Why Wall Street shorts economists and their DSGE models

22 Jan, 2014 at 11:15 | Posted in Economics | 34 Comments

Blogger Noah Smith recently did an informal survey to find out if financial firms actually use the “dynamic stochastic general equilibrium” models that encapsulate the dominant thinking about how the economy works. The result? Some do pay a little attention, because they want to predict the actions of central banks that use the models. In their investing, however, very few Wall Street firms find the DSGE models useful …

titanic-sinking-underwaterThis should come as no surprise to anyone who has looked closely at the models. Can an economy of hundreds of millions of individuals and tens of thousands of different firms be distilled into just one household and one firm, which rationally optimize their risk-adjusted discounted expected returns over an infinite future? There is no empirical support for the idea. Indeed, research suggests that the models perform very poorly …

Why does the profession want so desperately to hang on to the models? I see two possibilities. Maybe they do capture some deep understanding about how the economy works … More likely, economists find the models useful not in explaining reality, but in telling nice stories that fit with established traditions and fulfill the crucial goal of getting their work published in leading academic journals …

Knowledge really is power. I know of at least one financial firm in London that has a team of meteorologists running a bank of supercomputers to gain a small edge over others in identifying emerging weather patterns. Their models help them make good profits in the commodities markets. If economists’ DSGE models offered any insight into how economies work, they would be used in the same way. That they are not speaks volumes.

Mark Buchanan/Bloomberg

[h/t Jan Milch]

Splendid article!

The unsellability of DSGE — private-sector firms do not pay lots of money to use DSGE models — is a strong argument against DSGE. But it is not the most damning critique of it.

In the basic DSGE models the labour market is always cleared – responding to a changing interest rate, expected life time incomes, or real wages, the representative agent maximizes the utility function by varying her labour supply, money holding and consumption over time. Most importantly – if the real wage somehow deviates from its “equilibrium value,” the representative agent adjust her labour supply, so that when the real wage is higher than its “equilibrium value,” labour supply is increased, and when the real wage is below its “equilibrium value,” labour supply is decreased.

In this model world, unemployment is always an optimal choice to changes in the labour market conditions. Hence, unemployment is totally voluntary. To be unemployed is something one optimally chooses to be.

The D WordAlthough this picture of unemployment as a kind of self-chosen optimality, strikes most people as utterly ridiculous, there are also, unfortunately, a lot of neoclassical economists out there who still think that price and wage rigidities are the prime movers behind unemployment. DSGE models basically explains variations in employment (and a fortiori output) with assuming nominal wages being more flexible than prices – disregarding the lack of empirical evidence for this rather counterintuitive assumption.

Lowering nominal wages would not  clear the labour market. Lowering wages – and possibly prices – could, perhaps, lower interest rates and increase investment. It would be much easier to achieve that effect by increasing the money supply. In any case, wage reductions was not seen as a general substitute for an expansionary monetary or fiscal policy. And even if potentially positive impacts of lowering wages exist, there are also more heavily weighing negative impacts – management-union relations deteriorating, expectations of on-going lowering of wages causing delay of investments, debt deflation et cetera.

The classical proposition that lowering wages would lower unemployment and ultimately take economies out of depressions, was ill-founded and basically wrong. Flexible wages would probably only make things worse by leading to erratic price-fluctuations. The basic explanation for unemployment is insufficient aggregate demand, and that is mostly determined outside the labour market.

Obviously it’s rather embarrassing that the kind of DSGE models “modern” macroeconomists use cannot incorporate such a basic fact of reality as involuntary unemployment. Of course, working with representative agent models, this should come as no surprise. The kind of unemployment that occurs is voluntary, since it is only adjustments of the hours of work that these optimizing agents make to maximize their utility.

To me, this — the inability to explain involuntary unemployment — is the most damning critique of DSGE.

Added 23:00 GMT: Paul Davidson writes in a lovely comment on this article:

In explaining why Samuelson’s “old” neoclassical synthesis Keynesianism and New Keynesianism theories have nothing to do with Keynes’s General Theory of Employment, I have continually quoted Keynes [from page 257 of The General Theory] who wrote “For the Classical Theory has been accustomed to rest the supposedly self-adjusting character of the economic system on the assumed fluidity of money-wages; and when there is rigidity, to lay on this rigidity the blame for maladjustment … My difference from this theory is primarily a difference of analysis”. This is in a chapter entitled “Changes in Money Wages” where Keynes explain why changes in money wages can not guarantee full employment.

When in a published debate with Milton Friedman in the JPKE – later published as a book entitled MILTON FRIEDMAN’S MONETARY FRAMEWORK: A DEBATE WITH HIS CRITICS I pointed out this chapter of the General Theory to Milton his response was that Davidson refers to many chapters in the back of the General Theory that have some interesting and relevant comments – but are not part of Keynes’s theory, while fixity of wages and prices are essential to understanding Keynes.

In a verbal discussion with Paul Samuelson many years ago, I pointed out this chapter to Samuelson. His response was he found the General Theory “unpalatable” but liked the policy implications and therefore he [Samuelson] merely assumed the General Theory was a Walrasian system with fixity of wages and prices!

34 Comments

  1. “…the inability to explain involuntary unemployment”

    Is it possible that this is not the case for economies like China and it does hold for economies which depend on economies like China, i.e. for economies in which kids play with ipads all day long instead of learning how to produce hard goods?

    I’m just asking…

  2. “Obviously it’s rather embarrassing that the kind of DSGE models ”modern” macroeconomists use cannot incorporate such a basic fact of reality as involuntary unemployment.”

    This seems to be your favorite complaint, but it’s simply not true. Labor market with search and matching (where unemployment is *not* voluntary) has been introduced to RBC models already in 1990s (this note [1] has citations and a helpful diagram), and more recently combined with New-Keynesian DSGE models by Blanchard, Gali and others.

    [1] http://weber.ucsd.edu/~vramey/research/Search_and_Matching_Lit_History.pdf

    • Hey dude… any luck getting those private firms interested in, oh I don’t know, ANYTHING you or your colleagues do?

      No? Haha! Continue suckling at the teat of the public sector then, I suppose…

    • Also, I don’t think that was the sort of involuntary unemployment that Syll was talking about…

      • Hi Phil. So you’re defending free markets and marginal theory of value now? And worried about my employment prospects? I’m touched by such open-mindedness and empathy, really 😉

        Though if you also have anything actually relevant to add about the topic at hand (i.e. treatment of labor markets in contemporary macroeconomic models), feel free to do so as well.

  3. Also, I thought that this comment was a nice response:
    _________

    “Page after page of professional economic journals are filled with mathematical formulas leading the reader from sets of more or less plausible but entirely arbitrary assumptions to precisely stated but irrelevant theoretical conclusions.”

    “Year after year economic theorists continue to produce scores of mathematical models and to explore in great detail their formal properties; and the econometricians fit algebraic functions of all possible shapes to essentially the same sets of data without being able to advance, in any perceptible way, a systematic understanding of the structure and the operations of a real economic system.”

    Wassily Leontief (Nobel laureate, Economics, 1973), 9 July 1982, Letter to the Editors (re: Academic Economics), Science, Vol. 217, pp. 104-105

  4. Yeah, lots of DSGE models have involuntary unemployment.

    • Sure there are elaborations on the DSGE workhorse model. The problem, however, with these “New-Keynesian” DSGE models by Blanchard, Gali and others, is that they still have obvious shortcomings from a Keynes perspective. The model presented by e.g. Gali et al. (2007) shows behavioural patterns which, although at least a little reminiscent of Keynes’ line of thought, however [as I’ve argued e. g. my paper on microfoundations — Micro versus Macro — in the latest issue of Real-World Economics Review (January 2014)] builds on mechanisms that still are inadequate. The demand for goods may increase more strongly than in the basic DSGE model, leading to an increase in both nominal and real wages (because of the the relatively greater inertia in prices). As a result, people supply more labour in the labour market. But although this might look a little Keynesian, the mechanism behind it certainly isn’t. The changes in employment rates are still due to labour market reactions — people voluntarily change their supply of labour. So even with these elaborations — unemployment is still voluntary. And that’s the problem, since reality shows us something quite different.

      • Do you have any idea how ignorant you are? In many search models, people cannot find work (immediately) even if they lower their wage demands to zero. Moreover, as Roger Farmer has shown, it is easy to generate multiple equilibria, with very Keynesian (General Theory) properties. What nonsense you have written here.

        • I guess you would have said the same if commenting on e. g. Olivier Blanchard, when he a couple of years ago wrote: “One striking (and unpleasant) characteristic of the basic New Keynesian model is that there is no unemployment! Movements take place along a labor supply curve, either at the intensive margin (with workers varying hours) or at the extensive margin (with workers deciding whether or not to participate). One has a sense, however, that this may give a misleading description of fluctuations, in positive terms, and, even more so, in normative terms: The welfare cost of fluctuations is often thought to fall disproportionately on the unemployed.” So re ignorance and nonsense — well, why not just leave it to the readers to decide for themselves …

          • OK, Lars, fine, but still, you can’t say that the main problem with DSGEs is that they fail to include something that many of the most prominent, famous DSGEs do actually include. It looks like instead of one big unified critique you have a bunch of smaller critiques, and that’s fine, but it’s just not right to say that DSGE models don’t include involuntary unemployment. It’s just not correct.

            • Noah, of course I can’t vouch for every possible DSGE model (who can?), but please, do also notice what I did write (emphasis added):

              “In the basic DSGE models the labour market is always cleared – responding to a changing interest rate, expected life time incomes, or real wages, the representative agent maximizes the utility function by varying her labour supply, money holding and consumption over time. Most importantly – if the real wage somehow deviates from its “equilibrium value,” the representative agent adjust her labour supply, so that when the real wage is higher than its “equilibrium value,” labour supply is increased, and when the real wage is below its “equilibrium value,” labour supply is decreased.

              In this model world, unemployment is always an optimal choice to changes in the labour market conditions. Hence, unemployment is totally voluntary. To be unemployed is something one optimally chooses to be.”

              Nothing in the discussion on this issue for the last years has made me change this opinion one single millimitre!

              • How about the search models, where people look for a job and can’t find one at all? Isn’t that involuntary unemployment? And wasn’t that the central feature of a model that won the “Nobel Prize” a couple years ago?

          • And said readers could use a bit of context. Because, you see, Blanchard immediately continues:

            “The first question is then how to think about and introduce unemployment in a macro model. Here, fortunately, we can build (and are building) on a parallel effort, developed over the past twenty years by, in particular, Peter Diamond,Chris Pissarides, and Dale Mortensen (thus, the name DMP model). In this approach, unemployment arises from the fact that the labor market is a decentralized market, where, at any time, some workers are looking for jobs, while some jobs are looking for workers”.

            Which is exactly what people have been doing.

            • Oh indeed, these guys — “Nobel Prize” winners in 2010 — have been busy elaborating their “rigorous” and “precise” models!
              The problem, however, as usual, is that they — as e.g. Pissarides ”Loss of Skill during Unemployment and the Persistence of Unemployment Shocks” QJE (1992) gives ample evidence of – are as a rule constructed without seriously trying to warrant that the model immanent assumptions and results also are applicable in the real world. External validity (“parallellism”) is more or less a non-existent problematique sacrificed on the altar of model derivations. This is not by chance. For how could one even imagine to empirically test assumptions such as Pissarides ”model 1″ assumptions of reality being adequately represented by ”two overlapping generations of fixed size”, ”wages determined by Nash bargaining”, ”actors maximizing expected utility”,”endogenous job openings”, ”jobmatching describable by a probability distribution,” without coming to the conclusion that this is — in terms of realism and relevance — nothing but nonsense on stilts?
              If you’re really interested in evaluating these models from a science-theoretic and methodological point of view, I recommend Nancy Cartwright’s ”Probability machines: chance set-ups and economic models” (chapter 7 in her book The Dappled World (CUP 1999)).

              • Moving the goalposts, are we?

                • I prefer to think of it as information on the kind of theories and models we are talking about here. When you stick your nose deep down into technicalities, you sometimes, unfortunately, also loose sight of the essentials. If that contextualizing perspective is moving goalposts, well, be my guest 🙂

  5. “Obviously it’s rather embarrassing that the kind of DSGE models ”modern” macroeconomists use cannot incorporate such a basic fact of reality as involuntary unemployment. Of course, working with representative agent models, this should come as no surprise. The kind of unemployment that occurs is voluntary, since it is only adjustments of the hours of work that these optimizing agents make to maximize their utility.”

    I’m afraid that you are the one that will be embarrassed once you discover how completely wrong you are in this claim. There are many DSGE models (e.g., with labor market search) that display involuntary unemployment. Please do some reading and then, if you have any academic integrity, please issue a retraction. Thank you.

    • Well, maybe you should do some reading yourself, David. Why not start with a compatriot, macroeconomist Noah Smith, who wrote the following on his blog a couple of days ago (emphasis added):
      “But Eichenbaum also pointed me to this 2006 paper by Canzoneri, Cumby, and Diba of Georgetown (published version here), which provides simpler but more damning evidence against the Euler Equation …
      [T]he Euler Equation says that if interest rates are high, you put off consumption more. That makes sense, right? Money markets basically pay you not to consume today. The more they pay you, the more you should keep your money in the money market and wait to consume until tomorrow.
      But what Canzoneri et al. show is that this is not how people behave. The times when interest rates are high are times when people tend to be consuming more, not less.
      OK, but what about that little assumption that we know people’s preferences? What if we’ve simply put the wrong utility function into the Euler Equation? Could this explain why people consume more during times when interest rates are high?
      Well, Canzoneri et al. try out other utility functions that have become popular in recent years. The most popular alternative is habit formation … But when Canzoneri et al. put in habit formation, they find that the Euler Equation still contradicts the data …
      Canzoneri et al. experiment with other types of preferences, including the other most popular alternative … No matter what we assume that people want, their behavior is not consistent with the Euler Equation …
      If this paper is right … then essentially all modern DSGE-type macro models currently in use are suspect. The consumption Euler Equation is an important part of nearly any such model, and if it’s just wrong, it’s hard to see how those models will work.

      • I am not a macroeconomist.

        • Sorry about that. It should of course have been “Noah Smith, assistant professor of finance at Stony Brook University, often writing on macroeconomic issues.” 🙂

        • “Beyond DSGE Models: Towards an Empirically-Based Macroeconomics

          Paper for presentation at the 2008 American Economics Association Meetings Jan 6,
          David Colander, Middlebury College

          Peter Howitt, Brown University

          Alan Kirman, Université d’Aix Marseille and GREQAM

          Axel Leijonhufvud, University of Trento and UCLA

          Perry Mehrling, Barnard College

          Abstract :
          This paper argues that macro models should be as simple as possible, but not more so. Existing
          models are “more so” by far. It is time to step beyond representative agent, DSGE models and focus
          more on alternative heterogeneous agent macro models that take agent interaction, complexity,
          coordination problems and endogenous learning seriously. It further argues that as analytic work on
          these models continues, policy-relevant models should be more empirically based; policy researchers
          should not approach the data with theoretical blinders on; instead, they should follow an engineering
          approach and let the data guide their theory choice. ”

          Click to access complex%20macro6.pdf

      • Wow. You criticise DSGE for not having involuntary unemployment. I tell you there are versions of the model that do just that. Instead of asking me which papers, you cite Noah Smith’s reference to Eichenbaum’s critique of the Euler equation. Brilliant.

        • Brilliant or not — do indeed feel free to support us with links and references 🙂

    • My understanding is that in labor market search unemployment is voluntary with an involuntary component but cannot be fully involuntary.

      AFAIK involuntary unemployment at the prevailing wage is a fact of life but goes against the predictions of DSGE models. You can start adding auxiliary hypotheses to try to fix it, like it was attempted with the Cartesian model of gravitational vortex, but at some point you will hit a wall and the sound will be loud. Economics has no future as a science and it must be merged with politics. Actually, economics is politics driven by ideology. Anyone using equations to do economics has no respect for mathematics, for logic and for science.

      • “My understanding is that in labor market search unemployment is voluntary with an involuntary component but cannot be fully involuntary.”

        Your understanding is wrong. Unemployment is completely involuntary in these models (i.e. each unemployed individual would like to have a job at the prevailing wage but cannot find one). So please read up on the literature before you pass further judgement.

        “AFAIK involuntary unemployment at the prevailing wage is a fact of life but goes against the predictions of DSGE models.”

        Again that is wrong. See above.

        “Economics has no future as a science and it must be merged with politics.”

        As both your premises are wrong, so it your conclusion.

        This heterodox world is in a very very sorry state. I can only imagine what is must feel like realising that one has devoted your entire career tearing down a falsely held caricature. Is it hubris underlying it? So much regret must eventually follow.

        • You are being too dogmatic. First try to understand and differentiate between assumptions and predictions of models. Most dogmat6ic followers of various models of reality cannot understand the difference. For example, if you assume a condition and then you predict it at the end this is simply a petitio principi. or circular thinking, if your Latin is not fresh.

          To get up to date, I suggest the study by Shimmer that shows these models you purport underestimate unemployment http://www.jstor.org/discover/10.2307/4132669?uid=3738128&uid=2&uid=4&sid=21103288433711

          If they underestimate, there is an involuntary component.

  6. Whose is this Andolfatto guy spewing noise?! Even if you go back to Blanchard’s article, he notes that the search models used to try to bring the issue of employment in is still in its infancy and not entirely convincing!!!!!

  7. […] Why Wall Street shorts economists and their DSGE models Lars P. Syll […]

  8. I find these kind of economics discussions so amazing. To me, it is clear that economics is completely ungrounded, all of the supposedly fundamental concepts mere hypotheses about how to begin interpreting the data.

    Economics data is a subset of historical data. Add all the math and modeling you want, but the basic data only provides raw data points and correlations between them.

    An economy is a very large, very complex, open, evolving system. In a complex system, there are many causes for any effect. In an open complex system, new causes and effects can arise, and may not be noticed. In an evolving complex system, the causes of effects and their weights may change over time.

    There are thus an effectively-infinite number of correlations to be had from the economy. How to decide which ones are most fundamental?

    Certainly it is clear that there is no way to extract a cause-and-effect signal from those correlations. Science has progressed because controlled experiments allow extracting those kinds of relationships, and human understanding does not progress where that is not possible.

    As it is possible to find economists advocating many different courses of action for any particular economic problem, it is difficult to argue that economics has progressed much beyond an ideological stage.

  9. Simon-Wren adds savings to the mix. Why have DSGE models generally in discussing savings as a real thing!

  10. David knows of course darn well that this discussion is *not* about search unemployment, which is frictional unemployment, i.e. not the ‘Keynesian’ kind of thing. It’s about the kind of unemployment we now see in Spain and Greece and increasingly in other EA cuntries – countries with a central bank which uses a DSGE ‘new area wide model’ which *does not* include unemployment. It’s the kind of unemployment which increases when, what’s happening in Greece, the government uses billions upon billions of Euro to shore up balance sheets of banks and lets tax payers pay the bill, which in effect leads to deeply deflationary effects. It’s not just effective demand which goes down – the amount of M-3 money (or the velocity of money, depends) goes down, too, as this money is used for transactions on what Keynes called ‘a zero elasticity of production’. For your information: M-3 in Ireland and Greece is down with around 30% (debts aren’t, though, surprise!, bad debts are going through the roof in both countries)

    But let’s quote Frank Smets himself about this new area wide model (sorry, no link, from a 2011 paper with the title: ‘Unemployment-in-an-Estimated-New-Keynesian-Model’).

    “Over the past decade an increasing number of central banks and other policy institutions have developed and estimated medium-scale New Keynesian DSGE models…. However, as highlighted by Galí and Gerter (2009) and others, one of the shortcomings of these models is the lack of a reference to unemployment. This is unfortunate because unemployment is an important indicator of aggregate resource utilization and the central focus of the policy debate.”

    Let’s forget about his lack of social empathy: unemployment is, in combination with debts, ruining the lives and long term prospects of tens of millions of people in the EA, at the moment, and Smets does not seem to notice. But let’s quote some others:

    “In a series of papers J. Galí (Galí,2011c,b,a; Galí et al.,2011) explained how to include unemployment into New Keynesian DSGE models. Many New Keynesian (representative agents) DSGE models attempt to explain employment and wages. Employment is a demand oriented variable driven by some well identified shocks (related to technology, consumption preferences and foreign trade). Unemployment is neglected. …. The inclusion of unemployment rate into a model seems to have much more power to solve the issue” http://www.nbp.pl/publikacje/materialy_i_studia/144_en.pdf

    There really does seem to be a problem… But the point is: we have to look at the combination of lower wages in combination with a decline of the stock of money when we want to find ‘Keynesian’ unemployment. Do we find this in the DSGE models?

    Aside: it was and is ridiculous to call such models ‘microfounded’ as no micro data are aggregated to derive the macro functions. They are *not* sound, when it comes to this. Of course, a macro Cobb douglas shaped utility function might be an emergent property of modern western economies, which I doubt, but until now I’ve never read any neoclassical economist which even tried to make this plausible – in the end they just assume that the 1937 Samuelson formula, adapted to eternal life, holds for entire economies, too, even without estimating ‘revealed preference’ which, according to S. (read his Noble price lecture), was a condition sine qua non to use utility functions. Oh, can’t resist a Keynes quote which surely holds for the ECB new area wide model:

    ““The Conservative belief that there is some law of nature which prevents men from being employed, that it is ‘rash’ to employ men, and that it is financially ‘sound’ to maintain a tenth of the population in idleness for an indefinite period, is crazily improbable – the sort of thing which no man could believe who had not had his head fuddled with nonsense for years and years. The objections which are raised are mostly not the objections of experience or of practical men. They are based on highly abstract theories – venerable, academic inventions, half misunderstood by those who are applying them today, and based on assumptions which are contrary to the facts…Our main task, therefore, will be to confirm the reader’s instinct that what seems sensible is sensible, and what seems nonsense is nonsense.”

    • Many good points, Merijn. I especially appreciated the Keynes quote 🙂

  11. In this model world, unemployment is always an optimal choice to changes in the labour market conditions. Hence, unemployment is totally voluntary. To be unemployed is something one optimally chooses to be.

    Although this picture of unemployment as a kind of self-chosen optimality, strikes most people as utterly ridiculous, there are also, unfortunately, a lot of neoclassical economists out there who still think that price and wage rigidities are the prime movers behind unemployment.

    I think you’ve hit the nail on the head. An average person is going to think of unemployment as more like a matching problem, search problem, or result of high fixed cost of taking the risk to move to somewhere with more/better jobs which may dry up (they would not use these words) — than a non-geographic economy where falling into a job is more like falling into a hole than locking two proteins together.

    I would relate your critique of the neoclassical point to the simple models where interest-rate adjustments make the market clear. Definitely this can only happen in a simplistic world. And here we run up against the lack of philosophical / history-of-science training of economists, leading to a confusion of the role and meaning of models in argument.


Sorry, the comment form is closed at this time.

Blog at WordPress.com.
Entries and comments feeds.