Believe those who are seeking the truth. Doubt those who find it. Andre Gide


Friday, April 27, 2012

NGDP Targeting: Some Questions

Let me start by saying that the idea of a NGDP target does not sound outlandish to me. But I feel the same way about price-level and inflation targeting. The first order of business for a central bank is, in my view, is to provide a credible nominal anchor. Probably not  much disagreement about this out there.
  
Proponents of NGDP targeting, however, like Scott Sumner and David Beckworth, for example, seem to believe very strongly in the vast superiority of a NGDP target--not just as a policy that would mitigate the effects of future business cycles--but also as a policy that should be adopted right now by the Fed to cure (what they and many others perceive to be) an ongoing "aggregate demand deficiency." 

What I am curious about is not that they believe this, but how strongly they believe in it. I respect both of these writers a lot, so naturally I am led to ask myself how they came to hold such a strong belief in the matter. What is the theoretical underpinning for NGDP targeting? And what is the empirical evidence that leads them to believe that an NGDP target right now is a cure for whatever ails us right now?

One way to seek answers to these questions is to spend hours perusing their past blog posts. I'm sure they must have answered these questions somewhere. But I figure it will be more efficient for me to just state my questions and have them (or somebody else) point me in the right direction for answers.

First, let us consider the (or a) theoretical justification for NGDP targeting in general. Actually, David was kind enough to point me a nice paper on the subject: Monetary Policy, Financial Stability, and the Distribution of Risk (Evan F. Koenig). Here is the abstract:
In an economy in which debt obligations are fixed in nominal terms, but there are otherwise no nominal rigidities, a monetary policy that targets inflation inefficiently concentrates risk, tending to increase the financial distress that accompanies adverse real shocks. Nominal-income targeting spreads risk more evenly across borrowers and lenders, reproducing the equilibrium that one would observe if there were perfect capital markets. Empirically, inflation surprises have no independent influence on measures of financial strain once one controls for shocks to nominal GDP.
Alright, fine. The argument hinges on the existence of nominal debt obligations. Well, not just debt that is stated in nominal terms, but debt that is fixed in nominal terms (renegotiation is ruled out). This is, of course, a story that goes back at least to Irving Fisher (1933): The Debt-Deflation Theory of Great Depressions.

I've always liked the Fisher story. And it obviously has an element of truth to it. But admitting this is different than asserting that the mechanism is quantitatively important, especially for generating decade-long recessionary episodes.

First of all, as I alluded to above, people can and do renegotiate the terms of nominal debt obligations if things get too far out of whack. True, renegotiation (including outright default) is costly and imperfect, but it happens nevertheless. And to the extent it does, nominal debt is not as "fixed" as some make it out to be. It would be good to know how much renegotiation does or does not happen out there.

Second, even if renegotiation is quantitatively unimportant, we should consider the dynamics of debt creation and retirement. At any point in time there is an outstanding stock of nominal debt, with terms negotiated in the past on the basis of future price level paths (among other things, of course). We should also keep in mind that new debt agreements are being formed, and old agreements are being retired and modified (refinanced) continuously throughout time. How big are these flows relative to the outstanding stock of debt?

I think the answer to the previous question is important for understanding how long the real effects of a "negative price-level shock" can be expected to last. If "debt turnover" is high, then such a shock cannot reasonably be expected to generate a decade of subnormal economic performance.

We are presently more than 3 years out from the sharp decline in the price-level that occurred in the fall of 2008. How much new nominal debt has been issued since then--debt that would have presumably been negotiated with expectations of a new price-level path? Does anybody know?  In particular, if one is advocating a return to the old price-level path right now, what does this mean for the creditors who have extended loans over the past 3 years? Should we care? Why or why not?

I have not even touched upon the practical feasibility of NGDP targeting--I'll save this for another day. But for now, I'd like to know the answers to my questions above. Who knows, I too may become one of the faithful! 

A good weekend to all. 

43 comments:

  1. its not the price level per se, its home prices that affect residential mortgage debt, prices are are still declining nationwide. calculatedrisk has updated numbers but a significant fraction are underwater (corelogic says 22% i recall). calculated risks distressing gap chart is useful. how hard is it to renegotiate? you cant. try. once you default itll be impossible to get another for years, you have to rent. foreclosure is hard: theres about 1.5 yrs of inventory still in the pipeline. read about the foreclosure mess. again calculatedrisk is a decent source im sure if you email him hell be helpful. the homeownership rate is 66% ish, this affects a giant cohort of people.

    ReplyDelete
  2. oh yeah, read up on how hard it is to push through principal forgiveness via fannie and freddie: we the taxpayers cannot even agree to give principal foregiveness to other taxpayers who are 150% ltv (mtg/home value) with no hope of ever selling for what it would take to recover what they paid.

    the size of the mortgage market is larger than publicly held treasuries. http://www.federalreserve.gov/econresdata/releases/mortoutstand/current.htm

    the flows are small compared to the stock.

    ReplyDelete
    Replies
    1. dwd: The link you sent me provides stock information (thank you), but no data on gross flows (I'm not even sure it exists). How do you know flows are small?

      Delete
  3. older post but should get you started.

    http://www.calculatedriskblog.com/2011/11/lps-foreclosure-timelines-increase.html?m=1

    home prices are still declining, we are probably near the bottom, but have a few years before the current delinquent mortgages work their way through.

    ReplyDelete
    Replies
    1. OK, this is a bit better -- let me look at it (thanks).

      Delete
    2. industry experience (which i know is a hollow answer). People at the NY Fed have access to LoanPerformance, Corelogic, and other industry databases (when i want to know i have access to a Bloomberg terminal as well).

      there are also reams of Fed staff papers on the housing bust and mobility (see below).

      I think the answer is more than a simple stock/flow analysis: even if i were to refinance my house today because rates are low, there are significant closing costs (and I could not refinance if i had negative equity). For consumer debt, there are significant institutional barriers (banks dont like "principal foregiveness because they are afraid of incentivising borrower behavior to miss payments to get this), wealth barriers (you might pay tax on principal foregiveness!), social barriers, plus "renegotiation" os principal is synonymous with default and generally you lose the house and will not get a mortgage for 7 years.

      Corporate debt is much much easier to renegotiate - corporations are in my experience ruthless exercisers of there ability to renegotiate their loans.

      i wish i could devote more time, this is a complex subject (there are reams of research papers out there).

      But, i would honestly sum it up like this: 75% of deeply underwater (115-150% ltv) stay current. that means that the cost renegotiating must be on the order of 30% of the value of the house in those cases (closing costs and taxes along might eat up half that). so thats a big wedge.

      http://www.newyorkfed.org/research/staff_reports/sr526.pdf

      http://www.fhfa.gov/webfiles/23056/PrincipalForgivenessltr12312.pdf

      Delete
    3. another interesting graph: delinquencies by vintage (origination year).

      also: underlying this data, one can could at cumulative defaults.

      http://www.calculatedriskblog.com/2012/01/lps-2010-2011-mortgage-originations.html

      Delete
    4. and from the depths heres a fed paper that attempts to estimate the "cost" of default, which consistent with much other stuff i've seen (again, not sure that "turnover" captures the full depth of the stickiness of nominal mortgage debt).

      Also: i am reminded of an important point. State laws differ as to "recourse." Suppose I walk away from an underwater house. In some states, the bank is limited to the value of the collateral (market price of house). In other states, the bank can sue you for the deficiency (so you have to pay back the loan no matter what). In so-called recourse states, "renegotiation" is bascically precluded by law.

      Indeed, mortgages are very complicated.



      http://www.federalreserve.gov/pubs/feds/2010/201035/201035pap.pdf

      Delete
    5. David,
      ... one more. I finally found a dealer presentation online !!the internet is a wonderful thing.!!

      This is extremely detailed. I bet a campbells soup condended version can be had through NY Fed.

      {this is the kind of industry research i noted above}

      Notes:
      figure 4, p 93: WAC distribution. Note everything is to the right of the yellow arrow (the mortgages were issued when mortgage rates were much higher, in the past). Very hard to refi even when rates are low due to credit, negative equity, etc. Somewhere in here it might show vintage (origination year) as well, but the 5.5%+ seasoned coupons were probably issued mainly in 2008 and prior.

      Most of the tables like fig 26, p 45; & 27, p 48; 32, p 51 show vintages going back to 2004 because there is a substantial fraction of high coupon seasoned product out there that cannot be refi'd to lower rates.

      Figure 2&3 pages 92,93 are interesting (percent of refi-eligible mortgages)


      Again, lots here.

      as i said above, most estimates I've seen suggest the "cost of default" (i.e. renegotiating principal, not merely refi-ong) are in the neighborhood of 30-60% of the home value. Thats why so many underwater mortgages are still current!

      http://www.scribd.com/doc/75101537/4/Agency-MBS-GSEs-Money-Managers-and-REITs

      Delete
  4. Are they really saying that the problem is 'aggregate demand deficiency'? I thought market monetarists didn't believe such stuff.

    I don't entirely understand how a successful NGDP strategy is suppose to work, I mean, Bank of Japan was barely able to keep a positive inflation rate even less NGDP under control.

    And how would the public's inflation expectations be affected? It's much easier and faster to measure CPI than NGDP (which would probably be confused with real GDP as well). Would that be a problem?

    ReplyDelete
    Replies
    1. Market Monetarists believe there is an NGDP deficiency and NGDP is synonymous with aggregate demand.

      Delete
  5. David — Debt really is a very rigid institution. Sure, payment schedules are renegotiated and extended fairly frequently. But regardless of the term of a loan, regardless of whether one has explicitly contracted a non-callable 10-year loan or one is borrowing overnight and rolling the proceeds for 10 years, the principal of a loan is rarely forgiven unless there is a credible threat of default. "Turnover" results in renegotiation of interest to current market rates and credit spreads, but yesterday's short-period loan must generally be repaid in full with new money borrowed today. A reduction of principal amounts to a partial default, to which creditors assent only when they have very little bargaining power. Creditors find a great deal of bargaining power in the power to enforce payment in full at pain of bankruptcy, except when the debtor has little to lose.

    Of course, sometimes debtors do have little to lose. A game of chicken ensues, which results in a mixture of "voluntary" restructurings and outright defaults. These events do reduce aggregate leverage, but when "voluntary" fails, conflicts among claimants can provoke large external costs. Firms that might be viable under an all-equity capital structure are liquidated and employees are laid off; even firms that are reorganized lose customers and economic value; fixed capital pledged as collateral abruptly reverts from specialized use to bank-owned scrap; humans are thrown out of homes which then go to seed. Default and bankruptcy are microeconomic institutions designed for exceptional cases in a decent economy. They are ill-suited to mass use for macroeconomic management of depressions.

    There are some more flexible cases. Sometimes creditors do renegotiate the principal on outstanding loans, even when borrowers are healthy. When loans are non-recourse and undercollateralized, creditors are sometimes persuaded to negotiate the principal down towards collateral value. When loan contracts are entered as part of an ongoing commercial relationship, creditors may write down some debt in hopes of recouping the loss from future business. For example, suppliers sometimes forgive unpaid bills of customers to whom they have extended credit. Landlords often acquiesce to lease renegotiations when real-estate prices fall, in order to maintain good relations with and win future business from the tenants who occupy and control their property.

    But these cases really are exceptional. It is hard to win concessions from an arms-length creditor when a debt has become painfully burdensome but is not life-threatening. Nominal debt is often refinanced, but rarely diminished, except by actual payments out of income, or when there is serious threat of default. A "debt hangover" can last a long, long time.

    ReplyDelete
    Replies
    1. Steve, thanks for this. A lot of interesting material here. I am particularly intrigued with what you say in your second paragraph. Firms that might be viable under an equity capital structure, are not under a debt structure. And that while debt may have some role to play in aligning incentives, it is not a structure that is well suited to (say) rare macroeconomic events. I will have to think about that.

      Yes, I agree with you that a "debt hangover" can--in principle--last a long, long time. But for this to be true, the majority of debt contracts negotiated (say) prior to 2008 must still be alive and, more importantly, must be expected to stay alive for several more years.

      You have listed a number of reasons outlining why the institution of debt is rigid. I suppose I'm asking whether anyone can quantify this rigidity?

      There is a flow of new debt being issued, and old debt being retired (let's even forget about renegotiations). How big are these gross flows? The presumption seems to be that these flows are very small relative to the stock. This may very well be the case. I would have expected anyone making a case for NGDP targeting would have a ready answer to this quantitative question, since it seems vital (to me, at least) in supporting the strong view that adopting an NGDP target "right now" is going to help in some manner.

      I'm still in the process of organizing my thoughts on this matter. Still have lot's of reading to do over at your blog, as well as Scott's, David's, Nick's, etc.

      Delete
    2. I am confused about what kind of firm is viable under an all-equity structure that wouldn't be viable under an all-debt structure. Under pretty much any conditions I can imagine, equity is costlier than debt. The benefit of debt in a business is that it can force bankruptcy and liquidation or reorganization when all-equity would not, and thus all-equity allows for a longer period of wealth destruction than would otherwise be the case. I know this is an aside to the NGDP post, but it struck me as very odd.

      Delete
    3. Not all-debt, G**dammit. Partial debt-partial equity.

      Delete
    4. Under pretty much any conditions I can imagine, equity is costlier than debt.

      sure but i think you are oversimplifying it. theres all kinds non-debt financial and operational leverage (pension costs, employee costs, leases or long term assets, i could think of a lot), and off-balance sheet debt.

      also, i think its the opposite: sure, bankruptcy can force liquidation, but thats not really the goal. all-equity (really, firms with a high proportion of variable costs) can adapt their output/cost structure more quickly in a downturn. In a bankruptcy, you cancel your out-of-the money fixed contracts, and in many cases return to profitability with a different cost structure. If you maintain a high enough porportion of variable costs, there is no need to go through bankruptcy to do that.

      In some cases, debt itself forces companies to enter into fixed price contracts:

      say i want to buy gas generation plant. Bank A will lend me the money to the extent that the revenues/costs are fixed. So I go out and execute a fixed-price contract for the sale of power and a fixed-price contract for the purchase of gas. Bank A will now look through (to some extent) to the credit rating of those with whom I have executed contracts.

      Delete
    5. Prof J — Under the simplest theory, Miller and Modigliani capital structure invariance, the viability of a firm should be unrelated to its capital structure.

      In practice, most people think capital structure matters, but not because the lower cost of debt makes otherwise nonviable firms viable. The M&M intuition stands that, holding the asset side constant, the reduction in overt financing costs achieved by levering up a firm is matched by an offsetting increase in the cost of equity, due to the increased riskiness of the shareholders' claims.

      We have to think harder to understand how capital structure matters. Leverage can create a tax asset, the "tax shield" first described by M&M. There is the story that you point to: a leveraged capital structure might discipline managers and reduce agency costs, enhancing the overall value of the firm. But then there is the story that I point to, a leveraged firm may create distress costs, most especially when the going-concern value of the firm is positive but below the value of debt claims. A cumbersome legal process called "bankruptcy" often ensues. In theory, if the going concern value is greater than liquidation value, debt claims should simply convert to equity and firm value should be unimpaired. In practice, even in reorganization, bankruptcy often impairs business value through a variety of channels. Plus, there is a real risk that nervous senior creditors will force viable firms into liquidation, destroying going-concern value, which is always a risky, intangible, and contestable asset.

      Delete
  6. Williamson suggested here: http://newmonetarism.blogspot.com/2011/10/nominal-gdp-targeting.html that NGDP targeting isn't qualitatively different from inflation targeting since it fits in the same class of Taylor rules.

    Then here's George Selgin on NGDP:
    http://www.freebanking.org/2011/12/01/wide-off-the-mark-or-nonsense-about-ngdp-targeting/

    Mark Carney:
    http://www.bloomberg.com/news/2012-02-24/carney-speech-at-u-s-monetary-policy-forum-in-new-york-text-.html

    Mark Carney seems to have a lengthy critique of the whole concept of NGDP as an optimal source of information for central banks.

    ReplyDelete
    Replies
    1. Prof J: Thanks for a nice set of links. I think all three constitute essential reading.

      Delete
  7. payment schedules are renegotiated and extended fairly frequently

    not sure what you mean here. Companies do renegotiate their debt all the time. Not so with mortgages or auto loans.

    mortgages can be refinanced, but the new loan has to meet credit standards. So if the borrower falls behind or the mortgage is underwater, to bad so sad. We have programs like HAMP which have been completely inadequate. "renegotiation" is further complicated by the fact that the servicer rarely owns the loan, and there may be a second loan. So i would say that for borrowers with the ability to repay and positive equity, refinancing is viable, but when the home prices drops to the point where the loan is too high LTV, that route is not viable.

    Its the "double whammy" of negative equity and inability to repaya causes deliquencies/foreclosures. That is very important. I cannot find the statistics offhand, but 70% of fannie/freddie 110%+ ltv loans are current. when people can repay, they largely try to. when their income drops, then they default. generally at that point you lose the house through short sale or default (there is some ability to renegotiate, but usually not involving principal reduction, and the delinquency rate on renegotiated loans is very high so it postpones the inevitible).

    thats why in the midst of a housing price decline, allowing ngdp to drop is pretty caustic.

    heres a another good white paper:
    http://www.federalreserve.gov/publications/other-reports/files/housing-white-paper-20120104.pdf

    ReplyDelete
  8. David,
    I think the question is not just, "does targeting help alleviate a severe debt hangover?" Its also, "what conditions create severe debt hangovers, and how does targeting contribute to those conditions?"

    The prospect of default is bad without targeting. Knowing this, actors hedge against the probability of default. In "normal" conditions, these hedges -- liquidity reserves, excess capital, diversification -- prevent long hangovers. In other words, the system has built-in stabilizers that create robustness.

    By producing nominal certainty, NGDP targeting also reduces hedging -- stability -- during the upturns/booms. The result is a fragile system, one vulnerable to small shocks like the plateauing of house prices in 2006. Of course, if NGDP targeting is successful, then these shocks will not have an impact. The question is, "is there a level of reduced hedging that makes it impossible for a central bank to consistently hit a nominal target, begetting a long hangover as actors seek to rebuild hedges to higher levels?"

    Take away default fear, and one allows more catastrophic failure to occur. This is like preventing moderate fires in a forest, allowing the underbrush to grow, and losing the forest to a conflagration.

    ReplyDelete
    Replies
    1. This is a great point, Diego. I was planning to step back and take a broad view of the policy, but for now, I was mainly interested in how implementing an NGDP target right now was supposed to stimulate the economy.

      Delete
  9. Experienced bankers solved this targeting problem 200 years ago: Banks should only issue paper money in exchange for good bills, at not more than 60 days date. The 60-day maturity limit prevented maturity mis-matching, especially when combined with suspension clauses that allowed them to suspend convertibility of their paper money for up to 60 days. The 'good bills' provision assured that a bank that issued $100 received an asset worth at least $100 in exchange so that the money was adequately backed and would hold its value. Some (not all) bankers insisted that the bills be based on productive activity. This assured an "elastic currency'---one that grew and shrunk with the needs of business.

    Then around 1810, during the Bullionist debates, politicians joined forces with academic scribblers to force bankers to issue money according to Bullionist (i.e., quantity theory) principles. The resulting disaster is still unfolding, as each new crisis produces an alphabet soup of new proposals from the scribblers.

    ReplyDelete
  10. David: My response here: http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/04/three-arguments-for-ngdp-targeting.html

    CL: "Are they really saying that the problem is 'aggregate demand deficiency'? I thought market monetarists didn't believe such stuff."

    ?? We most certainly do!

    ReplyDelete
    Replies
    1. Nick, yes, I left a reply to your post. I still don't think you answered my question. You seem to be saying that you don't believe that the policy is not to be valued for curing what people perceive to be a huge "debt overhang" problem (what you call risk-sharing). But this is mostly what I hear people say when then urge the Fed to adopt the rule right now. Maybe I'm on wrong on this though. I'll sort it out eventually.

      Delete
  11. @David,
    "First of all, as I alluded to above, people can and do renegotiate the terms of nominal debt obligations if things get too far out of whack. True, renegotiation (including outright default) is costly and imperfect, but it happens nevertheless. And to the extent it does, nominal debt is not as "fixed" as some make it out to be. It would be good to know how much renegotiation does or does not happen out there.

    Second, even if renegotiation is quantitatively unimportant, we should consider the dynamics of debt creation and retirement. At any point in time there is an outstanding stock of nominal debt, with terms negotiated in the past on the basis of future price level paths (among other things, of course). We should also keep in mind that new debt agreements are being formed, and old agreements are being retired continuously throughout time. How big are these flows relative to the outstanding stock of debt?"

    This is a large question that demands dividing it into more managable portions. Let's consider just the household sector for now.

    At the end of 2011, there was some $13.2 trillion in household debt outstanding. Of that nearly three quarters, or about $9.8 trillion, consisted of home mortgages (Table D3):

    http://www.federalreserve.gov/releases/z1/Current/z1r-2.pdf

    Let's consider home mortgage debt.

    NGDP and RGDP both reached a trough in 2009Q2, so to answer the question we would need to find out how much home mortgage debt had been renegotiated, defaulted, retired and nwly created since then. The Mortgage Banker's Association (MBA) keeps track of the number of originations subdivided into new originations and refinances. Assuming there is no double counting (multiple refinances on the same mortgage) if one takes the sum value of those originations and subtract it from the outstanding mortgage debt we should have a rough estimate of the value of all the mortgage debt that was negotiated prior to 2009Q3. (We needn't worry about the principle balance since that will barely change in a new loan after 10 quarters.) Mortgage originations totaled roughly $3.5 trillion in 2009Q3 through 2011Q4, of which about 60% consisted of refinances (Figure 8):

    http://www.mortgagebankers.org/files/Research/IndustryArticles/FratantoniKan_0112.pdf

    So after 10 quarters (not counting 2012Q1) about 36% of outstanding home mortgage debt has been negotiated since the attainment of the new NGDP trajectory.

    ReplyDelete
    Replies
    1. Mark, this is great, thanks! These are the type of numbers I was looking for. Need to digest them now...

      Delete
  12. What about the remaining $3.4 trillion in household debt?

    Well, according to the NYFRB, it falls into the following categories (rounded, Page 5):
    1) Home Equity 18%
    2) Auto Loan 21%
    3) Credit Card 21%
    4) Student Loan 25%
    5) Other 11%

    http://www.newyorkfed.org/research/national_economy/householdcredit/DistrictReport_Q32011.pdf

    Assuming HE loans are newly created, retired, defaulted and refinanced at the same rate as mortgages then about 36% of HE debt has been negotiated in the ten quarters through 2011Q4.

    Theoretically there is an auto loan refinance industry. In practice it rarely happens because banks look at resale value. Thus for auto loans we only need to consider the rate at which they are paid off and newly created. Assuming a typical auto loan is for five years and assuming that the auto loan origination rate is proportional to the value of motor vehicles and parts expenditures in BEA Table 2.3.5 then approximately 43% of outstanding auto loan debt has been taken out since 2009Q3. For auto loans, given their short life, we have to consider principal balances. Taking that into account one comes up with a an estimate of 70% for the amount of outstanding auto loan debt taken out since 2009Q3.

    Credit card debt is revolving. The question really is how much of the principal balance has been incurred since 2009Q3? Who really knows? I don't even think banks keep track of such numbers.

    Student loans are generally speaking not renegotiable. So then the question becomes how much has been retired and newly created in 10 quarters. Given that student loan debt is a lot like a mortgage these days, although with a somethat shorter term, I would guess roughly 20% of the outstanding student loan debt has been originated since 2009Q3.

    Which leaves Other, which is probably unsecured debt similar to credit card debt. For sake of argument let's suppose that of the outstanding credit card debt and Other about half has been originated since 2009Q3.

    Now, a simple weighted average results in an estimate of 44% for the proportion of non-mortgage debt that has been originated/refinanced in the ten quarters through 2011Q4, or approximately $1.5 trillion.

    Thus a total of perhaps $5 trillion in debt has been originated/refinanced since the new NGDP trend has been established. Which means that about $8.2 trillion or approximately 64% was negotiated before the new trend was established.

    Assuming that the rate of origination/refinancing is linear (dubious) then it will take a least another five years before all household debt conforms to current NGDP growth expectations.

    One major caveat remains however. Just because one chooses to refinance a debt does not mean that the new terms are compatible with one's nominal income expectations going foreward. In many cases people are simply exercising an opportunity to modestly reduce their mortgage payments and/or hold on to their homes. And many of the people who might benefit the most from refinancing cannot because of underwater mortgages, reduced earnings, degraded credit ratings and tightened lending standards.

    Thus I suspect this is a very optimistic estimate of the degree of adaptation to new circumstances.

    ReplyDelete
  13. @CL,
    "Are they really saying that the problem is 'aggregate demand deficiency'? I thought market monetarists didn't believe such stuff."

    Fundamentally NGDP targeting *is* AD targeting. Please come to an MM blog and find out what you have been missing.

    and:
    "I don't entirely understand how a successful NGDP strategy is suppose to work, I mean, Bank of Japan was barely able to keep a positive inflation rate even less NGDP under control."

    Scott Sumner addressed this pretty well in 2009:

    http://www.themoneyillusion.com/?p=2583

    If the goal of ryōteki kin'yū kanwa was literal price stability then the Japanese seemed to have succeeded incredibly well.

    Moreover the real effects weren't too shabby either. During 1993-2002 RGDP growth averaged 0.9%, unemployment rose almost consistently every year from 2.2% in 1992 to 5.4% in 2002. The Japanese announced QE on March 19, 2001 and maintained it through March 9, 2006. RGDP growth averaged 2.1% during 2003-2007. Unemployment fell every year until it reached 3.9% in 2007. That sounds like it worked to me.

    ReplyDelete
  14. The most important thing is policy consistency. If the Fed followed a consistent inflation policy, nobody would be paying any attention to Scott Sumner. The inflation vs GDP argument would be confined to academic journals.

    Why did the Fed react to the recession by lowering its inflation target? Basically it's because we have a tug of war between "inflation hawks" and "inflation doves", and cyclical fluctuations provide an opportunity for one side to gain ground over the other. It's a very costly war.

    ReplyDelete
    Replies
    1. Max, it is true that until recently, the Fed had no explicit inflation target. But it was widely known that there was an implicit 2% target (recently made official). I have no recollection of the Fed lowering its inflation target in reaction to the recession.

      Delete
    2. http://www.clevelandfed.org/research/data/inflation_expectations/2012/April/image1.gif

      Forget what you know about Fed policy for a minute and just look at this graph. What do you see?

      What I see is a slow downward drift, which paused in the 2000s (well above 2%), and resumed in 2008.

      "The Federal Reserve Bank of Cleveland reports that its latest estimate of 10-year expected inflation is 1.47 percent."

      Delete
    3. Well, sure, I've made a similar point before; see here:
      http://andolfatto.blogspot.com/2011/10/inflation-expectations-downward-march.html

      The striking thing about this graph is the dispersion in expected inflation rates following the crisis.

      In any case, I stand by what I said before: As far as I know, the Fed did not lower its inflation target as you are asserting. Inflation may have drifted lower (below 2%) for a while, but this can reasonably be expected to be temporary. The foreign demand for USD and UST also impacts inflation; it is not just a matter of Fed policy.

      Delete
    4. The chart starts in 1982, so the downward drift has been going on for 30 years. How can this be unless the Fed has repeatedly lowered its (implicit) inflation target?

      Regardless of your interpretation - the fact is that the Fed has never followed a consistent policy for any significant length of time. The same is of course true of other central banks.

      The question of what the optimal target is, is secondary to the question of how do you get a consistent long term policy. Maybe it's impossible.

      Delete
  15. An incisive post. A similar point about NGDP targeting and debt occured to me recently ( http://www.themoneyillusion.com/?p=13841#comment-148678 ) - ie that NGDP targeting makes debt more like equity. Economic theory is not my strong point, but the apparent prejudice that some commentators have against debt (including money) seems to defy the rationality that they normally apply to economic analysis. For example, in point 1 of his reply to you, Nick Rowe simply asserts that debt in the presence of price targeting "doesn't seem to be an efficient or fair way to allocate aggregate risk". Why is it inefficient or unfair if consenting adults agree to a contract that is allocates risk in that way? It is not as if the debtors have agreed to offer a pound of flesh as an alternative to repayment! Presumably, for wholesale investors, there are ways of adjusting the price level risk allocation via derivatives (eg inflation swaps) anyway. What is the "perfect capital market" mentioned in the Koenig quote above?

    As I frequently remark on The Money Illusion and Macro Market Musings blogs, it seems to me that a major reason for the gathering support for NGDP targeting is exactly that if introduced now, it would effectively rewrite an array of debt contracts which are troublesome for the debtors and the authorities that would have to enforce them simply because that would be painful, rather than because inflation has proved to be unexpectedly low. I can see some merit in NGDP targeting, but if it is to be introduced, it should be introduced with a low target (eg 4%) that ensures that the motive is not the windfall that debtors would receive from the policy change. And the proponents of NGDP targeting would presumably have to accept that in future, debt interest rates would probably be higher to adjust for the shift in risk.

    That said, I must admit that I am so hawkish that I trace the relative decline of the British economy back to the closure of the Marshalsea prison in 1842!

    ReplyDelete
  16. @RebelEconomist

    "Why is it inefficient or unfair if consenting adults agree to a contract that is allocates risk in that way?
    ...
    rewrite an array of debt contracts which are troublesome for the debtors and the authorities that would have to enforce them simply because that would be painful"



    not exactly. unexpected inflation or disinflation will redistribute wealth, and as a number of people have observed, the Fed has opportunistically disinflated, lowering price level which inherently benefits creditors. http://economistsview.typepad.com/timduy/2012/04/bernankes-shift.html

    Second, an ngdp target actually prevents more rewriting of debt contracts: the persistent unemployment-default-home price decline is a vicious circle. When people can pay, they try to. But it's the dual hit of inability to pay with negative equity that essentially forces default, while the home stays in foreclosure for 2 years (700+ days). Some investor picks up the property at the new lower price. Its not a "trying to enforce them" thing. you can't "enforce" a debt contract when income has declined. period. People who say ngdp targeting is some scheme to prevent debts from being rewritten are... well its exactly the opposite of the facts.

    The goal of fed policy IMO should be to replicate as much as possible the flexible price equilibrium and minimize redistribution (no windfall to anybody). You can't do that with inflation targeting when house price debt contracts are so rigid. Thats both theory and reality

    incidentally, i do know of some hedge funds who have pushed the idea of restructuring the underwater portion of mortgages into an equity-like structure so that investors share in some upside if prices rebound. homeowners would probably willingly provide enter into such contracts. However, between the regulatory inertia and opposition (have to get it approved in 50 states, with different mortgage lein laws - and state and federal agencies are suspicious and skeptical of all newfangled products these days), banking industry inertia and opposition, and a whole host of other issues, the idea has gained zero traction. zip.

    ReplyDelete
  17. @dwb, I think the case that there has been any significant inflation or disinflation in the US is weak. Years ago, Greenspan used to say that the Fed aimed to hold inflation to a level where it was not a significant consideration in business and household decisions, which he informally put at about 2%. And that is about where we are now. Sure, there are some who select core measures or PCE inflation etc to make their point, but there are others (me among them) who would note that the Fed did not revise downward its inflation target to account for the Boskin commission changes in the 1990s that reduced the measured rate of CPI inflation by at least 0.75%. In the case of US mortgage debt, I dare say that many mortgage debtors have had opportunities to refinance at much lower rates than they expected when they started. Monetary policy has not been unkind to debtors.

    As David says, default is a form of renegotiation, and individual defaults are a much better targeted renegotiation than generalised inflation, that should promote more careful negotiation in future by both creditor and debtor. So get on with the foreclosures I say. I am not so familiar with the US, but in the UK, there are millions of young adults desperate to buy a house at an affordable price. Don't change the rules to deny them their chance. I am all in favour of helping mortgage defaulters, but help them to move and re-establish, not to hold on to the house that they wrongly gambled they could afford.

    ReplyDelete
  18. My first sentence should say "...unexpectedly low inflation or disinflation...." of course.

    ReplyDelete
  19. @RebelEconomist
    I am not so familiar with the US, but in the UK, there are millions of young adults desperate to buy a house at an affordable price.

    clearly.

    I am all in favour of helping mortgage defaulters, but help them to move and re-establish, not to hold on to the house that they wrongly gambled they could afford

    no one is talking about inflation. no one is talking about a keeping someone in a house they cannot afford or helping gamblers. Thats just false straw man argument and ignores the facts. 70% of underwater homeowners are current. Only when they get hit with unemployment or an income shock do they default. with 8% UE, stabilizing AD and closing the output gap significantly lowers the probability that they will get hit with an unexpected income shock that prevents them from staying current. Conversely, if you allow nominal income to decline, the unemployment-default-negative equity downward spiral continues putting even more people in a negative equity situation.

    ReplyDelete
  20. @ David: "the majority of debt contracts negotiated (say) prior to 2008 must still be alive and, more importantly, must be expected to stay alive for several more years."

    Mortgages comprise the majority of consumer debt, and are, as far as I know, mostly 30-years in term.

    Anon212

    ReplyDelete
    Replies
    1. Anon, even if this is true, there is a flow of 30 year mortgages that are being retired, right? And a flow of new 30-year mortgages are being issued, right? I am asking how big are these flows. Things are complicated as well when, for example, someone refinances their 30 year to a 15 year mortgage (say, to take advantage of lower rates). I think that the data provided by Mark, above, may help answer some of these questions. Thanks.

      Delete
  21. "I am asking how big are these flows"

    Should be answered by the Flow of Funds data from the Fed, which includes household real estate assets and liabilities.

    I'm sure there are better (i.e. more timely and accurate) sources of data - perhaps someone more familiar with US housing can recommend some.

    Anon212

    ReplyDelete