"New Models for a New Challenge"

Stephen Cecchetti, Piti Disyatat and Marion Kohler on "whether our macroeconomic models are still relevant," and if not, what needs to change:

Integrating financial stability: new models for a new challenge, by Stephen G Cecchetti, Piti Disyatat and Marion Kohler, September 2009: Introduction Reflecting on the financial crisis that is not yet over, it is natural to ask whether our macroeconomic models are still relevant. For all of their elegance and beauty, with their microeconomic foundations and complex endogenous dynamics, they provided the basis for monetary policy that delivered a quarter of a century of stability. The Great Moderation was great - inflation was low, growth was high, and both were stable. At least, that's what we thought. In retrospect, signs of smugness abounded. Academic journals are filled with papers explaining that this stability was, in large part, a result of good policy. And policymakers listened. The economy was inherently stable, with strong self-correcting forces. The financial crashes that were so common before the mid-20th century were banished by our deep and profound understanding that had been translated into mathematical models.

What a difference a year makes!

The models neither stopped the crisis from happening nor provided guidance on how policies could cushion its impact. They failed utterly in guiding our construction of an institutional framework capable of preventing systemic financial failure. Yes, there were warnings.¹ And yes, there were models that hinted at the sources of the difficulties we now face. And yes, the economic reasoning provides the lens through which we can start to understand what happened and why. But, in the end, we ignored the risks.

In this essay, we begin with a brief review of the pre-crisis consensus that provided the basis for stabilization policy as it has been conducted since around 1980. Our main conclusion is obvious: we need to build economic models that integrate the financial sector in a serious way, accounting for the role of intermediaries with all of their linkages, both with each other and with the real economy. And, most importantly, these models must be capable of endogenously creating financial stress that can build up until the pressure leads to a crisis - that is, models in which booms and busts are normal. ...

... 4. Conclusion For macroeconomics, the biggest lesson of the financial crisis is that our models need to find a more meaningful role for finance. Episodes of financial stress are too frequent, and seem too costly, to be treated just as events that are "bad luck" and therefore of little consequence to forward-looking stabilization policy, as suggested by Lucas (2009). Rather, we should ask whether policy can and should intervene to make financial stress less likely and less damaging when it inevitably comes.

While the New Keynesian workhorse models are built around a role for stabilization policies, they appear to have stopped too soon. Understanding how to deliver economic stability must include an understanding of how to avoid financial instability.

Modeling financial booms and busts requires a model where financial imbalances matter for the real economy. As we have suggested in this essay, this means questioning a number of fundamental assumptions of the current workhorse macroeconomic models, including whether capital markets function properly, whether individuals behave rationally, whether we can really rely on the fiction of a representative agent, and whether markets clear.

As daunting a task as this may seem, prospects for progress are encouraging. Not only is there a clear awareness of the challenge (Bean (2009)), but work is already under way: heterogeneous agent models are being solved, bounded-rationality and learning are being actively explored, agent-based models are being simulated, and incomplete financial markets as well as substantive financial frictions are being introduced.

It is our hope and expectation that successfully integrating financial imbalances into models of real fluctuations will yield a toolkit for policymakers. It will guide us in the creation of new stabilization tools as well in the improved use of old ones. It will help us understand how to measure financial stress in real time and allow for transparency and accountability of policymaking in the same way that price measurement is essential for holding inflation targeting central banks accountable. Getting there will not be easy, but then, the challenge to conventional monetary policymaking 50 years ago surely appeared daunting as well. Hopefully, this time it will not take as long to get things worked out.