Fed, GS forecasting models: nonexistent or very inadequate


I am not arguing to use any models of the Fed or Goldman Sachs for forecasting bubbles and crashes such as occurred in 2007, 2008. I am not aware of any Goldman Sachs model that would be relevant. I doubt if they have such a model.

I am pretty sure the Fed did not have any relevant models before the crash of 2008, and probably no good model yet after the crash of 2008. You can read conference papers at
Some are quite readable and interesting but the models are antiques: CGE VAR and ECT. Their “recent” technology is circa 1980. The Feds main model they admit is 1960s and 1970s. Long before derivatives became popular. The culprit derivatives were not even invented when the forecasting technologies were developed.

Further you can read the models have improved somewhat since the crash and may help prevent future crashes. But there is no description of model capabilities that would predict the crash http://www.federalreserve.gov/newsevents/speech/tarullo20140625a.htm

I do not expect that either the Fed or GS or any party will produce any good crash prediction models in my lifetime. Such a model would have to distinguish between what is a trend (such as faster computers) and what is a bubble (too many computers per capita). If something is good and healthy then it may stay around for a long time. But something else may produce disease in the longer run, people may die, and sales collapse. Only a good scientific model would be able to determine if there are too many derivatives or too little, and which ones will grow and which will collapse.

This lack of good models is caused by lack of good science and education in our universities. Only a handful of universities teach the needed classes. Then Phd students will rarely take enough of these needed classes to be able to do the correct modelling. I am not aware of any student or professor who has the relevant knowledge or interests to be worth talking to. Maybe the Fed in Washington DC has collected a few staff, but I have not seen any serious new models come out from them yet.

Greenspan, Bernanke, or Yellen do not have any background in Derivatives that would allow them to forecast the crash of 2008. I doubt if very many of their staff had the background. Maybe none. I have never seen any significant work on derivatives from the fed so never look at their stuff. They are mostly monetary economists who are not required to take Finance for their Phd. They probably have hired some since the crash as evidenced by their discussion on stress testing. After the fact.

But probably not nearly what is needed. I am going through several hundred pounds of journals on my living room floor, tossing most into the trash can. If I see anything good I will email you a link.

For the record, everybody I know knew a crash was probably coming, a repeat of recent bubbles. Bloomberg reporter was calling Greenspan DBCB double bubble central banker 2 years before the crash, quoting Anna Schwartz criticism of 2003-2004 loose money which was about time I began getting disturbed, coming so soon after the dot-com crash: http://www.bloomberg.com/apps/news?pid=newsarchive&sid=aouwvLwhMQRU

I would have to read Bernanke’s statements to see if he was aware of a bubble, say, in Feb 2008. I don’t care much what he ever said so may not do that. But it is worrisome that people who should have a gut feel for data cannot see when numbers are way out of equilibrium. Bernanke seems to have less senile dementia than Greenspan. This points to models that can do the required calculations and emit beeps. Like a horse that can find its way home even if the rider is drunk of shot.


On 6/28/14, Ron wrote:
> Your response appears very erudite but if the mathematical models used > by the Fed and Goldman Sachs were to be employed I would be inclined > to wager just the opposite of their claimed predictions. If I recall > it was Alan Greenspan and later Bernanke that stated in 2007 the > economy was “doing just fine”.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s