A Gaping Hole in EViews
Posted: Wed Dec 29, 2010 4:26 am
Hello,
EViews 7.1, with its friendly user interface, its add-ins and the possibility to dialogue with developers, has become the programme of choice for very many practitioners.
However, there is a gaping hole in EViews: the impossibility to estimate and simulate what has over the past two decades become the core of modern macroeconomics - the forward-looking DSGE model.
As a result, macroeconomists have had to turn to Dynare (and also RATS, which has a dsge instruction and for which Tom Doan of Estima has written a translator from Dynare to RATS [see http://www.estima.com/dynareconvert/dynareconvert.shtml]). Dynare is a free programme, developed by Michel Juillard of the Banque de France, essentially a pre-processor for Matlab, but which also works with Octave, a free Matlab clone.
Dynare estimates model parameters either by maximum likelihood as in Ireland ("A Method for Taking Models to the Data", 2004) or using a Bayesian approach as in Smets and Wouters ("An estimated stochastic dynamic general equilibrium model of the euro area", 2003). Both of these methods are well-developed and regularly used by macroeconomists, and it would seem (to the non-programmer!) but a relatively small step for EViews to incorporate them, thus dramatically expanding its appeal to macroeconomists.
From previous interactions with the EViews developers on this subject, I have the impression that doing so is on the "to do" list. This is a plea that it be given high priority.
Regards
Donihue
EViews 7.1, with its friendly user interface, its add-ins and the possibility to dialogue with developers, has become the programme of choice for very many practitioners.
However, there is a gaping hole in EViews: the impossibility to estimate and simulate what has over the past two decades become the core of modern macroeconomics - the forward-looking DSGE model.
As a result, macroeconomists have had to turn to Dynare (and also RATS, which has a dsge instruction and for which Tom Doan of Estima has written a translator from Dynare to RATS [see http://www.estima.com/dynareconvert/dynareconvert.shtml]). Dynare is a free programme, developed by Michel Juillard of the Banque de France, essentially a pre-processor for Matlab, but which also works with Octave, a free Matlab clone.
Dynare estimates model parameters either by maximum likelihood as in Ireland ("A Method for Taking Models to the Data", 2004) or using a Bayesian approach as in Smets and Wouters ("An estimated stochastic dynamic general equilibrium model of the euro area", 2003). Both of these methods are well-developed and regularly used by macroeconomists, and it would seem (to the non-programmer!) but a relatively small step for EViews to incorporate them, thus dramatically expanding its appeal to macroeconomists.
From previous interactions with the EViews developers on this subject, I have the impression that doing so is on the "to do" list. This is a plea that it be given high priority.
Regards
Donihue