insider in london macro firm here, we use DSGE models. can confirm other macro firms use them too.
For what purpose?
DSGE in the Private Sector
-
Insider in a small, boutique private equity firm. We do not use DSGE models, but we do use DSGE models as a screening device, in the same spirit as graduate programs use real analysis. It turns out that we can get these guys cheaper than MBAs, and they have similar levels of firepower.
-
insider in london macro firm here, we use DSGE models. can confirm other macro firms use them too.
For what purpose?forecasting/impulse response analysis primarily. we have an in-house model that we maintain/develop.
it's not our only tool of course, that would be stupid. but it forms an input into our analysis.
-
Noah Smith has clearly picked up on this thread with his latest post about how the lack of adoption by private sector of DSGE is the final proof of its worthlessness.
Anyone interested in this topic should check out the comments section. Very insightful. Both Cochrane and Williamson comment on it was well.
http://noahpinionblog.blogspot.com/2014/01/the-most-damning-critique-of-dsge.html
-
I think only a few people would deny that true scientific insight is a public good. Which is why science is concentrated in the academic sector which turns out to be the sector where "reputation" can be monetized. There is a guy out there who can tell you for a price, how business cycles will affect the toilet paper sector; its worth money, but because of scale diseconomies nobody else really enters that sector. I doubt that guy really makes more money than Larry Christiano, and no disrespect toilet paper bro, but I would rather be Larry Christiano.
There is no skill that pays off to the same degree as talking rich people into letting you invest their money for them. Why isnt Stanley Fisher as rich as Jordan Belfort?
-
Former IB economist here. We did not use DSGE models, mostly OLS and VARs and a lot of accounting exercises and carefully examining how different indices were constructed. Regularly ranked in top forecasters, which is not to say that our forecasts were any good but that serious models don't really help that much for that type of work.
-
Former IB economist here. We did not use DSGE models, mostly OLS and VARs and a lot of accounting exercises and carefully examining how different indices were constructed. Regularly ranked in top forecasters, which is not to say that our forecasts were any good but that serious models don't really help that much for that type of work.
And your track record has been impressively accurate, hasn't it.
-
Economist at central bank.
My own complaints about the DSGE method are based in hard experience. Our monetary policy committee hasn't had much confidence in the output of our in-house DSGE projection model for a long time. The amount of ad-hoc judgment and auxiliary modelling needed to make it usable at all would curl the public's hair if they knew the details. We don't even try to use it to model things like house prices or household debt, just for example. DSGE models have never been good at modelling asset prices or balance sheet variables.
-
Economist at central bank.
My own complaints about the DSGE method are based in hard experience. Our monetary policy committee hasn't had much confidence in the output of our in-house DSGE projection model for a long time. The amount of ad-hoc judgment and auxiliary modelling needed to make it usable at all would curl the public's hair if they knew the details. We don't even try to use it to model things like house prices or household debt, just for example. DSGE models have never been good at modelling asset prices or balance sheet variables.Another central bank economist here. Totally agree with you, word for word. Can you drop a hint as to which central bank you're at? We may not be strangers..
-
Lol ^^
For what it's worth I think ad hoc judgement really is the way to go. No one seriously thinks that date models (*dsge) model the economy. These committee guys are ultimately paid for this so their judgements should be the final word
Where the dsge model really helps is imposing internal consistency on their judgements. Without this they would just make a bunch of judgements that make no sense when taken as a whole