Sunday, November 08, 2015

"Panics and Bubbles" reading list


Tony Yates offers his reading list on "Panics and Bubbles".

The list has some good stuff on it. Diamond-Dybvig, Geanakoplos, and Cogley and Sargent especially stand out. There's also some good stuff on how "expectation shocks" can cause economic fluctuations - for example, Angeletos and Werning, Farmer, etc.

There were one or two incongruous things. (Money search? What does money search have to do with panics and bubbles??) But mostly good stuff to read.

However, there's also a lot that I think Tony left out. His list skews heavily toward macro and money-based models, usually with rational expectations but with a few learning-based models thrown in. It is also mostly made up of theory papers, with little empirical work. This is understandable, since macro theory is what Tony does. But there are a lot of good non-macro and empirical papers out there on the topic of "Panics and Bubbles". For example:

Harrison and Kreps (1978) model how overconfidence can lead to asset price volatility. Scheinkman & Xiong (2003) follow up. Barber and Odean (2001) provide some evidence.

In fact, there are a bunch of papers, both theoretical and empirical, on heterogeneous beliefs and their role in asset pricing. Here is a good overview by Xiong. Here is a classic 1994 paper by Morris and a classic 1993 paper by David Romer. Here is a good, short, simple discussion paper by Barsky, talking about the Japanese bubble.

Heterogeneous beliefs are closely connected with learning. Tony mentioned some learning-based macro models, but I'd also point people to this elegant 1999 paper by Zeira on how learning can produce bubbles even in very simple settings.

Noise trader models are another important strain of the literature. Start with DeLong et al. (1990) and the various offshoots of that paper. Another very important paper is Abreu & Brunnermeier (2003). Mendel and Shleifer (2012) is yet another good one. Then check out Brunnermeier and Nagel (2004) on hedge funds and the technology bubble for some evidence.

A mostly forgotten but incredibly interesting strand of research was the "information cascades" literature, that models bubbles as herd behavior. Check out papers by Avery and Zemsky (1998), Chari and Kehoe (2003), and Park and Sabourian (2009).

In terms of direct tests of bubbles, and why these are so hard to do, you'd want to check out the classic 1980s literature on variance bounds tests (here is a great brief overview), as well as the literature on other kinds of bubble tests (surveyed here by Refet Gurkaynak).

And of course, check out the surveys on bubbles and crashes by Brunnermeier and by Scherbina and Schlusche.

In terms of the interactions between financial market disturbances and recessions, in addition to Kiyotaki-Moore and Bernanke-Gertler, you should check out more recent papers by Curdia and Woodford, and Christiano and Rostagno. I'm mildly surprised those or others like them didn't make Tony's list.

So there you go. I left out stuff like emerging market crashes and capital flight. I also left out economic history (Kindleberger), non-mathematical treatises (Minsky), and various "heterodox" ideas like Austrian theory.

But I think this is a good start. I think this list also shows that much of the finance theory literature has developed in parallel to the macro literature, with incomplete communication between the two. To some degree I suppose that's inevitable.

16 comments:

  1. Not sure what's puzzling about including some micro-founded monetary theory on the list. Money is a classic example of a rational bubble: an asset that has value because everyone thinks it does, but can crash in value if people stop believing it will be accepted. Bitcoin comes to mind...Simplest and cleanest discussion of this is in this short survey,
    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1012865
    .

    ReplyDelete
  2. An amateur response here

    http://lund.co.uk/what-to-read-about-financial-bubbles-and-crashes/

    ReplyDelete
  3. Robert Flood7:47 AM

    The Shiller-type variance bounds tests were not bubbles tests - they were Euler equation rejections. See my 1986 paper with Hodrick. F&H do not have a proof that the so-called excess volatility tests are equivalent to chi^2 GMM-tests. I think Cochrane proved that somewhere

    ReplyDelete
  4. Your and Tony's lists are much better than I could have done.

    But, off the top of my head, I think I would have started with Bagehot.

    ReplyDelete
  5. Noah, tl;dr, can you summarize you post and all the material in your references in a line or two? Thanks.

    ReplyDelete
    Replies
    1. Sure!

      "Blah blah blah blah BUBBLES!!! blah blah" <-- how is this? :D

      Delete
    2. Yes, that's helpful.

      Delete
  6. Minsky is not non mathematical, he just doesn't do neo classical models. All the equations for the financial instability model are in "Stabilizing an Unstable Economy".

    ReplyDelete
    Replies
    1. Here's a PDF: http://digamo.free.fr/minsky86.pdf

      I scanned through hundreds of pages and found only one brief section containing one or two incredibly simple equations. So I stand by my description of the work... :D

      Delete
  7. Christophe Chamley wrote the book Rational Herds and has a bunch of interesting finance papers.

    http://chamley.squarespace.com/articles

    ReplyDelete
  8. I think his macro work on savings traps and paradox of thrift looks interesting too, deserves more attention

    ReplyDelete
  9. Congratulations on mentioning Kindleberger and Minsky without mentioning them. Nice hat trick there, even better than your fez, :-).

    BTW, Minsky has had more than just two equations, but I fully agree that they are all pretty simple.when he has them, which was indeed pretty infrequently.

    Oh, I could complain about the inadequacy of the reviews of past bubble tests and even egomaniacally point out how some of my own unmentioned papers resolved some of the problems raised, but I shall not go any further than this mentioning that I am not mentioning in doing so, and I am not even wearing a fez... :-).

    Barkley Rosser

    Barkley Rosser

    ReplyDelete
  10. "Manias Panics and Crashes: A History of Financial Crises" by Kindleberger is , regrettably, unreadable to anyone not already familiar with the history of financial crises.

    ReplyDelete
  11. Anonymous9:28 PM

    Rational expectations sounds like a REALLY intelligent way of understanding irrational behaviour in financial markets.

    You have one history book on there and give Minsky a token mention. Are you being serious? And, I am absolutely sure what was on Yate's reading lists before 2008 was very useful in understanding what was coming and it was really worth bringing up again a post that long ago should have been forgotten. I am sure that is great learning for the next bubble.

    ReplyDelete
  12. Anonymous4:19 AM

    This is just a laundry list. Boring. You should have given us a bit of the history of the literature. How did we forget what Keynes said about animal spirits (irrational behaviour) and the importance of the financial sector before the bubble? Do you think putting financial frictions into rational agent models after the crash is the correct way to understand this? Or is that like putting the theory of evolution into the Story of Creation? Do you think we should pay more attention to primary documented evidence which is likely to mostly be done by researchers outside the profession and much is likely to be qualitative in nature?

    "Heterogeneous beliefs are closely connected with learning." What is that supposed to mean? Is that what you get from reading this stuff? This is hardly enticing or enlightening stuff to potential readers. And Sargent? Nobody should be reading Sargent now.

    ReplyDelete
    Replies
    1. For a history of the literature, read Justin Fox' "The Myth of the Rational Market". It's an awesome book.

      People probably forgot about animal spirits because financial markets were really stable in the postwar period until the 1987 crash.

      Most of the models I linked to are not rational agent models. But no, I heavily doubt that "institutional" financial-friction models are going to get much of anywhere.

      "Heterogeneous beliefs are closely connected with learning" means that people disagree, so we want to know if and how they ever reach agreement.

      I admit that I struggle to find any Sargent papers that have had lasting influence besides convincing everyone to use DSGE models and rational expectations.

      Delete