort to correct this problem. Its successful completion will expand policymakers' ability to interpret ongoing movements in inventories and their understanding of how they influence and predict movements in other key macroeconomic series.

Project Report

Research funded by this NSF grant explores a series of newly awakened macroeconomic questions in light of the 2007 U.S. recession that demand careful inspection. There is growing consensus that this recession was caused or worsened by the financial crisis that began shortly before it. It was dramatic in its severity and otherwise unusual relative to other post-war recessions in the size of the drop in aggregate investment lending, in the large relative declines in investment, employment and GDP by comparison with the more modest decline in measured productivity, and in the fact that small firms contracted by much more than did large firms. Despite the common view that the unusual nature of this recession had something to do with the financial crisis, macroeconomics as a discipline has had very limited ability to provide a clear reasoning behind it, particularly given the fact that there were large aggregate cash holdings across nonfinancial firms at the start of the recession. 'Credit Shocks and Aggregate Fluctuations in an Economy with Production Heterogeneity,' sets out to solve this puzzle: How can a shock to the availability of investment loans cause or exacerbate a recession when the average firm in the economy does not need to borrow to finance its investment, as in true for U.S. nonfinancial firms? In a model economy where firms have differing needs for, and access to, investment loans, we show that a financial shock tightening collateral constraints on loans causes a recession with all of the unusual features noted above. This happens due to a misallocation of productive resources across firms, because small, young firms have unusual difficulty in accessing external finance to undertake necessary investment to expand to their efficient sizes. That misallocation itself pulls down aggregate productivity, lowering real interest rates and wages, and thereby discourages aggregate saving, investment, employment and production. The situation worsens over time, as a series of young cohorts are affected and their ordinary growth spells toward maturity are protracted, so aggregate productivity falls over several dates, as do aggregate quantity variables. While important for understanding how financial disruptions can cause recessions, the paper above omits two aspects of reality relevant to a more complete understanding of financially-related business cycles: changes in the numbers of operating firms, and changes in lending terms and repayment rates. Addressing these omissions, we find we can not only explain aspects of the 2007 downturn, but also do a far better job explaining the anemic economic recovery thereafter. Declines in the number of firms were dramatic in the 2007 recession. Our results in 'Entry, Exit and the Shape of Aggregate Fluctuations in a General Equilibrium Model with Capital Heterogeneity' help to explain the slow economic recovery following this recession expressly through large declines in the entry of new firms and rises in exit rates. We find that an aggregate shock raising firms’ costs of operating, a proxy for financial disruption, has large effects on the entry of new firms and disproportionately raises exits rates among small, young firms. This causes a large decline in the number of producers, reducing aggregate productivity and real economic activity. Moreover, it delivers a pronounced ‘missing generation’ effect whereby an usually small number of entrants fails to replace an increased number of exiting firms. This effect is most injurious several years out as the reduced cohorts of young firms approach maturity, a time at which they are most valuable to aggregate production and employment, and thus may help to explain the U.S. experience after June 2009. 'Default Risk and Aggregate Fluctuations in an Economy with Production Heterogeneity' extends the models above to consider default. Here, we assume that firms borrow in a competitive lending market where the loan interest rate each is offered takes into account the risk that it will default on its loan. Loan interest rates differ across firms, as default risk depends upon a firm’s productivity and the amount of capital and debt it chooses to take on. In this setting, we model a financial shock as a worsening of firms’ cash positions alongside a reduction in the fraction of a firm’s collateral that can be confiscated by lenders in the event of default. This shock worsens firms’ lending schedules, while pushing more of them into low cash positions where risk of default is high. The recession that follows is similar to that arising in the original model. However, the new model has useful added realism in that, like the U.S. economy, it experiences countercyclical default, procyclical firm entry and countercyclical exit. Because the endogenous credit tightening in this model causes substantial damage to the number and distribution of operating firms, and that damage takes time to repair, it does a better job of matching the slow post-2009Q2 U.S. recovery.

Agency
National Science Foundation (NSF)
Institute
Division of Social and Economic Sciences (SES)
Type
Standard Grant (Standard)
Application #
1061859
Program Officer
Georgia Kosmopoulou
Project Start
Project End
Budget Start
2011-09-01
Budget End
2014-08-31
Support Year
Fiscal Year
2010
Total Cost
$251,517
Indirect Cost
Name
Ohio State University
Department
Type
DUNS #
City
Columbus
State
OH
Country
United States
Zip Code
43210