Really Uncertain Business Cycles

Abstract

We propose uncertainty shocks as a new shock that drives business cycles. First, we demonstrate that microeconomic uncertainty is robustly countercyclical, rising sharply during recessions, particularly during the Great Recession of [2007][2008][2009]. Second, we quantify the impact of timevarying uncertainty on the economy in a dynamic stochastic general equilibrium model with heterogeneous firms. We find that reasonably calibrated uncertainty shocks can explain drops and rebounds in GDP of around 3%. Moreover, we show that increased uncertainty alters the relative impact of government policies, making them initially less effective and then subsequently more effective.JEL classification: E3

The Market for Used Capital: Endogenous Irreversibility and Reallocation Over the Business Cycle

“…Figure 9 illustrates the effect of endogenous irreversibility on the dispersion of the marginal product of capital, as measured by the average marginal product for high productivity (s H ) firms and that for low productivity ones (s L ). Consistent with a growing body of empirical evidence (Bloom et al, 2012), the model implies that returns are more dispersed in recessions than in booms. This fact is a manifestation of the lack of reallocation, because large unproductive firms downsize less than they would in a model with constant irreversibility, which prevents an equalization of marginal returns.…”

“…Thus, the market for used capital clears between investing and disinvesting firms only. 10 This assumption is equivalent to the assumption of total irreversibility in the aggregate, as, for instance, in Sargent (1980).…”

“…Parameter values are only meant to be illustrative and are shown in Appendix C. I solve the two-period model numerically and compute option values by numerical integration. Figure 4 illustrates the payoff of the put option value in equation (10), evaluated at k 0 and at the optimal choice k 1 , as a function of the initial idiosyncratic shock s 1 and for each of two values of z 1 . I label the low realization of z 1 "recession" and the high realization "boom".…”

“…However, in the literature there is no unanimous agreement on this parameter value and, in general, on the procedure to parametrize the idiosyncratic productivity process. For example, Bloom et al (2012) estimate time-varying volatilities of firm-level productivity and get values for volatility of up to 12% quarterly in high-uncertainty periods. This suggests that the amplification of aggregate productivity delivered by the present paper is a lower bound for the empirical effect of procyclical capital reallocation on TFP.…”

Interpreting volatility shocks as preference shocks

“…Proposition 1 shows that when the IES is greater than one, which is an empirically plausible assumption, 6 a large increase in volatility can decelerate the growth of output and consumption in the absence of any level shocks, a result consistent with related …ndings in the existing literature of uncertainty shocks. In the face of elevated uncertainty about future capital returns, the risk-averse representative household would prefer to spend a greater fraction of its current wealth on consumption, 7 with the resulting decreased investment in the …rm slowing the pace of capital accumulation and output growth after the volatility shock. Subsequently, future consumption falls relative to current consumption, implying a slower growth in consumption.…”