Abstract

This study employs a Monte Carlo simulation to see whether accounting identity problems are present in the Fazzari, Hubbard, and Petersen model (1988). The Monte Carlo simulation generates 50,000 random cash flows, Tobin’s Q, and error term variables, which shape an investment variable that is dependent on them. Cash flows and investments are linked by a partial accounting identity, also known as an accounting semi-identity (ASI). An accounting identity is, for example, an equality between the left and right sides of a balance sheet. An ASI is not a complete one since one or more components of the accounting identity are missing. The estimated coefficients of an ASI do not represent reality, according to the OLS estimations. The regression tells us less about causality the closer the data are to the accounting identity. This is the first time that the biases of OLS estimations in an ASI-based model have been demonstrated.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call