Inflation, deflation and QE, redux

I've suggested previously that QE could actually be deflationary. I looked at it from several perspectives - collateral effects, the monetary transmission mechanism, distributive effects, even Peter Stella's "deadwood" inhibiting bank lending. But I have to admit that the evidence in support of my deflationary hypothesis was thin and the case not proven. All I could demonstrate was that QE is not directly inflationary and whatever stimulative effect it has is weak at best.

Until now, that is. Soc Gen have looked at QE.....and they have concluded that its effects may indeed be deflationary. Their reasoning is somewhat different from mine. Here's their argument in full (their emphasis):
QE is by design set to be inflationary, yet we were asked several times last week whether the opposite could hold true; i.e. that QE is in fact proving deflationary. 
* No credit = no recovery
In theory, a permanent increase in money supply results in a proportional increase in all money prices. Central bank asset purchases boost money supply, but this “inflationary” impact of QE is only temporary as the assets are in the future set to either be sold to private investors or redeemed to central banks, thus exerting a “deflationary” impact. For QE to be efficient, this argument would thus suggest that central banks (Fed, BoE, BoJ … and even the ECB!) should simply forgive their considerable holdings of debt (mainly government debt) thus making the increase in base money permanent. In the case of the US, cancelling the $2tn of Treasuries held by the Fed would also offer a quick fix to the debt ceiling issue.
This somewhat tongue-in-check argument merits qualification. Indeed, there is base money and then there is broader money aggregates. While QE has boosted the former, the impact on the later has been modest to date due to still lacklustre credit channels.
This ties in with our long-held view that the key to sustainable recovery lies with corporates regaining sufficient confidence to borrow, to invest and to hire (and not to swap equity for debt via share buybacks).
Moreover, tempting as a cancellation (or a restructuring to perpetual zero coupon bonds) of government bonds held by central banks may sound; we are concerned that such a policy could ultimately prove inflationary in a bad way (think Weimar Republic). Indeed, this goes to the very heart of the argument as to why QE is not printing money.
*An unintended consequence of QE's external channel 
It can reasonably be argued that QE in advanced economies generated significant capital inflows to emerging economies, boosting credit. In China, this liquidity combined with a further boost from domestic government policies found its way to fixed asset investments. This led initially to a welcome demand boost for commodities and a wide variety of capex goods. Today, however, the result is that China suffers significant excess capacity and poor capital returns, not to mention a shaky shadow banking system that China Economist Wei Yao has written extensively about. Excess capacity is deflationary and the means to deal with it is to shut it down. Indeed, we expect China for now to exert deflationary pressure on the global economy.
While in China, the impact of QE was observed mainly on the supply side, other economies such as Brazil and India saw the effect concentrated on the demand side, via consumer credit channels. Initially, local currency appreciation masked this inflationary aspect of QE. When Fed taper talk earlier this year hit the tapes, however, a new dilemma appeared for central banks in these economies as local currency depreciation added to domestic inflationary pressures at a time of slowing growth momentum. Contrary to China's deflationary impact on the global economy, we do not expect to see these inflationary forces to be exported. As households struggle to deleverage balance sheets, the end result, however, could prove deflationary.
Unproductive investment is by nature ultimately deflationary. This is a point also worth recalling when investing in paper assets fuelled by QE liquidity and not underpinned by sustainable economic growth.

In effect, Soc Gen argue that the expectation that QE will be unwound in future negates the "inflationary" effect of the temporary increase in base money. This is a sort of Ricardian equivalence. I've long thought that Ricardian equivalence is too narrowly defined: it is perverse to assume that temporary tax cuts don't work because of the expectation of future tax rises, but temporary interest rate cuts DO work despite expectation of future interest rate rises. And it is equally perverse to assume that temporary increases in base money have a stimulatory effect when temporary tax cuts apparently don't. Kudos to Soc Gen, therefore, for pointing out that temporary increases in the monetary base due to asset purchases might simply be ignored and therefore ineffective.  

Cancelling the assets would make the increase in the monetary base permanent and therefore impossible to ignore. However, I'm personally unconvinced that cancelling the assets currently held by central banks would necessarily be inflationary, since central banks have a range of tools for controlling inflation even without the presence of large amounts of readily saleable assets on their balance sheets. 

But I do think Soc Gen's argument that unproductive investment is deflationary bears consideration. For whether temporary or permanent, if the money created by QE is not productively invested, it is useless. And if corporates don't want to invest productively - perhaps because as Soc Gen says, they lack the confidence to do so - then no amount of QE will make them do so. 

Furthermore, QE money that is not productively invested in the countries where it is issued but is diverted into over-investment in other countries will ultimately prove deflationary. Soc Gen's argument is that China's overcapacity problem arises from such over-investment. They further argue that as QE is withdrawn, other emerging markets will be forced to defend their currencies at the expense of their domestic economies. If they are correct, then Western QE can be said ultimately to have deflationary effects in emerging markets. 

Whether QE is "intrinsically" deflationary remains unproven. But the argument that it can indirectly have deflationary effects seems now to be well grounded. 

Related reading:

Inflation, deflation and QE - Coppola Comment

Thanks to Tom Bowker for providing the Soc Gen research.


  1. "This ties in with our long-held view that the key to sustainable recovery lies with corporates regaining sufficient confidence to borrow, to invest and to hire."

    The credit channel is maxed out it seems. Debt to GDP at or near all time highs is a huge impediment to getting lending markets going. Economic agents need more money but the credit channel isnt effective in transmitting money into the system.

    Money policy needs to be improved to create demand on a non debt basis. In recent history private sector debt growth is continually exceeding gdp growth which undermines balance sheets and accesibility to credit. Therefore the credit channel is maxed out as a result of unproductive lending and we need to cease depending on the credit channel too much. There needs to be a mechanism of expanding money that works alongside the credit channel in case the credit channel is not working (without relying on fiscal deficits if the gov is also against stimulus).

    1. There is such a process of expanding individuals spending power alongside the credit system. It is performed by governments borrowing in order to fund tax cuts. I acknowledge that you mention anti deficit governments, but they would not be anti , if they thought that increasing peoples spending power would go on to increase the tax base. If you don't believe government spending is self financing by increasing the tax base and not a burden, then its not good idea to implement it.

    2. The gov can stimulate the economy through deficits. Unless it sends money to cronies or spends on war or something.

      But monetary policy on its own regardless of what the gov does can be much more effective and balanced.

    3. Anon are you commentator Lxdr1f7,

      in this context tax cuts are synonymous with "monetary policy on its own"

    4. Yes I am lxdr1f7

      Deficit spending may be similar in effect to expansion of money. It also depends on what the gov spends on though. But you want to have options instead of just hoping on deficits. You want different tools to use in case one fails.

      Besides directly interacting with the public for monetary policy will make the system more participatory and make people more educated in economics and confident in the institutions. Apathy need to be addressed. If we leave everything in the hands of the gov representatives then the system will probably decline.

    5. From you earlier post on the blog I you said that "interacting with the public" was crediting balances.
      I was pointing out that tax cuts are the same effect.

    6. It wont be the same but can be similar. Who gets the tax cuts? Big corporates? Borrowers? etc..

      Equally crediting the accounts of all adults will place money in account of everyone even the unemployed or students or whatever.

    7. Well in the UK we did have an across the board drop in Income tax a few years ago. And In fact bush did send out cheques in 2008.
      But any way when they aim to intervene in the economy , which is what you are proposing , targeting tax cuts or across the board is what governments do it is a policy choice.

    8. If the central bank developed the mechanism it could be monetary policy.

    9. What you are proposing exists already. The gov puts money in with the public by borrowing and then giving tax cuts and then when appropriate the central Bank removes it from circulation by selling government bonds.

    10. It does not exist. The gov conducts fiscal policy and the central bank conducts monetary policy. The central bank cant directly place funds in the accounts of the public at present but should do to make MP efficient without hoping on the gov to conducts deficits when MP fails.

      The gov shouldnt be asked to fill in the for the deficiencies of MP when MP could be made more efficient. Central bank independance is vital just like the independance of the judiciary.

  2. > Frances

    The purpose of QE is to increase inflation in the money supply not in the general price level. A clarification in the terms of reference is required there. It is not an attempt to create stagflation. They say that unproductive investment is deflationary but I thought the opposite was true - that if creating money is not accompanied by an increase in goods then that is price inflationary, but it is complicated in that if people are involved in non productive projects then their spending power is not increased. I could just about envision that in competitive markets where costs of production establish prices then a lower interest rate on the businesses borrowing may result in deflated prices.

    1. Soc Gen correctly note that the reason why QE has not been inflationary is lacklustre credit creation. Monetary base has increased considerably, but money actually in circulation (broad money) has not increased to anything like the same extent. If the so-called "money multiplier" were working normally, monetary base expansion such as we have seen in the last few years would indeed cause price inflation unless it generated sufficient new production to counter inflationary pressure. But the money multiplier isn't working. Credit isn't being created, businesses aren't investing in anything productive, wages are not rising and people are not spending. Consequently QE is not having inflationary effects - anywhere, actually. In China it is causing over-investment and overcapacity, which eventually is deflationary, and in emerging markets it is causing currency appreciation, which is also deflationary. They have a point.

    2. Dinero

      'The purpose of QE is to increase inflation in the money supply not in the general price level"

      QE is aiming at increasing demand and hence the price level. You pretty much cant do only one.

    3. No when there is production capacity to meet demand then there is no price level increase. If QE caused price inflation they would stop doing it . As I said the purpose of QE is not to cause "Stagflation".

    4. Inflation is below target so they are trying to increase it along with GDP.

      I agree that when capacity is in excess then we shouldnt see increasing prices from increased money supply. There is always a little inflation feeding into the system though even when you have excess capacity or supply. You gotta employ new staff and things like this add a little to the cost of production. Also the financial markets are quite disfunctional interest rates and commodities can react to increases in demand quite strongly so there is another source.

      But in a proper economy not the one we operate in I agree with you.

    5. I mentioned production capacity but things in fixed supply will go up.

  3. "QE is not having inflationary effects - anywhere, actually. In China it is causing over-investment and overcapacity"

    This doesn't make sense. How do you make over-investment without demanding more investment goods and thus causing inflation? It needs to be first inflationary and then deflationary if you are not assume Ricardian equivalence on inflation too (which is rather silly). Anyway how the US QE ends up propping China's investments, is there a real channel or is this just the psychological channel?

    1. There was initially an inflationary burst - see Soc Gen's comments about "welcome boost to commodities and capex goods". But we are now beyond that. Over-investment resulting in overcapacity is deflationary.

      QE is a real channel. The money that replaces the purchased assets has to be invested somewhere, and much of it has gone into emerging markets including China. Soc Gen's point is that China has used that money for investment, whereas in other EMs such as Brazil it has fuelled domestic consumption.

  4. QE might be deflationary but still worth it? I think of it as an asset swap. The CB takes an asset with a long maturity and gives one with short maturity. In the situation of a balance sheet recession everyone wants to or even has to pay down it's dept. In cash I can do so at will. In any asset with longer maturity I will either have to make my creditor accept it as payment or I will have to take more or new debt to cover the time between the date of paying my bill and the maturity of the asset. QE reduces the need to do so, therefor I don't have to take the new debt, reducing the debt in the system. This may be deflationary, but I think it helps going through the balance sheet recession quicker, making it easier.
    But if there is a positive effect on asset prices or a lowering of long term interest rates, this would be mildly inflationary as any monetary easing and by the same channels too.

  5. I think inflation is starting now but its more supplier greed because they believe they can 'get away with' increasing prices so that's exactly what they do. Further, its press comments about our wonderful economy that is fueling there management meetings to agree price increases. Its tactics (or tic tacs as mum used to say) and psychology.

  6. "central banks have a range of tools for controlling inflation even without the presence of large amounts of readily saleable assets on their balance sheets"

    Such as?

  7. Frances, good post. I agree with you that temporary open market operations or interest rate changes would have almost no effect on prices. In fact, I haven't met anyone who would say otherwise.

    I agree that canceling assets could very well make the increase in the monetary base a permanent one. I also agree that canceling assets would not necessarily be inflationary, although perhaps for different reasons than you. Many of the central banks tools for controlling inflation would be incapacitated without readily saleable assets. Open market sales wouldn't work, since the central bank would have little to sell. Boosting interest on reserves would be tough since the central bank has few remaining income-generating assets. And issuing central bank sterilization bills may not work since the central bank lacks enough income generating assets to pay a sufficiently high return.

    The reason I agree that asset cancellations need not be inflationary is because the government will probably recapitalize the transgressing central bank, thereby providing it with the resources necessary to halt the inflation. Which is just another Ricardian equivalence argument.

  8. "For QE to be efficient, this argument would thus suggest that central banks (Fed, BoE, BoJ … and even the ECB!) should simply forgive their considerable holdings of debt (mainly government debt) thus making the increase in base money permanent. "

    More accurately it's spending the money that matters not its mere existence and most certainly not its duration. Any inflationary effect in the economy comes from actual spending that precedes the debt, i.e. the government deficit that led to the debt that QE swapped for reserves. Whether this currently exists as a zero duration asset (reserves) or a longer term Treasury makes no difference, unless of course government spending is affected by it. If for example the government couldn't spend (or reduce taxes) because of real (not self-imposed) constraints, then facilitation of said spending by the central bank would prove (hyper)inflationary. Since this is not the case at all, swapping the assets around makes no difference, except in inflating asset bubbles based on misunderstandings of what QE is.


Post a Comment

Popular posts from this blog

WASPI Campaign's legal action is morally wrong


What really happened to Signature Bank NY?