Evaluating UK Point and Density Forecasts from an Estimated DSGE Model: The Role of Off-Model Information Over the Financial Crisis

Fawcett, Nicholas; Körber, Lena Mareen; Masolo, Riccardo M.; Waldron, Matthew

doi:10.2139/ssrn.2639055

Cited by 23 publications

(16 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Its 15 endogenous variables also exclude credit and financial conditions. Fawcett et al (2015a) examined what the forecast performance of COMPASS would have been, working with real-time data, had it been in existence in 2007: it would also have failed very badly over the financial crisis. 15 Common sense about the economy, using a wider set of information, as embodied in private-sector expectations and by the MPC, forecasted better than COMPASS.…”

Section: (Ii) Forecasting Failuresmentioning

confidence: 99%

The future of macroeconomics: macro theory and models at the Bank of England

Hendry

Muellbauer

2018

Oxford Review of Economic Policy

View full text Add to dashboard Cite

The adoption as policy models by central banks of representative agent New Keynesian dynamic stochastic general equilibrium models has been widely criticised, including for their simplistic micro-foundations. At the Bank of England, the previous generation of policy models is seen in its 1999 medium-term macro model (MTMM). Instead of improving that model to correct its considerable flaws, many shared by other non-DSGE policy models such as the Federal Reserve's FRB/US, it was replaced in 2004 by the DSGE-based BEQM. Though this clearly failed during and after the global financial crisis, it was replaced in 2011 by the DSGE COMPASS, complemented by a 'suite of models'. We provide a general critique of DSGE models for explaining, forecasting and policy analyses at central banks, and suggest new directions for improving current empirical macroeconomic models based on empirical modelling broadly consistent with better theory, rather than seeking to impose simplistic and unrealistic theory.

show abstract

Section: (Ii) Forecasting Failuresmentioning

confidence: 99%

The future of macroeconomics: macro theory and models at the Bank of England

Hendry

Muellbauer

2018

Oxford Review of Economic Policy

View full text Add to dashboard Cite

show abstract

“…The fact that the headline policy rate was constant at 0.5 percent for a prolonged period in our sample (2009 onwards), while unconventional policy measures were adopted, presents our estimation with a challenge. To tackle this issue, we follow Fawcett et al (2015) and include a measure of the shadow rate (similar to those considered by Krippner 2013 andWu andXia 2016) in place of the headline policy rate in our estimation. Our measure of the shadow rate is based on estimates of the effects of QE presented in Bridges and Thomas (2012).…”

Section: Estimationmentioning

confidence: 99%

Deflation Probability and the Scope for Monetary Loosening in the United Kingdom

2016

View full text Add to dashboard Cite

We use an estimated DSGE model of the U.K. economy to investigate perceptions of the scope for monetary loosening since the 2007-08 financial crisis. We simulate our model under alternative assumptions about the constraints on monetary policy and compare the implied deflation probabilities with those expected by professional forecasters, financial markets (from options prices), and policymakers. We find that these real-time, external estimates of deflation probability are consistent with policy not having been perceived to have been constrained since the crisis. JEL Codes: E31, E37, E43, E47, E52.

show abstract

“…observed ex post) of the future paths of the technical assumptions. 2 Besides validating our measure of the gaps, this exercise also contributes to the literature that evaluates the practices of central bank forecasting, given the relevant role of conditional forecasts in the toolbox of central banks (see, for example, Del Negro and Schorfheide, 2013, Giannone et al, 2014, Fawcett et al, 2015, Iversen et al, 2016and Domit et al, 2016. We choose to primarily evaluate conditional forecasts based on the actual value of the assumptions (even though we also provide an assessment of the conditional forecasts based on real-time assumptions) because, from a statistical point of view, this is the most appropriate way to assess the accuracy of conditional forecasts (for an extensive discussion of this point, see Faust andWright, 2008 andMcCracken, 2014).…”

Section: Introductionmentioning

confidence: 99%