The Dynamic Effects of Firm-level Borrowing Constraints

SUNY-Albany Discussion Paper 00-02



Abstract. In this paper I develop a detailed dynamic model of firm behavior in order to see whether financial constraints and endogenous exit are important propagation mechanisms. To do this, I construct an economy where firms face financial constraints, fixed costs and persistent idiosyncratic shocks. Using numerical methods, I analyze how a large collection of these firms responds to aggregate productivity shocks. A common result is that financial constraints tend to dampen the economy's initial response to aggregate productivity shocks, but that equity accumulation and exit dynamics amplify the longer-term response. The relative sizes of these two effects, however, are sensitive to firms' environments.
 

  • Text
  • Data (Not a big part of this analysis, but here it is)
  • Computer Code

  •  

     

    Text

  • Scientific Word-generated *.pdf file
  • Working paper version, with lots of gory details

  • Data

  • ASM data set, cleaned up and annotated by Bartlesman and Gray.  (Note: To free up space, this item might be missing: Feel free to email me.)
  • But.... You are strongly advised to go to the NBER web site for the latest, cleanest update.

  • Computer Code   (Note: To free up space, some items might be missing: Feel free to email me.)
    Underlying GAUSS codes (zipped).  The two key files are the "clcns" ones. You can either solve the firm's problem from scratch, or generate time series from an existing set of policy functions.  To do the latter, you will have to set the variable "useold" to "1" and you will need the policy vectors contained in the "optpol" files, which you can download here.  It will be up to you to make sure the parameters of the model's current iteration are consistent with the parameters used to generate the policy functions. The "Markch" files comprise a GAUSS procedure library.  You should load them in the appropriate subdirectories of your "gauss" directory.

    This code utilizes work by John Rust. Anyone interested in numerical dynamic programming should visit his web page.
     



    Last Updated on November 8, 2004 by John Jones