A primer on DSGE Models

DSGE Models are being thrashed everywhere. These are actually based on New Keynesian thinking which had become a vogue till this crisis. The New Keyensian thinking merges Keynes ideas (markets can fail) with classical economics (rational human beings). Wikipedia nicely explains it:

Two main assumptions define the New Keynesian approach to macroeconomics. Like the New Classical approach, New Keynesian macroeconomic analysis usually assumes that households and firms have rational expectations. But the two schools differ in that New Keynesian analysis usually assumes a variety of market failures. In particular, New Keynesians assume prices and wages are “sticky“, which means they do not adjust instantaneously to changes in economic conditions.

Wage and price stickiness, and the other market failures present in New Keynesian models, imply that the economy may fail to attain full employment. Therefore, New Keynesians argue that macroeconomic stabilization by the government (using fiscal policy) or by the central bank (using monetary policy) can lead to a more efficient macroeconomic outcome than a laissez faire policy would. However, New Keynesian economics is less optimistic about the benefits of activist policies than traditional Keynesian economics was.

The DSGE model is basically a generic term but has become associated with new Keynesian thinking. These models though have included Keynesian thinking, ignore the role of financial markets. They assume markets to be efficient and self-correcting and not worthy of being included in the models (see this criticism by Munchau and Otmar Issing).

Now these models are being increasingly used by central banks and are highly fashionable. With most central banks failing to see the crisis coming, the models are being questioned. The fact that financial crisis are pretty common how can models ignore them? True, any model takes time to develop but then why should central banks rely strongly on it?

Camilo Tovar of BIS has written an excellent primer(no equation, no Latin terms )to explain the basics of this DSGE model. Tovar adds apart from not including financial markets comprehensively, it also has a very limited role for currency risk premia and fiscal policies (follows Ricardian equivalence, no wonder we understand so little about fiscal policies). He also adds on DSGE modelling issues with emerging economies.

Quite a decent read. Despite his best attempt to explain in English, one needs some understanding of econometrics to understand the paper fully (needless to say, I still struggled with the paper and just got a feeler). He also has added a huge list of references and papers explaining DSGE models used by different central banks. (Though doubt any paper would be as simple as this one ).

One can also read Mankiw’s new chapter explaining basics of DSGE Models.

PS.

I came across this interesting link. NIPFP, Delhi based economic think-tank along with Department of Economic Affairs, Ministry of Finance  has been organising training sessions on DSGE modelling! One is simply teaching about these models, it is altogether different to apply them for policies. Can DSGE model be applied in Indian setting? What changes does one make to make them suitable to Indian conditions? Anybody knows of any papers on any DSGE modelling in India, let me know.

But IMF does macroprudential analysis and it hasn’t worked

You pick up any report/paper/article on lessons from the crisis/preventing future crisis/reforming financial system and regulation  and one of the foremost suggestion you would get to read is – Implement Macroprudential approach to financial regulation.  It basically means the regulators need to shift from focusing on individual firms to the system as a whole. The individual firms may look well (just like before the crisis) but as a whole could pose risks to the entire system . Read this Bernanke speech for further insights.

FSA, UK conducted a conference to discuss Turner Review. I was reading this speech given by Adair Turner himself at this conference. It serves as a good summary of the review.

These words were of immediate interest to me.

The Review sets out eight sets of recommendation for regulatory reform. ……Fourth, the importance of macro-prudential analysis and intellectual challenge.  

however, one thing we clearly need is intellectual challenge to conventional wisdom. Because where we did do macro-prudential analysis, it often still failed to see the emerging problems.

Each edition of the IMF Global Financial Stability Report is full of macro-prudential analysis. But in April, 2006 it said this (Exhibit 16):

 ‘There is  a growing recognition that the  dispersion of credit risk by banks to a broader and more diverse group of investors, rather than warehousing such risk on their balance sheets, has helped make the banking and overall financial system more resilient.

The improved resilience may be seen in fewer bank failures and more consistent credit provision: consequently the commercial banks may be less vulnerable today to credit or economic shocks’

Which was not just wrong – but 180° wrong. So how do we ensure that we don’t in ten years’ time get it wrong again, going along with a dominant conventional wisdom? Market prices are subject to self-reinforcing herd effects: policymakers and policy intellectuals can be subject to intellectual herd effects; and there is no failsafe way to offset this human tendency to collective error. But we need as best possible to embed challenge into our institutions.

(IMF GFSR reports are here) This is excellent stuff again. Just like the crisis was a result of pessimism herd , the solutions also seem to following the herd  without actually checking whether it has been implemented already ? IMF has been using it and they have not been successful. IMF’s economists like Rogoff and Rajan did raise issues in their independent researches, but GFSR could not (ideally GFSR should have been able to see events as it sees the system as a whole).

Lord Turner is absolutely right in saying if we could not get it right then, what are the chances to get it right in next 10 years? Most of this macroprudential analysis is anyways going to be base on fancy models. Most of these models in turn will be based on the assumptions that have led to problems in the first place - people are rational and markets are efficient. And again we are most likely to get the same results – things are fine and this time is different (only to realise it is not).

This is not to say macroprudential exercise will not work etc. But to just point that it will be no magic potion. Bernanke in his speech had given a reality check on these models and they will be difficult to implement. Developed economies may still manage, it will be very difficult for emerging/developing economies.

Assorted Links

1. Urbanomics provides a snapshot of global trade slowdown

2. MR on how times have changed -earlier firms reported lower profits and now higher profits

3. WSJB on US Budget Deficit

4. Rodrik points to another economist’s works that deserves a reading

5. FMB on forecasting problems