Wednesday, November 11, 2020

Side Effects Of Stability Of DSGE Models — Brian Romanchuk

The stability of the solutions of workhorse dynamic stochastic general equilibrium (DGSE) models is a major drawback of these models when trying to explain recessions. This article looks at the estimated linear New Keynesian model that was used in the previous discussion of the estimation of r* -- the Holston-Laubach-Williams (HLW) model. 

As I have noted a few times previously, more interesting behaviour can be obtained by nonlinear DSGE models. However, this flexibility comes at the cost of increasing the difficulty of fitting the model to data. There is an infinite number of models one can use to tell stories with.

Models are general descriptions of possible worlds. Only data can determine how closely the possible world described by a model may be to the real world, and data determining this correspondence are exogenous to the model. Data gathering, processing and processing into information have their own problems.

An analogy would be a nautical chart that does not have information about reefs and shoals. The chart is fine most of the time, but.…

There is tension between model building and data that can lead to various issues set forth in the literature. One problem is that modelers often disregard this literature, if they are even aware of it, e.g., owing to cognitive-affective bias, ideology, and lack of awareness of priors such as unstated assumptions and lurking presumptions as meta-assumptions.

Why, especially in the face of much criticism and professional critiquing? This question has a variety of answers, many about the modelers, e.g., silos and group think.

Bond Economics
Side Effects Of Stability Of DSGE Models
Brian Romanchuk

3 comments:

Peter Pan said...

An analogy would be a nautical chart that does not have information about reefs and shoals.

So DSGE models are fraudulent?

Tom Hickey said...

So DSGE models are fraudulent?

Nautical charts are not fraudulent unless doctored, but they can be dangerously incomplete. Conventional econometric models are not fraudulent but inaccurate when it counts — at turning points.

Matt Franko said...

They’re not functional equations describing functional relationships...