Lumpy Investment and State-dependent Pricing in General Equilibrium Lumpy Investment and State-dependent Pricing in General Equilibrium

What are the aggregate consequences of microeconomic lumpy decisions? This is by now a classical question. Most existing general equilibrium analyses focus, however, on one single decision at a time. In the present paper we analyze simultaneous (S,s) pricing and investment decisions. Surprisingly, equilibrium dynamics are similar to what they would be in the absence of restrictions on price or capital adjustment. In that sense we generalize the prominent irrelevance result by Thomas (2002). Keywords: Lumpy Investment, Sticky Prices. JEL Classi…cation: E22, E31, E32 We thank the seminar participants at Norges Bank and Universitat Pompeu Fabra for useful comments on earlier versions of this paper. We are especially thankful to Ruediger Bachman. Part of this work was done while Sveen was visiting CREI and Universitat Pompeu Fabra. He thanks them for their hospitality. The usual disclaimer applies. The views expressed in this paper are those of the authors and should not be attributed to Norges Bank. 1 1 Introduction Many microeconomic decisions are lumpy in nature. Caballero and Engel (2007) note that examples include not only infrequent price adjustment by …rms but also investment decisions, durable purchases, hiring and …ring decisions, inventory accumulation, and many other economic variables of interest. Recently, some important progress has been made in understanding the consequences of microeconomic lumpiness for aggregate dynamics (See, e.g., Dotsey et al. 1999, Thomas 2002). It is, however, striking to note that most existing theoretical analyses of (S,s) decisions have focused on one particular lumpy decision at a time. This motivates our paper. We try to understand the macroeconomic consequences of simultaneous lumpy decisions in general equilibrium. To this end we study macroeconomic consequences of state-dependent pricing and lumpy investment. We obtain two main results. First, the restriction on capital adjustment does not have quantitatively important implications for equilibrium dynamics. Second, the same is true for the restriction on price adjustment. Our …rst …nding generalizes a prominent result by Thomas (2002). She observes that (S,s) lumpy and frictionless investment imply similar dynamics for aggregate quantities in the context of a standard technologydriven RBC model. Interestingly, we con…rm her equivalence result in a framework featuring sticky prices combined with monopolistic competition. The second result is in the spirit of Golosov and Lucas (2007). They show that (S,s) price stickiness has small aggregate consequences in the presence of (ad hoc) idiosyncratic productivity shocks. We demonstrate that endogenous changes in …rm-level productivity resulting from (S,s) lumpy investment can micro-found this. What is the relevance of our results? We show that the dynamic consequences of simultaneous (S,s) decisions di¤er crucially from what they are if it is assumed that investment and pricing decisions are made according to time-dependent adjustment rules. The latter assumption is pursued in Sveen and Weinke (2007). In that paper we obtain equivalence between a time-dependent monetary lumpy investment model 2 and a speci…cation featuring a convex capital adjustment cost at the …rm-level. Speci…cally, monetary disturbances are found to have large real consequences in a way which is empirically plausible. In the present paper we obtain a diametrically opposed result which is due to the (S,s) nature of the decisions under consideration. Of course, there is no agreement in the literature on which way of modeling lumpy decisions is most appealing (See, e.g., Woodford 2008). Our results suggest, however, that this is one of the most important questions in monetary economics. So far very little work has been done on the aggregate consequences of simultaneous lumpy decisions at the microeconomic level. The paper which is most closely related to ours is the recent work by Johnston (2007). Like our paper he also analyzes simultaneous (S,s) pricing and investment decisions at the …rm level. An important di¤erence with respect to our work lies, however, in the way in which the heterogeneity among …rms is reduced in order to make the problem tractable. In that regard our assumptions are more general, as we are going to see.1 Johnston concludes that the Thomas (2002) result regarding dynamic consequences of technology shocks is overturned in the presence of a price-setting decision at the …rm level, whereas we con…rm her result. On the other hand, Johnston also …nds that the presence of an (S,s) investment decision results in a reduction of the real consequences of monetary disturbances and, like us, he relates that to the above mentioned result by Golosov and Lucas (2007). The remainder of the paper is organizes as follows. Section 2 outlines the model. Section 3 presents the results and Section 4 concludes. 1Another important di¤erence with respect to our model lies in the speci…cation of monetary policy. He assumes a stationary process for the growth rate of real balances (combined with an interest rate inelastic demand for real balances), whereas we assume that monetary policy takes the form of an interest rate rule.