Wednesday, 18 November 2009

Does QE cause deflation?

In the leader column of City AM today, Allister Heath argues that quantitative easing should be stopped because it will lead to inflation. Set aside for a moment the fact that this is the whole point of it. Is he right that it will actually succeed in causing inflation?

Scott Sumner (consistent with mainstream monetary theory) points out that an increase in the money supply is only inflationary if it is expected to be permanent.

And is QE a permanent increase? Not necessarily. On the contrary: QE in isolation makes the money supply smaller.

Under QE, the Bank of England has issued £200 billion of new currency and used it to buy bonds (mostly gilts) in the private market. (The Federal Reserve has done something similar, with a higher proportion of corporate and mortgage-backed bonds.)

So there's now £200 billion more sterling than there used to be - it's fair to say the money supply is bigger. What will happen next year - or in five years time, or thirty years when all the bonds have matured?

As those bonds come due, their issuers will have to pay the face value, plus interest, to the holder of the bond - the Bank of England. A £1 billion 10-year gilt maturing in 2010, which the Bank of England may have purchased for £1.35 billion, will require the government to pay £1.4 billion to the Bank next year (one more year of interest having accrued).

What will the Bank do with the money that comes back into its vaults? Nothing, unless they explicitly decide otherwise. The money is an asset of the Bank and as soon as it's paid back to them, it will disappear.

Thus, the money supply will be reduced by £1.4 billion - cancelling the £1.35 billion increase and a further net reduction of £50 million.

The Bank is buying gilts which mature in between 5 and 25 years. If the average is 15 years, and the size of the programme is £200 billion, then around £100 billion will be drained from the money supply as a result of QE. This needs to be discounted either by the predicted Bank policy rate or by expected inflation, so the exact size of the effect is hard to predict. But a reasonable estimate is £50 billion or about 30% of the UK's monetary base.

So, unless the Bank says otherwise, far from being a permanent increase, QE is actually a permanent net reduction in the money supply. Orthodox monetary theory says this should cause not inflation but deflation.

If this logic is correct, then either:
  1. QE is not causing inflation
  2. QE is causing inflation because the Bank is expected to make the money supply increase permanent
  3. QE is causing inflation because the theory of permanent increase is wrong
If case 1: it's a clever way of boosting short-term activity without (apparently) an inflationary impact. Nice idea, but completely counter to the goals of QE.

If case 2: The Bank can do whatever it wants in future years, including increasing the money supply to counter the effects of maturing bonds. If markets are assuming they will do this, it provides another explanation for why long-term interest rates are falling: the long-term path of policy rates will be lower than it would otherwise be. The other reason for falling rates is that QE provides extra demand and keeps the prices of bonds higher, but the interest rate path is probably a stronger effect.

If case 3, there's a behavioural explanation: the extra money in the economy boosts immediate spending even though people should rationally expect a reduction in the future money supply. Clarity on this may need to await a theory of expectations.

So which is it? Is QE a mistaken policy? Is orthodox theory wrong? Or are the policy and theory both right, and markets are revealing an expectation of future policy? If so, are their expectations right? Time for the central banks to clear this up.

(This argument is strengthened by the fact that the expansion of the monetary base does not seem to be going into general circulation, but that is a distinct problem).

8 comments:

Don said...

"And is QE a permanent increase? Not necessarily. On the contrary: QE in isolation makes the money supply smaller."

That's why you have a Reinforcing Stimulus. The Govt Borrows money. It spends, but does not raise taxes in a downturn. In this sense, people who say that it doesn't matter what the money is spent on are correct. But only in this sense. It is the debt that is held to be inflationary.

Why? Expectations. It's all about expectations and incentives. If you keep short term interest rates very low, there's a disincentive to invest in them. Of course, in a Deflationary environment, there wouldn't be. So, you want longer term interest rates to be higher, giving investors an incentive to invest in longer term bonds, or stocks, or riskier bonds that have higher returns.

In essence, you want higher prices or inflation to be counted on in the future. With the stimulus, you should spend money in a way that reinforces these expectations. In that sense, it matters greatly what you spend the money on.

Frankly, Bernanke's position makes no sense to me. He seems to agree with what I'm arguing, but then is pulling the rug out from under the expectations that are being created. I don't get it.

Here's the best idea for QE that I've found:

http://www.nber.org/~wbuiter/helijpe.pdf

It seems to me that Nick Rowe has given a number of different ways to accomplish QE.

Don the libertarian Democrat

PS To counter fears of excessive inflation, Milton Friedman expressly, and Keynes implicitly, put forth plans that would reduce debt as the economy grew again.

Anonymous said...

I'm a complete layman when it comes to economics but I read your blog with interest.

I was wondering if some of the banks had been allowed to go to the wall last year would the BoE still have had a QE programme or would the proverbial horse have bolted and any attempt to prevent a deflationary spiral would have been fruitless?

Whats prevents the BoE selling bonds before maturity if the market position is favourable and therefore reduce the debt earlier?

Also, what is toxic? The bonds per se or the association of the bond on the balance sheet\national debt.

Tschäff said...
This comment has been removed by the author.
Tim Young said...

I followed your link from the FT Money Supply blog.

Nice try at saying something different, but I don't think you are right. As long as the central bank is independent of government and does not permit (routine) monetary financing, debt service payments have to be funded out of taxation, and would not increase the stock of reserves. If the central bank owns government bonds, and wishes to maintain the stock of reserves, as it receives debt service payments, the central bank must actively top up the supply by buying more assets, otherwise the stock of reserves contracts. This can be used as a way of passively contracting the stock of reserves if the central bank considers this appropriate. The interest margin of the central bank's assets over the reserves paid for them would make the central bank progressively richer, but in practice is transferred to the government. Under QE, the situation is qualitatively the same, but larger.

Also, I am not sure that you are right to presume that QE is intended to increase inflation. Its aim is to increase nominal expenditure, which if prices are sticky in the short-term, means real expenditure. Rising inflation would be a sign that QE is working, because it would suggest that an output gap has closed, but inflation per se is not the aim.

Leigh Caldwell said...

Hi Tim. Thanks for the comment.

I think we're saying the same thing. As debt held by the central bank matures (whether it's government debt or private debt) the money supply reduces, unless the central bank tops it up by buying more assets.

Thus, QE inherently reduces the money supply in the long term, and is in theory deflationary.

In reality it is very likely that the central bank will top up its purchases and keep the money supply from contracting. But it would seem sensible for them to explicitly say so, saving the markets from having to guess.

You're right that increasing nominal expenditure is the preferred goal. But even if the policy doesn't succeed in directly increasing nominal spending, inflation is a second-best outcome and definitely preferable to deflation. Inflation reduces real interest rates and provides a stimulus to investment, which in turn will contribute to nominal expenditure - just taking a bit longer than the direct route.

Tim Young said...

Thanks for the discussion, Leigh, but no, we are not saying the same thing. QE is always expansionary (assuming that the reserves added are not automatically mopped up by the central bank's interest rate maintenance - but that's another discussion), but is naturally self-eliminating as the bonds purchased mature. QE is always more expansionary than not doing it.

Leigh Caldwell said...

Hi Tim

In practice I agree with you.

My point is really that the standard theory is not being borne out by events.

Conventional monetary theory combined with the permanent income hypothesis (Scott Sumner and Nick Rowe provide a more authoritative discussion of this on their blogs than I can) says that inflation is only created if a permanent increase in the monetary base takes place. Or, at least, one that is expected to be permanent. If it's expected to be withdrawn again (say, as bonds mature) then the money will not be spent.

However in practice, QE does appear to be creating inflation, or at least helping to combat deflation. Thus my hypothesis is that either:
1. people expect the increase to be permanent (that is, that the money supply will be topped up as the bonds mature)
2. for reasons of liquidity constraints, or bounded rationality, people are spending the money even if they expect it to be temporary

There is a third possibility which might be implied by your comments: a temporary increase in the monetary base creates extra nominal expenditure even without affect inflation expectations. I don't think I'd agree with that, but I'm also not certain if that's what you mean.

In any case, yes I agree that QE is - in practice - expansionary. I just think that standard monetary theory doesn't adequately explain that fact.

Tim Waring said...

I read this discussion with interest. Tim wrote "QE is always more expansionary than not doing it." I do not think there are any absolutes with QE what so ever and no one really knows its effect. I wrote recently about how I thought QE2 could be deflationary (see http://grittyeconomics.blogspot.com). Clearly the Fed wants it to be inflationary but my fear is that it creates more uncertainty. In percentage terms, here in the UK were are in a far bigger debt hole than America, but the measures announced by government have given some semblance of stability. Things are going to be awful here for years but we essentially all have to knuckle down and work like oxes to get out of this horrendous mess (if we are lucky). Watching from abroad, the American people have not been able to get to this point as the Fed is creating such instability.