Sunday, July 12, 2015

Woodford vs. the Neo-Fisherians


I was trying to think of a good metaphor for Mike Woodford's role in the macro theory world. Dumbledore? But then I'd have to make someone be Voldemort, and I'm not that big of a jerk. Maybe Ed Witten? But far fewer people know who Witten is than know who Woodford is. I give up. Woodford is Woodford. And right now, at this moment in time, Woodford certainly seems like the most influential person in business-cycle theory. Maybe the most dominant influence since Robert Lucas.

One big challenge to the paradigm Woodford has built - which has won near-universal adoption at central banks - is the Neo-Fisherian idea. This is the idea that holding interest rates low for a very long time will either A) make the economy explode, or B) eventually cause persistently low inflation. Looking at the experience of Japan since 1990, (B) doesn't seem so crazy. John Cochrane explained the Neo-Fisherian idea in an epic blog post back in November, and the idea is supported by a more formal model by Schmitt-Grohe and Uribe. It's a big challenge to the Woodford paradigm because 1) the core of the idea is pretty simple, 2) it seems to fit with recent Japanese and possibly American experience, and 3) it says that central banks working in the Woodford paradigm are achieving the exact opposite of what they intend to achieve.

At the recent NBER Summer Institute, Woodford struck back. Actually that makes it sound too confrontational, since Woodford is the consummate nice guy. What he actually did was to address Cochrane's arguments directly, and give some reasons why he thinks they don't apply.

The question of whether interest rates affect inflation in a Woodfordian way or a Neo-Fisherian way depends on whether people's expectations are infinitely rational. Woodford's new idea - which will certainly be a working paper soon - is that people don't adjust their expectations to infinite order. He essentially puts bounded rationality into macro. He posits a rule by which expectations converge to rational expectations.

So to all you guys who ask "When will behavioral economics have a big impact on macro?" The answer is: Right now. It just did. Behavioral macro is now a reality. (Well, really it was a reality with learning models like Evans and Honkapohja, or even Sargent, but Woodford is using it to think about policy in real time, for big stakes, and his presentation will undoubtedly be influential). 

Anyway, I don't understand everything about the new bounded-rationality Woodford model, but from reading his slides, here's what seems to be happening. A permanent interest rate peg ends up making the economy explode. When the peg begins, people think it's a temporary peg. As it continues, people never quite believe it's permanent, but their estimation of its duration keeps getting longer. This makes expectations of the eventual interest rate (infinitely far in the future) diverge, so the economy basically explodes.

So that's the theory, anyway. It's not clear how well this theory applies to Japan, or to other economies that have had very low interest rates for a while now. It's also not clear how well the macro world will accept a behavioral theory as the workhorse model for monetary policy. I guess we'll see, especially after the paper comes out and people (hopefully) start to fit it to data! In the meantime, expect a response from Cochrane. Should be interesting to watch.


Updates

Cochrane has a brief preliminary response. Here is an excerpt:
This is a particularly important voice, as it seemed to me that standard New-Keynesian models produce the new-Fisherian result. i = r + Epi is a steady state in all models. In old-Keynesian models, it was an unstable steady state, so an interest rate peg leads to explosive inflation or deflation. But in new-Keynesian models, an interest rate peg is the stable/indeterminate case. There are too many equilibria, but if you raise interest rates, inflation always ends up rising to meet the higher interest rate. 
What I can glean from the slides is that Schmidt and Woodford agree: Yes, this is what happens in rational expectations or perfect foresight versions of the new-Keynesian model. But if you add learning mechanisms, it goes away... 
[I]f one has to resort to learning and non-rational expectations to get rid of a result, the battle is half won.
Actually, I'm not sure that this is what Woodford is saying. It's hard to tell from looking at his slides, but it looks like he's saying that if we restrict ourselves to stable paths, the Neo-Fisherian result holds in a rational expectations equilibrium. But the indeterminacy in rational-expectations New Keynesian models might also allow for explosive paths, of the kind that Cochrane calls "old-Keynesian". In fact, I'm fairly certain this is the case, since the "rational bubble" literature shows that explosive paths for inflation are a fairly general result in rational expectations models. I think what Woodford is saying is that if you even slightly relax the assumption of perfect rational expectations, there's no way to get a stable path with a permanent interest-rate peg.

Also: I removed a paragraph making a comparison between Woodford and Nakamura/Steinsson/McKay. Though the problem is similar in the two cases (one is about forward guidance in the infinite future while the other is about belief in a permanent interest rate peg), the two papers use fundamentally different techniques. So the analogy is not a great one, and wasn't that important to the post anyway.

Also: Nick Rowe has a great simplified explanation of the Woodford model. I'm pretty sure he's right, and this is what's going on. It's really impressive how Nick is able to capture monetary econ in these little simplified models...that's a skill I never learned and don't possess. Also, I agree with Nick that if your model isn't robust to an infinitessimal departure from rational expectations, you should be worried.

Also: Scott Sumner comments. He says the problem is the New Keynesian model itself.

Also: Brad DeLong comments. He hypothesizes that Neo-Fisherianism is basically just a face-saving way for economists who predicted QE-->inflation back in 2011 to admit their worldview was wrong without admitting that Paul Krugman etc. were right.

59 comments:

  1. Woodford is probably one of my favorite economists right now, I hope he chairs the Fed some day.

    ReplyDelete
  2. Jonathon Hazell5:59 PM

    Noah,

    I'm always a little confused when people speak of Woodford as the most influential living macroeconomist. My question is - to whom is he the most influential? Macroeconomists or policymakers?

    My sense is that (by publications in top journals, general acclaim) there are quite a few more influential macro theorists right now. Farhi, Werning spring to mind. Maybe Bernanke, from the earlier generation. Woodford has published relatively few articles in top journals. Though this is not necessarily important, it's a good measure of how the elite of the profession think of him.

    Curious to hear your thoughts.

    ReplyDelete
    Replies
    1. I think people mean that he's the academic that policymakers at central banks listen to the most. In terms of influence within the profession, I think he's also very influential, though to some degree it's just because he codified the NK paradigm that people like Gali, Gertler, Rotemberg, and others invented. I don't think we can measure influence in terms of journal publications, but I also don't have a good quantitative yardstick.

      Delete
    2. Anonymous1:34 PM

      Woodford is rank 49 on repec, which is well above Farhi (729) and Werning (1434). He has 83 publications, most of which are in top field or general interest journals: https://ideas.repec.org/e/pwo3.html. I think Noah's description was accurate.

      Delete
  3. Why would low interest rates cause low inflation? I can definitely see how low inflation can cause low rates, but not the reverse.

    ReplyDelete
    Replies
    1. The same way wet streets cause rain.

      Delete
    2. Anonymous4:10 AM

      Hey genius, guess what -- inflation rates and nominal interest rates are both endogenous, so neither causes the other. Every time you post you sound dumber and dumber. Maybe you and Henry could get together and have a dumbfest.

      Delete
  4. Anonymous7:03 PM

    "This makes expectations of the eventual interest rate (infinitely far in the future) diverge, so the economy basically explodes."

    It seems to me that such a theory is too narrow and places too much store on expectations around interest rates. What about expectations around income and other behavioural factors such as rank optimism. Seems more realistic.

    Henry

    ReplyDelete
  5. Anonymous8:00 PM

    "When will behavioral economics have a big impact on macro?"

    The problem with economics is that it has been usurped by people who would feel more comfortable modelling the flow of oil through a permeable sandstone. Human behaviour is very much less tractable in economic modelling than is hydrogeologic behaviour. The need for "scientism" in economics has been driven by technocrats who somehow fell in love with solving the problems of the world. While their methodologies may have solved the operational and logistical problems of military organization in WWII, they translate clumsily into the solution of real world macroeconomic problems. Yet stochastics is the new reality. Correlation is king and economic insight is moribund.

    Henry.

    ReplyDelete
    Replies
    1. Economics is so difficult because people are forward-looking. To solve models you need to deal with this problem. We need models to quantify things and because things don't always work the way intuition suggests because prices move and people respond to those prices.
      Model consistent expectations was a way to close the model and get a solution. As math and computing power increase this is being relaxed and learning models and bounded rationality are two examples.
      The true thing is that it takes a model to beat a model.

      Delete
    2. Anonymous4:11 AM

      Why do you need models to quantify things? For the most part, psychologists get along fine without them.

      Delete
    3. "The problem with economics is that it has been usurped by people who would feel more comfortable modelling the flow of oil through a permeable sandstone. Human behaviour is very much less tractable in economic modelling than is hydrogeologic behaviour."

      The path of individual oil molecules is still intractable. That doesn't prevent quantitative modeling at a macro scale. Similarly the path of individual gas molecules in a container is intractable. But we still have a successful ideal gas law... one that was formulated in the 19th century well before we even knew much about what a molecule actually was or had any computing devices to model large collections of them. Turns out it hardly matters. We just need to count the degrees of freedom of the molecules to apply the law.

      Delete
    4. "psychologists get along find without them." Not all of them apparently.

      Models are a key component of knowing true things reliably. You create them and then continuously check them against reality until you prove they're false. No other technique is more reliable.

      Delete
    5. Anonymous5:23 PM

      "The path of individual oil molecules is still intractable."

      I disagree with the allusion. Human behaviour is even less tractable. Attempting to define human behaviour with stochastic models is an act of delusion. Regarding modelling hydrocarbon flow, I was talking to a mathematician last week as it turns out. He told me had been contracted to do the very same. He told me the results of his modelling were very successful and yielded the predictive results required.

      Henry

      Delete
    6. Anonymous6:14 PM

      "Models are a key component of knowing true things reliably. You create them and then continuously check them against reality until you prove they're false. No other technique is more reliable."

      A correlation may have been found but what economic insight is gained? Probably zero.

      Henry

      Delete
    7. @Henry, suppose there exists a statistical model (actually a full framework) for analyzing certain emergent characteristics of macro economies, and that framework makes predictions (and has made predictions in the past) about what we'll see (or should be seeing now) in several macro economies around the globe, and those predictions are falsifiable (i.e. the author of the model has stated his model is false if the predictions don't work out), and some of those past predictions can be checked today. And the framework continues to produce more predictions, all of which can be checked by anybody in the future. What more can you ask for (except of course that all the predictions come true)? That's a reliable epistemology in action, and I don't see it happening on econ blogs very frequently.

      Delete
    8. Anonymous7:50 PM

      Tom,

      Sure, your prediction can meet the test if the confidence band is wide enough. But when is a prediction a prediction? And what if it doesn't meet the test. It's on to the next iteration of the model. Unfortunately for modellers, reality is volatile and ever changing. Chasing correlation is like a dog chasing its tail. I am not saying this kind of activity is a complete waste of time. I think its value should be kept in perspective. As far as I am concerned economic knowledge has not advanced much since Keynes finished the last sentence of his GT.

      "That's a reliable epistemology in action, and I don't see it happening on econ blogs very frequently."
      I'm sure you can work up a model to predict when that might happen. At least then you will know when to be around to see it.
      :-)

      Henry

      Delete
    9. Henry, my point is that whether that model (really a framework) turns out to be correct or not is not as important as the way it's presented: with falsifiability conditions clearly presented. If you want to ask the author how far outside the error bands constitutes falsification, I'm sure he'll respond. I've asked him several times for other predictions.

      Try getting other econ bloggers to tell you precisely what conceivable future events will demonstrate that their models and theories are wrong. It seems to me that most avoid falsifiability like the plague. They're much better at explaining things after the fact when low and behold it turns out some unobservable factor must have been at play (obviously, else their ideas might be wrong which is inconceivable). Noah has written before about this problem of falsifiability being avoided in the field of econ.

      A common one is "expectations." It goes something like this: open market operations don't matter really, unless the central bank also changes expectations. How do you know if expectations were changed sufficiently by the central bank? Because if the CB did so then they would hit their inflation / price level / NGDP level target... and BTW, that's the ONLY way to tell if the proper expectations were generated.

      "Expectations" in this context is the mystery ingredient that explains away any problem, yet conveniently isn't independently observable so that we can put this mystery power of expectations to the test. At least not observable in any way that it's proponents would agree would let you falsify their "models."

      Delete
    10. Some of Noah's previous posts on the subject of falsifiability.

      Delete
    11. Anonymous5:43 PM


      Tom,

      “...my point is that whether that model (really a framework) turns out to be correct or not is not as important as the way it's presented: with falsifiability conditions clearly presented. If you want to ask the author how far outside the error bands constitutes falsification, I'm sure he'll respond. I've asked him several times for other predictions.”

      In what way does this validate a model? How long will the model be valid for (however valid is defined)? – no-one can say. As far as I am concerned it’s called moving the goal posts. Models contain statistically defined error terms which allow for random exogenous events. It might constitute sound modelling practice but it is not economics. Economics is dead.


      “A common one is "expectations." It goes something like this: open market operations don't matter really, unless the central bank also changes expectations. How do you know if expectations were changed sufficiently by the central bank? Because if the CB did so then they would hit their inflation / price level / NGDP level target... and BTW, that's the ONLY way to tell if the proper expectations were generated.”

      I’m not sure I agree with the last sentence. Who knows what happens in a real world situation. The REH is based on the notion of the CB second guessing how the “independent agent” behaves. What if in reality the “independent agent” second guesses what the CB is doing. And on it goes. What is the end result? How can what happened be discerned? As soon as a model is defined by its assumptions, reality is out the window.


      “ "Expectations" in this context is the mystery ingredient that explains away any problem, yet conveniently isn't independently observable so that we can put this mystery power of expectations to the test. At least not observable in any way that it's proponents would agree would let you falsify their "models." “

      When you say expectations are not observable I suspect you really mean they can’t be modelled. And it’s almost as if it can’t be modelled it doesn’t exist. This is where economics insight should enter the equation.


      Henry

      Delete
    12. "How do you know if expectations were changed sufficiently by the central bank? Because if the CB did so then they would hit their inflation / price level / NGDP level target... and BTW, that's the ONLY way to tell if the proper expectations were generated."

      Nonsense. There's a much simpler way: just ask what the expectations actually are. TIPS exist. So do investors.

      Suppose during a period of low interest rates, a CB credibly announces that starting in five years, they will pursue a 10% inflation target with all available means, but there is no change in policy until then. If your theory says this has no effect on 10-year interest rates then your theory is wrong. Expectations are not ineffable, they're obvious.

      Delete
    13. @Dave, regarding TIPS, take a look here.

      Delete
    14. Sorry, not interested in "Information Transfer Economics."

      Delete
    15. Anonymous3:43 AM

      Wow, Henry is kind of dumb and that retarded Information Transfer guy just will not go away.

      Delete
    16. Anonymous7:54 AM

      At least I'm not completely dumb in your estimation. However, I have no doubt that with due deliberation you will conclude that I am completely dumb.

      Anyway, do you have anything sensible to contribute along the lines of:

      Equilibrium Price = Expected Price from prior period + error term.

      That sort of thing got someone a Nobel Prize.

      Given the penetrating intellect you display, I'm sure you can do even better.

      Henry


      Delete
    17. Anonymous2:56 PM

      "Henry is kind of dumb"

      I would actually appreciate seeing your reasons for saying so.

      Henry

      Delete
  6. In all seriousness - who thinks far into the future? Business mangers focus on the current quarter. Workers focus on the current pay period.

    I'll posit that the interest rate awareness of most people is exactly nil. They live in a cash-flow - i.e. paycheck-to-paycheck - world. How much is the monthly payment for this item going to be, and if i make this purchase, can I still afford to feed my cat? That is practical economics for at least half the population.

    And that's for people who actually have some DISPOSIBLE income. People at up to maybe 2x the poverty level have little or none, ditto net worth.

    When economists talk about expectations, whose expectations do they mean?

    Other economists? Business leaders? Speculators? The Fed? Or the ca. 90% of us who live in the real world?

    And why would anyone ever expect any of them to be rational?

    Full Disclosure: I do not have cat.

    Cheers!
    JzB

    ReplyDelete
  7. Nice post. It got a little confusing but let me see if I can help exposit the confusing part.

    "The idea that keeping interest rates low for a "very long" time will either a. cause the economy to explode or b. cause persistently low deflation."

    In a rational expectations model with choice "a" "very little long" means a finite period of time as short as 2 years. Explode simply means unreasonably large numbers. If very long in "a" meant infinity this is not true. A forever interest rate peg has multiple equilibria, but is stable. But "b" is a steady state phenomenon which means "very long" means infinity. No 10,000,000,000 years is finite. Weird but true. There are basically two steady states. This occurs because there is a discontinuity in the policy rule because of the zero lower bound for interest rates. The low inflation equilibria occurs because of the jump and the Fisher equation. Nominal rates = real rates + inflation. If every one expected interest rates to be at zero forever, inflation = -r.
    President Bullard popularized this argument and heuristically argues this would occur with a long finite period of time. But that was not modeled. It really is an infinity problem in models.
    Many consider "b" idea fragile because if the central bank could commit to some T no matter how large where rates will rise before then this equilibrium disappears.
    This shows the weirdness of infinity in these models.

    ReplyDelete
  8. Anonymous4:09 AM

    As a mathematically literate outsider, I think it's shocking that these people are put in places of influence.

    The behaviour of interest rates at infinity? The emperor has no clothes.

    ReplyDelete
    Replies
    1. Anonymous1:22 PM

      "As a mathematically literate outsider"

      a.k.a "as someone who has literally no idea what they're talking about"

      Delete
    2. First anon, so you don't believe in using boundary conditions, eh?

      Delete
    3. Anonymous4:09 AM

      first anon here. Let me answer your question with a question. Do you think every field of inquiry is amenable to mathematical analysis? One could conceivably argue this case in a reductionist, we're-all-made-of-protons kind of way.

      However if you accept that mathematics has limitations, in practice if not in principle, I think it's only reasonable that we question whether fundamental mathematical analysis is the best tool for understanding the aggregate behaviour of large groups of people.

      Delete
    4. Anonymous2:06 PM

      I too am a mathematically literate outsider, and I have two thoughts: expectations is such a big part of the game in econ that if you don't have boundary conditions you've essentially given up, which seems to be what first anon is advocating here. Also, Witten is recognized by far fewer people than Woodford? This would surprise me. How many BBC or Nova specials have mentioned Woodford? Maybe you meant econ blog readers specifically in which case OK.

      Delete
    5. first anon here. Let me answer your question with a question. Do you think every field of inquiry is amenable to mathematical analysis?

      Let me answer YOUR question with a question. Why do you think fat cats and fat rabbits are cute, while fat dogs and fat mice are generally not cute?

      I mean, as long as we're answering questions with questions...that is one that I have seriously wondered about.

      I think it's only reasonable that we question whether fundamental mathematical analysis is the best tool for understanding the aggregate behaviour of large groups of people.

      Maybe so. But that's hardly settled enough to take one piece of said math and declare that "the emperor has no clothes", don't you think?

      expectations is such a big part of the game in econ that if you don't have boundary conditions you've essentially given up, which seems to be what first anon is advocating here.

      Yeah. I mean, you can do OLG, or assume short horizons, but you still have boundary conditions.

      Also, Witten is recognized by far fewer people than Woodford? This would surprise me. How many BBC or Nova specials have mentioned Woodford? Maybe you meant econ blog readers specifically in which case OK.

      I don't watch these newfangled television machines. Give me my good old-fashioned social media sites any day.

      Delete
  9. ReturnFreeRisk8:04 AM

    Good article.
    Woodford fails to look at the downside of perpetually low interest rates. Mechanically, ZIRP funds all kinds of useless investment and turns a good high productivity economy into a speculative low productivity growth economy. The US Fed is doing it right now and has been since Bernanke joined the Fed in 2001. This delusion of perpetually low real interest rates as being the cure for secular stagnation and NOT the cause for it will be the downfall of Woodford and his theories.

    ReplyDelete
    Replies
    1. ReturnFreeRisk8:05 AM

      as evidence, see Japan. A 4% economy got transformed into a 1% economy (even before the demographic hit)

      Delete
  10. Obviously very interesting from a theoretical point of view. But what is the solution? I think the empirical evidence is there that interest rate hikes that happen too soon, will revive a recession. Maybe low-interest rate economies grow slower because their recessionary periods are less harsh? In the long run, growth isn't all that different between low-rate and high-rate economies?

    ReplyDelete
    Replies
    1. ReturnFreeRisk7:45 AM

      Not raising rates will not guarantee continuation of recovery. Brewing bubbles will burst (are bursting) and take the economy down again. Fed has trapped itself listening to people like Woodford.

      Delete
  11. This may look to you like "striking back," but it seems to me Woodford is just going off in another direction. He's not addressing Cochrane's critique. You may find this direction exciting, but I feel like I'm living in 1975.

    ReplyDelete
    Replies
    1. A) He's definitely addressing part of Cochrane's critique. Cochrane says "If Neo-Fisherianism is wrong, a Fed interest rate peg makes the world explode." Woodford says "Sure, yeah, the world explodes, and here's why it explodes."

      B) Why do you feel like you're living in 1975? I was dead at that point, so I don't remember...

      Delete
    2. "Sure, yeah, the world explodes, and here's why it explodes."

      Sure, yeah, and the problem is the "here's why."

      " Why do you feel like you're living in 1975?"

      That's the way a lot of people did macro then. Reduced forms and "adaptive" expectations.

      Delete
    3. This notion that the world explodes with an interest rate peg is bizzarre. The problem with an R peg is there are too many stable eigenvalues. You only get the world exploding without forward-looking behavior. I agree with David. I am in a very scary time warp with such statements.

      Delete
    4. Sure, yeah, and the problem is the "here's why."

      Write a blog post about it! Blog POST. Blog POST.

      This notion that the world explodes with an interest rate peg is bizzarre.

      What it would really mean would be that the models would stop working, because things would go outside the bounds of the model's domain of applicability. It's not that bizarre to think a price control would cause a market to break down if maintained for long enough.

      Delete
    5. "Write a blog post about it! Blog POST. Blog POST."

      Sometimes there are better things to do. This isn't as earth-shaking as you seem to think.

      Delete
    6. Oh come ON, Steve! How many of your blog posts are complaining about Krugman? Surely Woodford has a bigger impact on the world of academic macro than Krugman. Plus Neo-Fisherism is your thing!

      Delete
  12. I awoke with a fright this morning with worry about what happens at t=infinity. But then I realized the universe will experience heat heat in 10^100 years, which was a great relief!

    ReplyDelete
    Replies
    1. Anonymous5:26 PM

      Entropy rules.

      Henry.

      Delete
    2. Shoot, not "heat heat" but "heat death."

      Delete
    3. @Henry, you write "entropy rules." I think that might more applicable than you think. See my response to you above.

      Delete
  13. Anonymous9:14 PM

    "The question of whether interest rates affect inflation in a Woodfordian way or a Neo-Fisherian way depends on what people think is going to happen at infinity. Woodford's new idea - which will certainly be a working paper soon - is that people don't really look all the way to infinity."

    Keynes had a simple solution to the problem - as we know, he said in the long run we are all dead.

    Henry

    ReplyDelete
  14. This comment has been removed by the author.

    ReplyDelete
  15. I like this. It's a step closer to reality relative to previous academic models, and I'm on the side of reality over simplicity. I'm not sure it will be enough closer to reality to make big inroads.

    But I think some common sense observations are the best way to refute the idea that low interest rates cause low inflation, or what you call the "Neo-Fisherian" theory. I don't like the name because Fisher didn't believe any such thing.

    The only sensible argument for low interest rates causing lower inflation is that lower rates foster lower expectations of growth and inflation. So if that's true we should observe markets adjusting downward their expectations for growth and inflation after interest rates are lowered or QE is announced or rates are promised to be low. A review of the market data will agree with the traditional theory, that loosening of monetary policy increases expectations of growth and inflation, but not by much.

    And here's where common sense needs to come in. Central bankers are considered to have very well-informed, strong analytical teams, at least in most developed markets. So their optimism or pessimism can somewhat move the market consensus. But they're not gods, or even philosopher kings. They're not possessed of any deep secrets or super-human insight. They're never very much against consensus and wouldn't be taken seriously if they were.

    I think explanations for the seemingly small-to-zero response of growth and inflation to monetary loosening are to be found in other common-sense ways:

    - In most developed countries >40% of spending is public sector, and public spending does not much respond to low rates (due to fiscal conservatism in the US, the same and currency union in Europe, and fiscal laxity being already at its perceived limit in Japan).

    - Population graying and slowing/reversing of growth. Crossing the line of 66 years after V Day was a very big macroeconomic event.

    - If interest rates are lowered without any corresponding weakening of expectations, they will be capitalized into real estate and equity prices, partly negating the reduction of investment costs.

    - Market funding rates are mostly long-term and not all that closely related to the short- to mid-term course of policy rates. Bank lending and mortgage bond regulations have tightened.

    I could go on but those seem enough. Not easy to work all that into a model, but these are factors that have to be taken into consideration if you want to get closer to reality.

    ReplyDelete
  16. Jesus, it's not so fucking complicated; The Fed keeps rates low coz they rightly scared shitless that raising them will precipitate a recession. The low interest rates are certainly partly responsible for financial instability building up but only partly. And they certainly don't explain why we went into the shit in the first place.

    Basically, interest rates are not nearly as powerful a variable as we think it is. Except in the financial speculative world and when that world suddenly interacts with the real world, usually via crises.

    ReplyDelete
    Replies
    1. Anonymous3:46 AM

      In this post, Frederic makes unsubstantiated and unsupported assertions. News at 11.

      Delete
  17. Anonymous6:56 PM

    If I could nominate you for the Nobel Prize for Economics I would. (May be even the prize for Literature.)

    Henry

    ReplyDelete
  18. Rates are a function of inflation expectations. If inflation expectations are low, interest rates will be low. An American buying a 30-year bond in 1983 had very different expectations than in 2015. A Japanese investor has expected low inflation for a longer time.

    Unfortunately a lot of mysticism has crept into the profession, with ZLBs and liquidity traps and other means of obfuscating the simple fact that CBs can always inflate, but this seems mainly a function of the fact five of the six richest counties in America now adjoin DC.

    ReplyDelete
    Replies
    1. Anonymous10:31 PM

      CBs can always inflate? Then why haven't they?

      (I understand why they may not want to now, but we sure could have used some 4-5% inflation since the late financial crisis. Wasn't quantitative easing supposed to produce inflation?)

      Delete
  19. Apart from weighing in on the ZLB and Neo-Fisherian discourse, WGS make a subtle contribution to the pure behav econ camp: separation of bounded rationality analysis into "long enough" and "not long enough" is basically a horizon selection problem which is still an open question in the modern post-Prospect theory world. They don't quite solve the "optimal horizon" issue, and I don't think they intended to find the cut-off period after which bounded rationality equilibria diverge (if such fixed point exists), but on a conceptual basis their idea is basically an application of the horizon selection problem to monetary economics. In principle, it may have a very large impact on the way we view asset pricing and even as simple as market volatility. Assets are priced by markets under a set of beliefs and expectations on future path of policy. If policy horizon is long enough, beliefs across agents DO NOT converge and, as WGS claim, diverge explosively. So, the probability of severe asset mispricing when commitment time is long is presumably higher. This leads directly to the argument that stocks and indeed fixed income instruments in certain parts of the world are severely mispriced now. Periods of volatility hikes can be rationalized as moments when WGS' Bayesian solution "explodes", or when multiple beliefs collide in an event of unbounded uncertainty. Either way, lots and lots of things to play around with.

    ReplyDelete