Monday, January 18, 2016

Please: Lets Agree to Speak the Same Language

Olivier Blanchard finds the drop in the value of American stocks hard to explain in a framework where only fundamentals matter. He concludes that ‘herding’ is to blame.

Can we please agree on terminology? Animal spirits, confidence, sunspots, self-fulfilling prophecies and, sentiments have all been used to mean shifts in markets caused by factors that are non-fundamental. Now Olivier adds herding as one more term. (To be fair, that term too has been used before in the finance literature). Why this smorgasbord of synonyms?

Beatrice Cherrier and Aurélian Saïdi h
ave a nice survey of the history of sunspots. The idea that non fundamental factors can have real effects, was developed at the University of Pennsylvania in the 1980s at about the same time that the Real Business Cycle model took off. According to the RBC school, recessions are times of technological regress caused by shocks to technology. According to the sunspot school, recessions are times when belief shifts move the economy from one equilibrium to another. Initially, both paradigms flourished but the RBC agenda pulled ahead and stayed ahead for thirty years. That is now changing.

Why this divergence in fortunes? According to Cherrier and Saïdi there were three major reasons why the RBC program pulled ahead. 1) there was no single strong individual to promote the sunspot agenda and the three initial leaders, Costas Azariadis, Dave Cass and Karl Shell could not even agree among themselves. Azariadis used the term self-fulfilling prophecies. Cass and Shell used sunspots. 2) The literature on sunspots was technically demanding and the protagonists made no attempt to explain it to a non technical audience. 3) Cass, Shell and Azariadis were not interested in empiricism and they did not make an effort to promote their agenda at central banks or at applied groups such as the National Bureau of Economic Research.

My own work on self-fulfilling prophecies began at Penn in the early 80s and I have made the effort to promote this agenda at central banks and to promote these ideas in a series of books and coauthored papers. My research agenda has been devoted to explaining these ideas to a non technical audience and to providing a link with empirical work. My coauthors on this agenda include Michael Woodford,  Jess Benhabib and Jang-Ting Guo. Why weren’t we more aggressive in promoting the agenda? I can only speak for myself; but I realized early on that, if we accept the idea that business cycles are just autocorrelated disturbances around the natural rate of unemployment, that a sunspot shock is one more disturbance that takes its place alongside productivity shocks, tastes shocks, news shocks and monetary disturbances.

In a survey paper published here, I distinguish between first and second generation models of endogenous business cycles. I use that term to mean models driven by non-fundamentals (yes, I too am guilty of adding one more synonym). In first generation models, sunspots are one more shock that temporarily shifts the economy away from the natural rate of unemployment. In second generation models, movements away from the natural rate are permanent. There is no self-correcting mechanism.

If you are a graduate student or a researcher who is working, or planning to work, in this area, I have a plea. Can we at least agree to add no more words to refer to the same idea? Please: Lets agree to speak the same language and, in so doing, give credit to those who laid the foundations for this agenda.


  1. "Sunspots" is a bad name: because Jevons used it first to mean something quite different; because only experts with telescopes can see sunspots, so they can't be common knowledge. But maybe it's too late!

  2. Great blog, many thanks for this, and will certainly add some of the links to my reading list. Graduate students like myself are very skeptical of the pure instrumental rationality of mainstream models. It is obvious that confidence plays a big role in economies and in driving investment, though the discipline ought to be careful because in some instances, confidence is being appropriated to mean Ricardian equivalence and perfect foresight, when it means the failures of these!
    A good research agenda would be to formalize these theories into simple models and diagrams that can be taught in the classroom.


  3. Sun spots really do impact the climate. The sun's activity modulates cosmic rays which impact cloud formation.

    If you think there are too many ways to say "animal spirits" you could get a laugh out of how many ways there are to say "printing money":

    You may also be amused at how many different (and good) ways there are to explain hyperinflation (which like market crashes has positive feedback loops):

  4. I like "positive feedback loop" as a general way to view things like "self-fulfilling prophecies". I have a good collection of examples of positive feedback loops in nature to help people get the idea:


Note: Only a member of this blog may post a comment.