Wednesday, September 10, 2014

Optimal quantity of money, achieved?

Here are three graphs, presenting inflation, long-term interest rates and short-term interest rates in the US, Germany and Japan.




Now, suppose you just returned from a long trip in outer space, started around 1979. What would you say of these three graphs?

If you didn't "know" anything and just look at these graphs, your response would most likely be, "Hoorray!,"at least if you blasted off somewhere near the University of Chicago.  It looks like our economies vanquished inflation and are all on a steady global trend towards the Friedman "Optimal  Quantity of Money."

You might sensibly forecast that the trend, so clearly established for 2 to 3 decades, will continue. Inflation will continue to trend down, to zero or slightly negative values. The short term nominal rate will stay at zero, or maybe rise to at most a percent or two.  Long term rates, read as expected short rates plus a risk premium, signal this future and might end up slightly positive.

You might suppose our central bankers are all off retired to write memoirs at think tanks, enjoying the accolades of a grateful public, and cutting ribbons at statues being built to their honor. You would be wrong, but that's another story.

The Friedman Optimal Quantity and Financial Stability

Milton Friedman long ago wrote a very nice article, showing that the optimum state of monetary affairs is a zero short-term rate, with  slow deflation giving rise to a small positive short-term real interest rate.

Friedman explained the optimal quantity in terms of "shoe-leather" costs of inflation. Interest rates are above zero,  people go to the bank more often and hold less cash, to avoid lost interest. This is a socially unproductive activity.  Bob Lucas once added up the area under the money demand curve to get a sense of this social cost, and came up with about 1% of GDP. Not bad, but not earth-shattering.

As I think about it, however, there are financial stability benefits to zero rates  far beyond what Friedman imagined.  This thought reoccurred this morning as I was thinking about Dan Tarullo's testimony on capital requirements.

Why do banks load up on debt? Well, one answer, interest payments are tax free and dividends aren't, so the "tax shield" leads to excessive debt. But if interest rates are zero, the value of the tax shield is zero, and this incentive to undercapitalization vanishes! 

Positive inflation induces all sorts of pointless tax arbitrage. Close to home, universities issue tax-free bonds, and invest in hedge funds.  But the whole profit-non-profit distortion in investing vanishes if interest rates are zero. If interest rates are zero, and you earn money from deflation, all interest is tax free.

The real costs of inflation are not shoe-leather trips to ATM machines. They are the fragile structures of overnight funding, which built up before the financial crisis, and crashed spectacularly, much of it designed to make sure "cash" earns interest. At zero rates, it is all needless.

Zero interest rates. Zero or slightly negative inflation. It's hard to tell just where long-term inflation is anyway. Would you really trade your imac for 1,000 Apple IIs? What's not to like?

Why not? 

So, why do so many people look at my graphs with deep foreboding and a sense of something wrong? Why is the "optimal quantity of money" and the "non-distorting interest rate" suddenly the "zero bound," as welcome at macroeconomic discussions as an ebola patient in an emergency room? What's wrong with an economy that has zero or slight deflation, and zero or very low interest rates? Why are central banks fighting so desperately to avoid their apparent victory?

One view, espoused frequently by Paul Krugman, sees the quiet approach of zero inflation or deflation with great foreboding, as it puts us in danger of "deflation spiral" or "vortex" about to break out at any time.   A little extra deflation raises real rates, which lowers "demand," which through a Phillips curve leads to more deflation, and the whole thing spirals out of control.

But it never happened, not even in Japan, though feared for nearly 20 years now. I don't know of a single historical event where a deflation "spiral" ever happened. (Deflation has happened, as in the US in the great depression. But it did not "spiral" out of control. It looked a lot like money demand went up, money supply didn't the price level fell, end of story.) And in my view of the world it can't happen. Real rates lowering "demand," are a tenuous idea, the Phillips curve is a correlation not a theory of price level determination specifying cause and effect from output to prices, and a serious deflation means governments must raise taxes to pay off higher real values of debt, which simply is not going to happen.

Another view is that we stand on a cliff of monetary-policy induced inflation or hyperinflation about to break out.  The zero bound is being held too long. Reserves have exploded from $50 billion to $4 trillion. Just wait.

The long trend and calm behavior of the data belie this view too.

A more nuanced view holds that we need positive inflation and positive rates so that the Fed has room to lower rates to ward off deflation spirals, as well as to counteract recessions. I'm dubious. This is like the view that you should wear shoes that are too tight, so it feels good to take them off at night.  A few monetarists have called for deliberately stifling financial innovation so the Fed could control the money supply. The high inflation target so we can lower rates is is the Keyensian (or interest-rateian) analogue. But do we really need to lose 1% of GDP in Lucas shoe leather costs, and the far larger financial stability costs that artificially high rates imply, just so the Fed can jigger around rates when it wants to do so?

At least for inflation, the graphs do not scream the necessity of this view. They certainly do not endorse the view that the disinflationary trend was caused by a Taylor rule: You do not see interest rates moving 1.5 times as much, or in response, to inflation, and you do not see rates dropping more than 1.5 times inflation to ward off deflation.  Producing a coefficient above one takes a lot more fiddling with a regression. You see pretty much a Fisher rule -- interest rates move one for one with inflation. The graphs are just as consistent with the story that talk policy somehow "anchored expectations" and then central banks slowly lowered rates.

Our astronaut, on hearing all these views, might well conclude that none has a good handle on just why inflation is falling to zero, what central banks or other parts of the government actually did to bring about these great trends. And he would be correct. But that emptiness surely means that chicken-little "the sky is falling" about this three-decade trend suddenly exploding is overstated.

(Someone will quickly point out that I too have worried about inflation. But my worries have nothing to do with monetary policy or the level of nominal rates. My worry has to do with fiscal policy, and is more like a worry that low mortgage backed security rates in 2006 could not last. )

What about wage stickiness? A standard answer to "what's wrong with slow deflation" is "wages are sticky so you'll get a secular stagnation." Now, wages arguably are sticky at the 1-6 month horizon, and when we're talking about large, say, 20% shocks, like if a country's banking system implodes.

But that's not what we're talking about here. Does wage stickiness really get in the way of 1-2% steady deflation?

Now, nobody likes to have their wages cut.  But nobody has to. As Alex Tabarrok points out in a splendid Marginal Revolution post, half of US employees have changed jobs since the bottom of the Great Recession.  This is one of many ways in which the popular imagination of having one job all your life butts up against the reality of huge churn in the labor market.

Now stickiness fans will come up with some new story about people not wanting to take lower wages at new jobs, or social limitations to hiring new people at lower wages and so on. But that's a new and different story than "employers don't want to cut people's wages." Again, we're thinking about the long run here, not recessions.

Moreover, each individual can ascend an age-earnings profile while wages overall are declining. And productivity growth adds to the spread between wages and inflation.  If each individual's wages grow 2% per year as they age and move up the ladder, if aggregate productivity grows 2%, then we can have 4% deflation before anyone takes a wage cut.

So what is the problem? Yes, the reduction in inflation is associated with slower growth, see again Japan. But it's far from settled that zero inflation, butting against some sort of stickiness, caused the slow growth and everything else in Japan was a smoothly functioning market.  Anil Kashyap thinks Japan had zombie banks. Fumio Hayashi and Ed Prescott point to low TFP growth.  And similarly with us.

Bottom line

So, back to our graphs and returning astronaut. If you just look at the graphs, I think our astronaut would think there is a good chance this trend continues.  And, perhaps, we should see a long period of zero rates and slight deflation as a great achievement in monetary policy.  If only we honestly understood why it happened and therefore had more faith that it will continue.








49 comments:

  1. If we rely on prices for information, stability of the price level is better than slight deflation. I like the way that Evan Koenig (2011), also Kevin Sheedy, explain our need for monetary policy: it's because we use debt contracts that are not state-contingent, and need some way to manage balance sheets ex post. Some people are looking for ways to remove the zero lower bound. I would much rather go for the first-best option to improve the distribution of risk: innovation in debt contracts.

    ReplyDelete
    Replies
    1. Anwer,

      The market value of equity is continuously determined using all available information including ex post information as well as forecasts of future results.

      Why wouldn't you take the same approach with innovation in debt contracts? Debt contracts should be adjusted based upon both ex post information and forecasts of future results.

      Debt contracts that are adjusted based solely upon "looking in the rear view mirror" are informationally inefficient.

      Delete
    2. Frank I think you are advocating ARMs - adjustable rate loans. But experience shows that it is bad for most households to bear a heavy exposure to interest-rate risk. Prof. Cochrane suggests in his proposal for floating-rate Treasury debt that the government would also use swaps to manage interest-rate risk. Many borrowers don't have such financial sophistication and access to suitable instruments. We need an innovation that would be useful, and comprehensible to them.

      Delete
    3. Anwer,

      See my earlier post about making the value of deposits contingent on potential GDP growth (forward looking counter-cyclical index) and making the value of debt contingent on measured GDP growth (rearward looking pro-cyclical index).

      If only debt is to be indexed (not deposits), then debt would need to be indexed in such a way that it takes into account both past information and future projections. Otherwise, with pro cyclical / rearward looking debt, there is no bottom for it - it's value can fall for the entire term of the debt.

      I know you had mentioned that debt would be indexed to Real GDP and the value of the real economy can never fall to zero. But debt contracts typically are not of the infinite life variety. Continuous losses throughout the term of the loan are bad enough to create zombie banks.

      Delete
    4. Anwer,

      "But experience shows that it is bad for most households to bear a heavy exposure to interest-rate risk."

      I think what you mean to say is that it is bad for most households to bear a heavy exposure to macroeconomic variable risk. If the interest rate that a household pays on its debt is indexed to a macroeconomic variable (like GDP growth), then the financial position of some households situation could worsen during periods of high growth.

      Consider the mortician that borrows to finance his operations. More people dying = less GDP growth = more business for the mortician. The fortunes of some households will be counter cyclical to the index being used.

      Delete
    5. That's a good point, Frank - the choice of a suitable benchmark for indexed debt is important. I just want to note that if we have people borrow Trills, the unit of account is bounded - unlike the currency units we use now, which are not bounded away from zero. The value of Trills due in payment is the present value of the economy, determined by future projections of performance over its lifetime. And a diversified economy should be quite stable in value, especially for moderate discount rates. That's one reason why people are concerned about the downward trend in rates revealed in the charts above. There certainly is no financial death spiral for the real economy; life goes on after a stock market crash, or even a total currency collapse.

      The contracts that we use now - written in nominal terms - expose us to more tail risk than if they were indexed probably. Indeed we have Grumpy economists warning us about the tail risk of inflation. The event itself would favor debtors, but having them insure the event might be inefficient.

      Delete
    6. Anwer,

      "I just want to note that if we have people borrow Trills, the unit of account is bounded - unlike the currency units we use now, which are not bounded away from zero."

      The U. S. Bureau of Economic Analysis calculates nominal GDP (in dollars) from the sale of all goods. It maintains an index of the prices of various goods and with some weighting / hedonic adjustments, back calculates the price change (aka the GDP deflator) over the time period that has elapsed. It also maintains a reference year (currently 2009) when it equates the real GDP level and the nominal GDP level.

      First question: To set a trill equal to a portion of the real GDP level, we need to define a reference year where real GDP is equivalent to nominal GDP - what determines the year to be used?

      Second question: If the value of the unit of account is changing with time based upon the level of real GDP, how do you accurately measure real GDP? It would be like trying to measure how fast an ice cycle grows / melts by comparing it to the ice cycle right beside it.

      Third question: If we have people borrow Trills, and the value of all borrowed Trills exceeds the real GDP level, either the value of those Trills has to fall or the real GDP level has to rise. What prevents the value of the Trills to fall? What limitation is there on the number of Trills that can be borrowed?

      "There certainly is no financial death spiral for the real economy; life goes on after a stock market crash, or even a total currency collapse."

      You really think it is easy to go from a common currency to a barter system? Is that what you are implying? I think you are taking Friedman a little too serious. Money may be neutral in the long run, but the liquidity and divisibility functions provided by a common medium of exchange are definitely not.

      Delete
    7. Frank we already have indexed bonds called TIPS, which means that there are established methods to address the measurement problems that you are noting. Some countries also issue GDP-linked bonds.

      Even if we change the unit of account for loans and deposits, that would not eliminate the dollar, which is the unit of account for risk-free government debt. We would still use that for prices and measurements based on prices. You might want to listen to some of Prof. Cochrane's interviews on the Euro and how it should be preserved as a numeraire, even if some countries in Europe decide to reform their banking systems, possibly by changing units. He is also very capable of handling your asset pricing questions ... but he might require enrolment in his MOOC first!

      Remember that there are 1 trillion Trills, hence the name. I don't know how many of these could be borrowed, but I think you can see that it's impossible to borrow more than all of the wealth that exists. If they don't work as a currency, people will start using a currency that does work. For now, that seems to be the dollar, but the Fiscal Theory of the Price Level tells us that the dollar has a nonlinear relationship with government finances, whereas Trills are a linear function of aggregate wealth, which makes them a good way to reduce the destabilizing effects of tail risks on contracts. If we make the transition to Trills, we always have the option of going back to the cyclical economy described by Keynes and Minsky in their writings, but I doubt that we would want that.

      Delete
    8. Anwer,

      "Remember that there are 1 trillion Trills, hence the name."

      Okay, then there are always and forever one trillion Trills, none can be created or destroyed - ever.

      "I don't know how many of these could be borrowed..."

      In a fractional reserve system of Trills, the same Trill can be borrowed over and over again. There is no limit unless a "central bank" of sorts sets a reserve requirement.

      In a full reserve system of Trills, a Trill can be lent out once and only once.

      "but I think you can see that it's impossible to borrow more than all of the wealth that exists."

      No, I cannot see that. New wealth is created every day. At the same time existing wealth depreciates, wears out, and eventually return to dust. Real GDP measurements only take into account new goods created.

      Under both a fractional reserve and full reserve system of Trills it is possible and indeed likely that the populace will borrow more than all the wealth that exists because:

      1. The same wealth can be leant more than once under a fractional reserve system
      2. Even under full reserve, Trills only take into account new wealth while loans can be made to purchase both new and existing wealth - see housing bubble. Do you think all of those mortgages were written on newly constructed homes?

      "Even if we change the unit of account for loans and deposits, that would not eliminate the dollar, which is the unit of account for risk-free government debt."

      Now you are describing a dual currency system (Trills for private enterprise, dollars for government). IMHO, that creates more problems than it solves.

      "the Fiscal Theory of the Price Level tells us that the dollar has a nonlinear relationship with government finances, whereas Trills are a linear function of aggregate wealth"

      Trills are not a linear function of aggregate wealth. They may be a linear function of new wealth.

      Delete
    9. That's a good point, Frank: Trills are the present value of domestic income that is yet to be earned, and they don't include the value of real assets that already exist. My rudimentary understanding of lending tells me that it is done against future income, which imposes a limit on how much can be borrowed. We currently don't have a good process to adjust loans for changes in prospective income, but indexed debt gets proposed now and then.

      Delete
    10. Anwer,

      "My rudimentary understanding of lending tells me that it is done against future income, which imposes a limit on how much can be borrowed."

      An individual person has a limit on how much he / she can borrow (his / her lifetime income is finite). Infinite life agents (governments, too big to fail financial institutions, monopoly corporations, etc.) are not limited.

      Also, loans can be classified into two categories - collateralized and non-collateralized. Mortgages, car loans, home equity loans on the consumer side and capital improvement loans and business mortgages on the production side constitute collateralized loans. If the borrower defaults, the lender has a legal claim on the good(s) that were purchased with the borrowed funds. Non-collateralized loans (credit cards, interbank funding, commercial paper) are claims on income only.

      Finally, there are different categories (tranches) of debt on the corporate level. The highest level of corporate debt often has legal claim on trademark / patent rights and other intellectual property of a company should the company go bankrupt. This is a type of collateralized debt where the property used as collateral does not have an observable market price.

      With all of that going on, trying to index all private debt becomes problematic. Money is often borrowed against goods that already exist (homes, used cars, etc.) and goods that don't have an observable market price (trademarks, patents, intellectual property, etc.).

      There is a solution. Prior to the evolution of marketable debt (banks retain all loans), the solution was for the central bank to set a reserve requirement. That reserve requirement put a hard limit on how many loans could be made by the banking system. That solution is not viable anymore because banks are permitted to sell their loans to make more loans.

      A solution that could work today would be for the federal government to require all loans to be collateralized (no more credit cards, unsecured debt) while also stipulating the amount / type of collateral that could be borrowed against. A bank could still sell off its loans but it could only re-lend those funds against collateral approved by the federal government.

      Delete
  2. "Friedman explained the optimal quantity in terms of "shoe-leather" costs of inflation. Interest rates are above zero, people go to the bank more often and hold less cash, to avoid lost interest. This is a socially unproductive activity."

    Why would banks lend at zero? Whats the point? It would only happen if they are avoiding neg IOR or avoiding other costs which come about in a dysfunctional economy.

    People placing money at a bank to earn interest is like lending to a bank so it can lend onwards, its not unproductive at all. Intermediation.

    The optimal Q of money is one where prices are stable (0% inflation) with positive rates. No costs of inflation or deflation, perfect stability for forecasting and perfect balance between creditors and debtors for stable real value of debt.

    ReplyDelete
    Replies
    1. The idea was that the nominal rate is 0, but with deflation so real rates are positive so banks do make profit (in real terms) lending.

      Delete
  3. "You see pretty much a Fisher rule -- interest rates move one for one with inflation."

    As you move further up away from the zero bound, the correlation between nominal interest rates and inflation becomes less pronounced precisely because of the tax advantage associated with debt.

    That (and the elimination of housing prices from the CPI in 1983) is how Volcker "conquered inflation". The central bank sets a nominal interest rate. An economy will gravitate towards the cheapest form of financing. As nominal interest rates rise (in both public and private sectors) the tax advantage enjoyed by the private sector grows until private financing becomes less expensive than public financing. Private debt actually crowds out public debt at high interest rates and tax advantaged debt.

    "Yes, the reduction in inflation is associated with slower growth, see again Japan."

    And the reduction in nominal interest rates is also associated with slower growth, see again Japan.

    ReplyDelete
  4. Very interesting.
    Question is, would current economy model allow that? Current economy model want inflation, because then holding money "in your sock" is losing option, and holding it in the bank is only slightly better. People are encouraged to invest. In zero-rate/low-deflation situation, people would save much more. And if people save more, and take less debt, they may even cut on consumerism habits!
    (When I say allow that, I mean there would be strong push from those with money and influence, much stronger than today's to rise rates.)
    As for sticky wages, I read once one economist who talked to some owners and managers about sticky wages some time after the start of the great recession, and they answered that wage cuts are last on the table, because they greatly hurt morale of the workers, so explanation "employers won't cut wages" should explain sticky wages for at least first 6, if not 12 months of the crisis. And from my (very limited) experience (personal and from people I know), employers do like to cut everything else (including bonuses and perks, which does reduce earnings) before touching wages, and people react much more negatively to even 2-3% wage cut than to, for example, 5-year flat wage.
    Interestingly, in some companies and industries in Northern Europe wages were cut relatively early and without much trouble, but in all the cases I have heard of either workers were part-owners (even majority owners) or union was strong, and wage cuts were preceded by clear explanations what management is doing, how it will help, what else is cut and when (in terms of market conditions) wages will grow back.

    ReplyDelete
  5. 1. Note that if deflation comes on the heels of increased productivity of labor, labor, on average doesn't have to fall. People just figured out how to make more shirts with the same amount of work. Shirt prices fall, but wages might even rise.

    2. Note that even in cases of severe deflation, corporate borrowing rates are always going to be postive. That's because debt has an implied short-call, that increases in value with volatility. As long as their is cash-flow uncertainty, the rate will have to be postitive to offset the uncertainty.

    ReplyDelete
  6. I hate to nickpick, but nextime have each short-term, long-term, and inflation rates the same color across each country's graph.

    ReplyDelete
  7. As a practical matter, no major modern economy has flourished at zero inflation. The US economy did pretty well from 1982 to 2007 with mild inflation.
    Then we have the problem of measuring inflation there is no assurance that a measure is accurate over any period of time, such as 5 years or more. Why obsess over a subjective measure of inflation? Should not the point of monetary policy be robust economic growth?
    FInally, how will commercial banks be able to lend on real estate in a deflationary environment?
    A switch to deflation in real estate would eviscerate the financial health of most Americans. I guess equity values would have to fall as well, at least is GDP measured nominally was also falling.
    I cannot understand the obsession of modern economists with microscopic rates of inflation.

    ReplyDelete
  8. This is a great article, Thank you Mr Cochrane again. Thanks for pointing out that mild deflation is good for consumers, that "deflation spirals" never happened. I think what is missing from this article is a criticism of the the idea of "resource slack" that is the prevalent defence to the ones that feel that this scenario is not good, that "we must do something". For these, I leave a question: what is the purpose of the economic system? To make sure "resources are 100% used", or, meet consumers demands ?

    ReplyDelete
  9. one answer, interest payments are tax free and dividends aren't, so the "tax shield" leads to excessive debt. But if interest rates are zero, the value of the tax shield is zero, and this incentive to undercapitalization vanishes!

    I suppose that statement is true so far as it goes - that one particular incentive is gone. However, equity still currently costs about 4% after tax at the corporate level (say 6% before tax at the corporate level) so with debt at zero (or very low) interest rates there is an incentive to maximize debt and minimize equity.

    So far as the difference between +1% inflation and -1% deflation, it seems to me that our definition and measurement of inflation is sufficiently uncertain that the debate is almost meaningless.

    ReplyDelete
  10. But if you were an economics professor from Chicago looking at Japan, you might also look at the following data:

    http://freakonomics.com/2008/04/23/the-economics-of-happiness-part-5-will-raising-the-incomes-of-all-raise-the-happiness-of-all/

    Somehow the very low inflation environment coincides with low life satisfaction, and also with low and occasionally negative real per-capita gdp growth.

    Here's the best way I can build a story around these facts and the editorial in the blog post. There's an exogenous shock in late 80's Japan that causes everyone to be unhappy - big property crash. This unhappiness causes low GDP growth - the unhappy population don't want to get out of bed and GDP growth falls. But unhappy people and depressed economies are more predictable, so the central banker's job gets easier, their forecasts get better, and inflation gets more tameable. Central bankers get inflation under control because central banking is now a doddle. But the central bankers get unhappy too because they're curmudgeonly souls only truly happy when they're pulling their hair out, and decide to wreck everything in a depressive fit by printing money.

    I suspect it might be easier to build the chain of causation the other way.

    ReplyDelete
    Replies
    1. It seems more likely that the story of Japan is: In 1950 you have a country with most of its housing and industry destroyed; a high savings rate; a disciplined population that valued education and hard work; experienced engineers, accountants and managers; access to American markets and technology; low wages and a low standard of living. For about thirty years all those savings get usefully plowed into real estate and industry and the economy grows rapidly. By 1980 - the capital backlog for housing and industry has been filled, wages and technology have caught up to the US; Korea and then China are coming on line as competitors.

      The wonder is not that Japan's growth rate fell. The wonder is that anyone could think it might have continued. Japan has a high standard of living and a low unemployment rate. All of the attempts to "stimulate" the economy are probably doomed.

      Delete
  11. John, and interesting post. You express this sentiment more than once I notice:

    "If only we honestly understood why it happened and therefore had more faith that it will continue."

    I know of an explanation that you may not have heard before. Does it have merit? You might want to check it out see what you think. The idea is based on a generalization of information theory to complex systems. There are admitted limitations to this approach, but macro-economics seems like it might be a perfect fit. The best part (to me) is it largely removes from the equation individual human-scale motivations and their resultant conflicting "just so" stories we often here from economists. The general price level model that falls out of the core hypotheses looks like this:

    P/Po = (alpha/kappa)*(M/Mo)^(1/kappa - 1)

    Where P is the general price level, M is the currency component of the base, and kappa goes as log (M/Mo) / log (NGDP/Mo), and Mo and Po are normalizing constants (Po being truly arbitrary). Honestly, I've forgotten what alpha is, but like kappa, and Mo, I believe it's slowly varying (slow compared to P and M).

    When kappa = 1/2 (a real possibility, and one that matches fairly closely the economies of Russia, Sweden, China, etc), the QTM applies. That's clear if you look at the above formula. But as economies mature and M becomes large compared with NGDP and kappa thus approaches 1 (because kappa ~ log M / log NGDP) , then the QTM no longer applies (Switzerland, US, Japan). Also, there is everything in between (e.g. Canada). This *could be* the missing explanation you're wondering about.

    This is not just curve fitting, although the normalizing constant Mo needs to be fit to the data. A one or two parameter model? Not much to ask, is it? Here's an abstract about the underlying information theory derived model which has been adapted to economics by the author of the above model for P.

    The author has had pretty good success with fitting his models to in-sample data and then "forecasting" out of sample data. And he's compared his model against the empirical data from many different global economies, and against other well established models.

    At least he's got an explanation! And he's humble enough to know that his hypotheses might yet be proved false. And he's put himself out there in terms of making predictions about big trends in multiple world economies: in Europe, Japan, Canada, US, Russia, China, etc. Predictions which he has gone on record as stating expose his hypotheses to the possibility of falsification. And he's not trying to sell you something: he's interested in economics for its own sake.

    Anyway, I think he needs some established economists such as yourself to critique him. He's has some feedback from Scott Sumner, Nick Rowe, David Andolfatto, Noah Smith, David Glasner and others, but nobody has actually tried to understand the information theory related model behind it, AFAIK. This is a shame, because it's not that hard. I'm no genius, and I can understand a good bit of it. I understand that you're a bit of a math wiz: you should have no trouble quickly digesting it.

    ReplyDelete
    Replies
    1. I don't see any 'why' in this theory, it simply pushes the 'why' under the carpet.

      I'm not saying it doesn't have its merits, im actually a big fan...but economics today is about understanding actors, otherwise we won't know what is right. We of course have a moral theory hidden in it (check the homogeneity of degree 0 results) but a hit to one parameter is so far as meaningless as a hit to another.

      Delete
    2. LAL, it sounds like you are familiar with Jason's blog. I'm not sure I fully understand your comment: what's the "hit to one parameter" you write about?

      In terms of "why" I think that's one of the attractive things about Jason's approach: you can answer why without understanding actors. That's kind of his point! The usual approach of looking for answers in human-scale motivations seems to lead to endless stories by different schools of thought: all of which sound great, but which are incompatible: it's hard to sort out the "just-so" type stories from the legitimate explanations.

      As an analogy, in a recent post Jason wondered if the famous physicist Richard Feynman got it wrong when he said:

      "Imagine how much harder physics would be if electrons had feelings!" - R. Feynman

      Perhaps it wouldn't make a whit of difference if electrons had feelings or "free will," or had the ability and motivation to "maximize their personal utility functions." ... maybe none or very very little of that would survive at the macro scale.

      So I don't see Jason as abandoning the principle of telling stories to explain why, but he's trying an approach in which certain kinds of stories that are traditionally told are skipped or deemphasized. However, he does try to find as many correspondences in his results as he can with traditional macro.

      So what do we have instead? How about a nice story drawing an analogy with a pile of sand (for example)?. :D I just happen to have read that one tonight (one of his posts I'd missed).

      I don't see what's wrong with developing a new kind of intuition for this stuff which largely skips the human part. We've had to do it in other fields to make advances: for example quantum mechanics and relativity: very very counter-intuitive... but who cares? You build a new intuition based on what actually works, rather than stick with something that's intuitive but wrong.

      Delete
    3. But economics is fundamentally different from the physical sciences. Whether or not it works is completely undefined without the stories, this is not quantum mechanics this is morally right from wrong. Works means nothing without knowing what is morally good.

      Who cares about predicting money, as long as there are inconsistent stories underneath the predictions, we are just as ignorant as before. We only care whether the inflation rate is optimal to the underlying agents, we do not care about predicting inflation unless we have a coherent theory as to who it helps and doesn't. Otherwise there is no need for economics.

      The parameter hit concept just shows that as the economy evolves parameters will be refit, and the parameters will mean nothing to anyone in the economy, bc we refuse to give them a moral meaning in the first place.

      Delete
    4. "But economics is fundamentally different from the physical sciences."

      Jason might agree with you there, I'm not sure, and his opinion is more valuable than mine. However, I will only concede to it being "very different." And that difference is in every respect strictly a matter of degrees: relative abundance of useful data, repeatability of "experiments," numbers of available "laboratories" or test subjects, legal and financial constraints, and of course, degrees of complexity. (And one other factor: us being, in some sense "too close to the problem" and thus too easily beguiled by our own self-deceiving sense of "we're human, so we should know implicitly how humans in aggregate behave.")

      Every experience in my life points to one overriding conclusion: this is a mechanical universe. I don't believe any kind of spooky magic happens regarding "human consciousness" or "free will" which would cause economics to fundamentally be different than physical sciences (though practically this may be the result, because of various (current) limitations). At the same time, I agree with Jason that we need to be very cautious about drawing analogies with other physical systems: there are many differences. Conservations laws being (and the lack of them -- in economics) being amongst the biggies.

      "Whether or not it works is completely undefined without the stories, this is not quantum mechanics this is morally right from wrong. Works means nothing without knowing what is morally good."

      You and I have very different outlooks on this. Take what John is asking for in this post: an explanation for what is happening and what has happened. This is the same explanation that Nick Rowe admits he doesn't have (but would like to have). What does a true explanation have to do with morality (other than it being true: honesty is a key (only?) part of the scientific "ethic"). Perhaps the answer is we're in our current state because there's not enough wrongful homicide. If that's truly the answer, then lets hear it: morality has nothing to do with it. It doesn't mean we *act on* this information by making a policy of murdering more (or less) people. To bring up the morality now before we even have a good sense of the mechanics is putting the card before the horse, IMO.

      "Who cares about predicting money, as long as there are inconsistent stories underneath the predictions, we are just as ignorant as before."

      Delete
    5. I agree that a morass of inconsistent stories is a bad thing (so why not escape that dead end by going off in a human-free orthogonal dimension?). I don't agree with the rest however. Take my simplified version of Jason's story to explain what John is asking for (Jason, if you're reading this, I apologize in advance... this will probably be at least 50% outright wrong, and 25% mischaracterization). Call it "Tom's hypotheses" to let Jason off the hook. He's... er,... I'm saying an economy gets into a state where it's information transfer index (kappa) is somewhere along the QTM to non-QTM spectrum of possibilities, provided the CB is reacting to market forces (i.e. it's part of a feedback loop). He has another solution to his fundamental equations for cases where the CB is truly exongenous and is ignoring the market, but I'll ignore that now. If this hypothesis is true, this is huge! Since kappa goes as log M / log NGDP, it's slowly varying, and not easy for policy to change it in short order. It means all the human-centric storytelling about why NGDP futures markets are better than NGDPLT which in turn is better than IT which is better than price level targeting, etc. and vice versa amounts to a lot yammering about very little! Perhaps just a few scraps after Tom's model explains 90% of what's going on. And it's a morality-free explanation. Does it mean we're stuck here (if we don't like it?). No: recall there are other solutions to the underlying equations, and they suggest policies which *might* extract us from this state -- for a price. Let the moralists squabble over the boring part, but lets have a firm foundation in explaining realty first. "Is the hypothesis true?" at this stage is a far more interesting (and sensical) question than "Is the hypothesis morally correct?" I claim the latter question is meaningless at this point.

      Delete
    6. My personal opinion is that discussions of morality ruin almost everything worth doing. Dr. Frankenstein would have never flipped the switch if he'd been listening to a bunch of finger-wagging incurious kill-joy Luddite moralists. In the spirit of the late Christopher Hitchens (but admittedly carrying it to extremes), Morality poisons everything! :D

      Delete
    7. I can't tell if you are being ironic...Did you finish reading Frankenstein?

      Delete
  12. Clearly, this is a short-run/long-run distinction involving expectations. If everyone believes the price level will decline by 2% per year, and it actually does, year on year, on average, so that those beliefs turn out to be well founded, I know of no economic theory, repeat NO economic theory, that can claim this to be harmful. And yes, we would then save all those shoe-leather costs and what not else.

    ReplyDelete
  13. But Steve Williamson makes a good argument why the Friedman rule may not be optimal. Maintaining a currency system is costly (replacing old currency with new, preventing counterfeiting, etc.) and currency transactions are the preferred mode of operation of criminal enterprises. Because of these costs it may be optimal to tax those who hold currency, by maintaining a small but positive rate of inflation.

    ReplyDelete
    Replies
    1. Plus a modest amount of inflation is a good tax on wealth: it's a tax on people who have large piles of cash and can't think of anything to do with it.

      Delete
  14. Wages can fall rapidly if they need to as long as regulations don't prevent them from doing so. During the depression of the early 1920s in the U.S., wages fell 19 percent in one year. Today it would be more difficult for them to change this rapidly because of minimum wage laws, pro-labor union legislation, and welfare for the unemployed.

    Japan's problems in recent decades haven't been because of its slowly changing prices, but because of its failure to quickly adjudicate bankruptcies and its massive "stimulus packages" that undermine saving and capital accumulation.

    Rapid rates of economic progress can occur even when prices are constant or falling. We saw this in the late 19th century in the U.S. Prices fell during roughly the last quarter of the 19th century and the rate of economic progress was one of the highest experienced in the U.S. (if not the highest) during any period of the same length.

    To properly understand inflation, one must understand it as an undue increase in the money supply. For deflation, one must understand it as a decrease in the money supply or spending. The focus must be on money and spending, not prices. If one has this understanding, one would see that inflation is high today, and given the propensity to spend by the U.S. government has the potential to be much higher.

    I discuss all this in great detail in my recently published book Money, Banking, and the Business Cycle.

    ReplyDelete
    Replies
    1. To properly understand inflation, one must understand it as an undue increase in the money supply.

      That is a cop out embraced by people who have been predicting "inflation" and now that it has failed to appear, they are trying to change the definition. "Inflation" is a general rise in prices, the precise definition and calculation of which is open to debate.

      Prices fell in the late 19th century because of rapidly rising productivity. The productivity gains were driven by new technology. The resulting reallocation of resources involved a lot of pain at the level of individual workers and farmers.

      Delete
    2. No cop out here, Absalon. It’s based on a proper method of defining a concept, which involves identifying the essential characteristic(s) of the concept. The definition of inflation based on rising prices doesn’t focus on the essential characteristic and, as a result, has led to confusion and false conclusions regarding what causes inflation (demand pull, cost push, wage-price spiral, etc.), a lack of understanding of the effects of inflation (mal-investment, overconsumption, the withdrawal-of-wealth effect), and a poor understanding of the causes and cures of recessions and depressions. Note also that those who have a proper understanding of inflation aren’t predicting inflation because it’s already here. I discuss all of these in great detail in my recently published book Money, Banking, and the Business Cycle.

      Regarding economic progress in the late 19th century, you are right that the progress was due to productivity gains and new technology. The freedom and competition that made these possible can lead to temporary losses for a very few but lead to gains for all in the long run. In fact, very quickly do those who might experience a loss due to a specific instance of competition start to gain from the general improvement of economic conditions. In addition, all workers and farmers today have benefitted tremendously from the freedom in the late 19th century and the remnants of freedom that still exists today.

      What you are implicitly advocating for is stagnation. You are advocating for a standard of living that existed prior to the late 19th century (or worse). You are also implicitly advocating for the government to initiate physical force and violate individual rights to prevent the economic progress. This, of course, violates freedom and restricts competition, which is what leads to the stagnation. I discuss these topics in detail in a book I wrote in 2005 titled Markets Don’t Fail!

      Delete
    3. Brian, when you say "The focus must be on money and spending, not prices." I think it is pretty clear that you are trying to redefine "inflation" to suit your argument.

      The last part of the nineteenth century did not see "temporary losses for a very few" but rather wide spread and long term hardship among, for example, small farmers. The numbers were large enough and the hardship severe enough to give rise to political movements in the United States. The "Cross of gold" speech still reverberates in American politics.

      I am not advocating stagnation. I want faster progress. In 1880 my ancestors were mostly mostly poor farmers. I'm a self employed millionaire with a couple of university degrees. I am grateful for the progress. My points are that the deflation of 1880 was the result of progress - not the other way around; and that while our grandchildren will probably benefit from current progress we should not turn a blind eye to how harsh and painful the adjustment process can be in the here and now.

      Delete
    4. Absalon, you say it is clear that I am redefining inflation to suit my argument. Please show me how it is clear that I am doing this.

      In actuality, I am redefining inflation because the definition I propose is a better definition, one that adheres to the appropriate method of defining concepts and therefore makes possible a better understanding of inflation and what is occurring in the economy. As an example of the poor understanding to which the bad definition of inflation leads, consider your claim that “deflation” (viz., falling prices) created a harsh economic climate during the late 19th century. This couldn’t be farther from the truth. As I have stated, during this period economic progress and the rise in the standard of living occurred at some of the most rapid rates experienced in U.S. history.

      People think the late 19th century was a period of stagnation because they believe deflation means falling prices and since this was a period of falling prices, it must have been a period of deflation and depression. Falling prices do occur during deflation but it’s the falling spending during a deflation that causes the problems, not the falling prices. Falling prices are merely a symptom of deflation. Deflation also leads to falling revenues and incomes and makes it harder for debtors to pay off debts. Falling prices due to increased production, which is what generally occurred during the late 19th century, do not lead to these problems because revenues and incomes do not decline and it is not harder for debtors to pay off debt. This is true, for example, because even though prices fall, the goods that businesses have available to sell increase (due to the greater ability to produce) and therefore businesses can generate the same revenue. In fact, under a gold standard, as prices fall, revenues and incomes increase due to the greater ability to produce gold. This is what generally occurred during the late 19th century. I discuss this in my book Money, Banking, and the Business Cycle.

      Note also that your desire for faster progress will not solve the problem according to your standards. You say that “the deflation of 1880 was the result of progress.” This means that if you have more progress you will have more “deflation” (viz., falling prices). According to your standards, this will only add to the problem.

      To begin to solve the problem of misinterpreting the economic conditions in the late 19th century (and in other periods), people need to abandon their invalid definitions of inflation/deflation. Note also that the definitions of inflation/deflation did not always focus on prices. In the late 19th and early 20th centuries they focused on money and spending. So the definitions I propose would merely take us back to the valid definitions of inflation/deflation.

      Delete
  15. If a central bank implemented the Friedman rule, how big would the central bank's balance sheet need to be, in equilibrium?

    If central bank liabilities are more liquid than other assets, and if people are never satiated in liquidity (they always prefer a more liquid asset to a less liquid asset, other things equal), then the central bank would need to own all the assets to achieve monetary equilibrium plus the Friedman rule.

    If the government owns the central bank, that means the government owns all the assets.

    ReplyDelete
    Replies
    1. People “always prefer a more liquid asset to a less liquid asset, other things equal..” What, so the entire population is likely to sell their houses to the central bank and live in caves surrounded by piles of $100 bills? Or would they prefer living in a house owned by the CB and rented from the CB? This is totally unrealistic.

      As to shares, under the Friedman “zero interest on base money and government debt rule” people would still hold shares because the latter produce an above zero return. Indeed, we’ve had near zero interest rates for the last three years or so, and there hasn’t been a stampede to sell shares. Quite the reverse.

      Thus I’m baffled by the claim that under the Friedman rule the CB ends up owning all assets.

      Delete
    2. Ralph,

      Nick missed a step here.

      1. If the central bank liabilities are more liquid than other assets and
      2. If people are never satiated with liquidity and
      3. The only way for the amount of liquidity to increase is for the central bank to buy assets (no helicopter drops)

      Then Nick maintains that the government owned central bank would own all assets (aka communism).

      The problem that I see is that Nick set's an absolute preference of liquidity overall all other goods, but does not allow for a relative preference between other goods. I may not sell my apples for either an apple orchard or liquidity because I want to eat today. I may not sell my apples for liquidity because I found someone to barter with (my apples for his apple orchard). I can skip the middle man (central bank / government).

      A liquid asset (like money) makes trade easier. That doesn't mean that trade is impossible without money. To get to communism, a central bank / government would have to outlaw barter.

      Delete
    3. If you look at the front of every U. S. Federal Reserve Note you will see the following inscription:

      "This Note Is Legal Tender for All Debts, Public and Private"

      Notice that that this does not say:

      "This Note is Legal Tender for The Purchase of All Goods and Services"

      Delete
  16. Nick I think you raise an important issue: how much more liquidity would there be if the government nationalized the banks, including shadow banks? I think that there would be a small increase in money, as public holdings of bank stock are replaced with liabilities of the central bank. This is a small change in money holdings because bank capital is a small fraction of bank funding. Yet there would be a significant decrease in price stability, according to the fiscal theory of the price level.

    The point I'm making is that private production of liquidity is very effective, better for price stability, and obviously more efficient than having the government own the economy. We should look for a way to extend this function of banks to a wider class of assets, by developing a suitable form of indexed debt to fund the intermediation of equity investments. Instead of people buying index funds, they should buy the index, from intermediaries.

    ReplyDelete
    Replies
    1. “How much more liquidity would there be” has nothing to do with whether banks are nationalised or not. Government and central bank can create and spend ANY AMOUNT of “liquidity” (i.e. base money) into the economy whenever they want. In fact they’ve just done that big time over the last three years in the guise of fiscal stimulus followed by quantitative easing.

      Re your claim that “…private production of liquidity is very effective, better for price stability..” are you being serious? We just been through the worst recession since the 1929 crash all caused by irresponsible bank lending or “liquidity production”.

      Commercial bank produced money is a liability of such banks, and it is “runnable” as John Cochrane puts it. I back his proposal to ban runnable liabilities from the liability side of banks’ balance sheets. That means commercial banks no longer create liquidity / money. Only the central bank does.

      Delete
    2. Ralph I thought you were willing to accept Anat Admati's 20-30% capital requirement for banks. The full-reserve system only works in complete fiscal and monetary unions like Prof. Cochrane's United States, or the one here in Canada, or your own United Kingdom. Other places need a model that works without fiscal integration. They also need better hedging instruments to avoid the balance sheet problems that we all have had lately.

      Delete
  17. My understanding of Krugman's deflation spiral is that output and unemployment spiral out of control - prices not so much. The basic argument is that the equilibrium interest rate is negative, so the zero lower bound is binding. Planned investment drops and savings are forced to match planned investment through reduced output. Wages are highly sticky and unemployment rather than diminished wages result. Deflation isn't even necessary for the spiral to get out of hand. As long as expected inflation plus the equilibrium interest rate is still negative the spiral of decreasing output and employment will continue.

    ReplyDelete
  18. did you read latest article by Paul Krugman on Scotland independence from the UK??

    laughable.....He bashes the scots but he is unable to suggest a plan to avoid the turmoil in case of a secession...

    ReplyDelete
    Replies
    1. he is unable to suggest a plan to avoid the turmoil in case of a secession

      Maybe that is because there is no plan that would work in both the short and long term.

      A Scottish vote for secession would be a kick in the teeth for the English with long term costs for the English. Why the Scots would think that the English would go out of their way to accommodate Scottish secession is a mystery. The only bargaining lever the Scots have is the express or implied threat to do even more damage to the Scottish and English economies if the English don't make concessions. I don't think that would work out for them.

      Delete
  19. have you read latest article by Paul Krugman on scotland's independence??

    laughable!! he bashes scots in any possible manner, but he is unable to elaborate on a plan to achieve a successful secession ....

    http://www.sunherald.com/2014/09/16/5803335/paul-krugman-the-scots-should.html

    ReplyDelete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.