The financial accelerator in an estimated New Keynesian model

Abstract

This paper estimates and simulates a sticky-price dynamic stochastic general-equilibrium model with a financial accelerator, à la Bernanke et al. [Bernanke, B., Gertler, M., Gilchrist, S., 1999. The financial accelerator in a quantitative business cycle framework. In: Handbook of Macroeconomics. North-Holland, Amsterdam], to assess the importance of the financial accelerator mechanism in fitting the data and its role in the amplification and propagation of transitory shocks. Structural parameters of two models, one with and one without a financial accelerator, are estimated using a maximum-likelihood procedure and post-1979 US data. The estimation and simulation results provide quantitative evidence in favor of the financial-accelerator model. The model without a financial accelerator is statistically rejected in favor of a model with it. The presence of the financial accelerator amplifies and propagates the effects of demand shocks on investment, but it dampens those of supply shocks. However, we find that the importance of the financial accelerator for output fluctuations is relatively minor. Crown

Macroprudential rules and monetary policy when financial frictions matter

“…14 14 Given the importance of the parameters related to financial frictions and the financial shock in our study, it is useful to compare our estimates of these parameters to those from previous studies. Our estimate of χ is close to the estimates (χ = 0.04) in Christensen and Dib (2008) and Covas and Zhang (2010). For the financial shock, we compare our estimates with those in Dib et al (2008), a paper that also estimates a DSGE model with a financial shock for the Canada economy.…”

“…See, for example, Christensen and Dib (2008) who estimate a sticky-price DSGE model with a financial accelerator using U.S. data and find that their model is better able to account for the key features of the U.S. data than the same model without a financial accelerator mechanism. Covas and Zhang (2010) estimate a sticky-price DSGE model that includes debt and equity market frictions using Canadian data and find a similar result.…”

“…For examples, Christiano et al (2003), Gilchrist and Saito (2006), Christensen and Dib (2008) and Covas and Zhang (2010).…”

Financial shocks and the US business cycle

“…They argue that other features of the model, such as investment adjustment costs, are more important. Christensen and Dib (2008), on the other hand, use a maximumlikelihood procedure to estimate a new Keynesian model with and without a …nancial accelerator mechanism, and incorporating a wider set of shocks compared with Meier and Müller. In contrast, they …nd that the quantitative signi…cance of the FA mechanism is somewhat more important in understanding monetary shocks, although it is less important for understanding output volatility.…”

“…As we are interested in estimating three shocks, we need data on and at least three variables contained in Y , the choice of which we discuss presently: 9 7 We use the King and Watson solution algorithm. 8 Here, k t is the capital stock, w t the hourly real wage, q t is Tobin's q, r t denotes the real interest rate, nw t our measure of entrepreneurial net wealth, i t is the federal funds rate, c t is consumption and m t are real money balances.…”