The Bank of Canada says: "Quantitative easing if necessary, but not necessarily quantitative easing"

The Bank of Canada released its eagerly-anticipated Monetary Policy Report (pdf) today, in which it laid out its contingency plans for quantitative easing. But it seems pretty clear that - notwithstanding its jaw-dropping estimate of annualised 2008Q1 GDP of growth of -7.3% - it doesn't intend to be implementing those plans anytime soon.

Here's why:

It's instructive to dig out our intermediate macro textbooks at this point. Here is what we'd expect from an economy that is hit by negative demand shocks and by negative supply shocks:

Negative demand and supply shocks both have the effect of reducing output and income. But while negative demand shocks are deflationary, negative supply shocks are inflationary. If Canada's recession were due to a reduction in aggregate demand, then deflation would be a real and present danger, and the Bank of Canada would do well to prepare us all for a wave of quantitative easing.

But it's not clear that our recession is best characterised as a negative demand shock. As I argued here, it may make more sense to think of the current recession as a negative supply shock generated by a deterioration in our terms of trade. 

In the MPR, the Bank continues this line of thought:

Significant structural changes are under way in the Canadian economy. These changes are most apparent in the automotive and forestry sectors but are occurring in other sectors as well. Owing to their scale and abruptness, these changes will have an important effect on the current level of potential output and on its short-term growth rate. The closing of plants and other establishments will render a significant proportion of the capital stock obsolete. There will also be adjustment costs as labour is gradually redeployed to other activities. It is estimated that these developments will reduce the level of average labour productivity in the first half of 2009 and, hence, the level of potential output, 1.0 percentage point below what they would otherwise have been.

In addition to the terms of trade shock, the Canadian economy is undergoing the equivalent of that staple of macro exam questions: the sudden reduction of its productive capacity. The capital stock in those sectors may not have disappeared in an earthquake, but neither is there any realistic possibility that it will ever be put to use again.

This story is also carried over to the revision in the Bank's estimates for potential output:

[P]otential output growth is expected to slow to 1.2 per cent in 2009 and then pick up gradually to 1.5 per cent in 2010 and to 1.9 per cent in 2011. Previously, potential output had been expected to increase by 2.4 per cent in 2009 and by 2.5 per cent in both 2010 and 2011.

The obverse side of this revision is that estimates for the output gap have also been trimmed by a substantial margin. This reduction in its measure for deflationary pressures explains why the Bank doesn't feel obliged to send the launch codes for its quantitative easing scenario just yet.

There's lots more in the report, including the strategy for credit easing. I, or maybe Nick, or maybe both of us will get to that soon.

Comments

[Scott Sumner]

Stephen, I would make several observations:

1. If the problem is a supply shock, monetary stimulus is not appropriate, but then neither is fiscal stimulus (unless it is cuts in MTRs.) Does Canada have fiscal stimulus? (Sorry, I am the typical ignorant American.)

2. A reduction in AS is probably part of the problem.

3. I would be absolutely stunned if AD isn't also declining sharply. What is the rise in the GDP deflator? I'm guessing it is less that 12.3%, in which case NGDP is rising below trend (well below trend for the U.S., but I'd guess Canada's trend rate is similar.)

4. Look at nominal wages. If aggregate nominal wage growth is expected to slow this year and next, then you have an AD problem. If not, then monetary policy may be fine.

So I agree with your theoretical analysis, but (without knowing the data) my hunch is that Canada would be a bit better off with more AD.

[Stephen Gordon]

Oh, I wouldn't go so far as to say it's a pure supply shock. But enough of it is to warrant hesitating before implementing QE.

[Nick Rowe]

Scott: I can think of 6 differences between Canada and the US that are relevant for the current discussion:

1. We have a similar fiscal stimulus, but went into the recession with a lower debt/GDP ratio, a declining debt/GDP ratio (even a small surplus), so I am less worried about fiscal policy than I would be in the US.

2. House prices did not bubble as much as in the US on average (though they did bubble a bit), and prices have only recently begun declining. There were fewer subprime mortgages, generally higher home equity ratios (no mortgage interest deductibility), and mortgages are (except in one province) recourse mortgages (the bank will come after you for the negative equity difference if you just mail them the house keys).

3. The banks (touch wood) seem to be in OK shape. We had one market in Asset Backed Commercial Paper freeze up totally, and an argument on whether or not the banks should be on the hook, but it got sort of resolved. But the non-bank sector of financial markets is not good.

4. We are much more of a resource based economy, and the change in the world prices of commodities is something we can't do much about, and means a reallocation of demand for labour, both between industries and geographically.

5. We are smaller and hence much more open. Net Exports around 40% of GDP, and so the exchange rate plays a bigger role in our thinking about AD. And a world recession has a greater impact on the distribution of demand.

6. Since 1991, we have had an explicit inflation target, and both the Bank and the government have signed onto that target, and people have gotten used to it. It's an explicit commitment, not just a recent forecast of Fed members' judgment of what they think might be consistent with a dual mandate. That's not to say that expected inflation is fully anchored at that target (it isn't), but it is less unanchored than in the US, I think.

I'm more in favour of QE in Canada than Stephen is, but I agree that the case here is less clear-cut than it is in the US.

[Nick Rowe]

Off-topic, but while we are on the subject of Canada-US differences that are relevant to your own interests: The Bank of Canada has paid interest on reserves for several years. The "deposit rate" has always been set 0.25% below the target for the overnight rate, but in Tuesday's announcement, which lowered the overnight rate target to 0.25%, the deposit rate was kept at 0.25% also. But banks have small deposits at the Bank of Canada (unlike the massive deposits in the US). And the weird thing about the liability side of the Bank of Canada's balance sheet is not the ballooning of bank reserves, but the ballooning of government of Canada deposits at the Bank of Canada.

[JKH]

The interesting thing is that both the Fed and the Bank of Canada had a choice of paying 25 basis points or zero at the zero bound.

They both chose 25 basis points.

Why?

One of these days, we're going to see a more or less official explanation from both of them as to why they made that choice.

And I predict the explanation will be consistent.

The common theme will be that neither of them is concerned with misperception that 25 basis points is somehow relevant or substantial to the idea of “hoarding” of reserves. It isn’t. Nor is the idea of “hoarding” reserves itself.

In the Fed’s case, the reason to pay 25 basis points is to compensate the banking system for the fact that the Fed has forced the massive excess reserve system upon them.

In Canada’s case, although the excess reserve levels are relatively insignificant right now, Carney has indicated he wants to be prepared for the contingency of more aggressive QE, however unlikely this may be. If that happens, the Bank of Canada may want to use an expansion of excess reserves as an alternative funding source, rather than additional government deposits, and in that case will be in exactly the same position as the Fed with respect to the reasoning for paying 25 basis points. So they’re doing it now to be prepared for that possibility.

[JKH]

I meant:

"compensate the banking system for the fact that the Fed has forced the massive excess reserve levels upon it"

[westslope]

Further weakness in the Canadian dollar will keep pressure up on Canadian prices in the intermediate term. It all fits quite nicely with Stephen's deterioration of the terms of trade, and negative supply side shock story.

The Bank of Canada needs to stick to its inflation targeting mandate. At some point the target CPI rate should be reduced to 1% or even 0%. The long term is now and that bit of slogan wisdom should apply above all to economic policy.

As an aside, I'd like to lament how "quantitative easing" is becoming another misused buzz word like 'governance' or 'employability'. Somebody should send a memo to the mainstream media explaining exactly what QE is and that while it is not a fancy, clever synonym for 'business as usual' monetary policy of the past, it is not that fundamentally different.

Another related reflection: When are Canadians going to lose their 'monkey see, monkey do' attitude with respect to the USA?

[Nick Rowe]

"As an aside, I'd like to lament how "quantitative easing" is becoming another misused buzz word like 'governance' or 'employability'. Somebody should send a memo to the mainstream media explaining exactly what QE is and that while it is not a fancy, clever synonym for 'business as usual' monetary policy of the past, it is not that fundamentally different."

YES! Monetary policy has always been "printing money" and getting it into circulation. The only difference is in HOW it is put into circulation. What does the Bank of Canada buy with it?

But we've already lost that terminological battle.

[Scott Sumner]

Stephen and Nick, I don't know enough about smaller countries in general, or Canada specifically, to have an intelligent opinion on whether they need more AD. But I do have a few general observations.

1. If Canada is engaged in substantial fiscal stimulus, it is clear that the Canadian government thinks they need more AD.

2. I have never in my entire life heard a good argument for using (demand-side) fiscal stimulus. It seems to me that monetary stimulus is superior to fiscal stimulus in every way. It is far more powerful and it doesn't run up huge budget deficits. So my recommendation to Canada would be to stop its fiscal stimulus and replace it with monetary stimulus. That is the sense in which I think Canada needs QE.

3. I assume that small countries never have any trouble depreciating their currencies if they really want to. The could set the Canadian/ROW basket exchange rate as low as they want, and supply unlimited CAN$ at that rate. BTW, I'm not saying that is what they should do. I don't believe that it is, it's just that that sort of thought experiment make me think that people in smaller economies, or even mid-sized economies like Canada, don't have to worry about liquidity traps.

I have a feeling you 2 will tell me I don't know what I am talking about when it comes to Canada, and you are probably right. I am throwing out these rather sweeping statements because I am interested in the Canadian reaction to my perspective.

[Nick Rowe]

Scott:

1. Agree.

2. For stabilising AD, I generally much prefer monetary policy, rather than fluctuations in fiscal policy. But, we did need to loosen fiscal policy *a bit* sometime, because you can't have the debt/GDP ratio falling forever. It's not obvious to me what the optimal debt/GDP ratio is. If there's a strong demand for government bonds from some quarters, why not satisfy it? (You could even argue the Samuelson Exact Consumption Loan model at some point, and say that if the interest rate on government bonds is less than the growth rate of GDP, on a long-run basis, the market is telling you to run a Ponzi scheme, and not let the supply of bonds drop further.

Plus, since I confess to uncertainty (any of our theories may always be wrong) there's something to be said for a "belt and braces" (belt and suspenders?) approach. And a money-financed deficit is belt and braces.

(It's not obvious whether a fiscal deficit would *in fact* be money-financed under the inflation-targeting regime, but I think it will be, at least in part.)

3. Yes. But Forex Market operations would be very unwise at the moment, especially given the risks of further inflaming "buy domestic" policies, and our vulnerable position. Plus immoral to beggar thy neighbour. On the other hand, if the exchange rate fell because we had an expansionary monetary policy, without reducing any other country's money supply, that would be OK.

[Andrew F]

Nick, is there a good reason that a country can't have a falling debt:GDP ratio forever? I thought the debt in question was net debt, so even if there was demand for government bonds, the proceeds from those bonds could be invested elsewhere. I will agree though that debt repayment is not necessarily wise. The ROI of repaying debt is lower than many other investments the government could make.

[Scott Sumner]

Nick, Thanks. A few comments. Too much government saving is certainly not a problem in the U.S., and I suspect even Canada probably saves too little. I have a theory that almost all welfare states save far too little, and that an optimal budget would look more like Singapore's (which saves close to 40% of GDP and runs huge surpluses.) If you do that you can get rid of the welfare state and cut taxes to low levels, and still have universal coverage of health, welfare, etc.
I understand the belts and suspenders argument, but my approach is slightly different. First you set monetary policy at a level expected to produce on target AD growth (Lars Svensson's criterion). Then you see if there is any need for fiscal policy. But you then notice that any fiscal action would simply move expected growth away from the target.

On exchange rates is there really much difference between using monetary policy to depreciate the CAN$, say by buying euro bonds with CAN$, or simply doing OMOs with Canadian government bonds with the hope of depreciating the Can$. I had always assumed the one tool one target idea meant that monetary policy couldn't simultaneously hit two targets.

[Patrick]

Singapore also uses personal medical savings accounts, which are not without major problems:

http://www.ncbi.nlm.nih.gov/pubmed/15312709

http://www.cmaj.ca/cgi/content/full/167/2/143?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&fulltext=medical+savings&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&sortspec=relevance&resourcetype=HWCIT

[Scott Sumner]

Patrick, Thanks for the info. I don't agree with the view that demand-side cost containment would not work. Most of the medical expenses that I incur would not happen if I paid out of a HSA. And I don't think I am that atypical. Singapore spends 5% of GDP on health care, vs. 15% in the U.S. (What is Canada, about 10%?) They have universal coverage like Canada, we have 45 million uninsured, they have the longest life expectancy in the world, Canada and the U.S. don't. I'm sure that there are many flaws in the Singapore system, but it's not obvious to me that there are any better systems out there. But I admit that this may reflect my American perspective. When you are spending 15% of GDP on health care, it's pretty easy to see cost advantages in systems spending only 5%.

[Patrick]

https://www.cia.gov/library/publications/the-world-factbook/rankorder/2102rank.html

Not to quibble but Singapore is #4, Canada #8. The difference is .75 of a year. The US is #50. Though life expectancy is probably not a great indicator - too many confounding factors rolled-up in the number (e.g. gun violence, traffic fatalities, etc ...).

In general, my bottom line is that sick people should not be left to die or languish because they can't pay, and nobody should be forced into bankruptcy due to illness or injury. There are probably all sorts of ways to do this.

The way Singapore uses mandatory savings is interesting. In some ways not so different from taxes - a big chunk of your pay is deducted right off the top. I guess the trick is regulating how those massive pools of capital are managed so they don't get pilfered or clobbered when something like a financial crisis hits. I can just imagine what would happen if a Bernie Madoff got his paws on the money ... torches and pitchforks all around.

[Nick Rowe]

Andrew F: If a country has falling debt/GDP ratio forever, debt will go negative, keep on declining, and eventually the government owns everything. Fiscal conservatism leads eventually to communism!

Now, of course, it could asymptote the debt/GDP ratio to zero (or some such level). But if you hold taxes and government expenditure constant as a share of GDP, and you have a surplus, the surplus will get bigger and bigger over time, as you pay off the debt, and interest payments on the debt decline. So you need to cut taxes or raise spending eventually. Why not do it now?

Scott: I agree that US government saving has been too small. But it's not at all obvious you can say the same for Canada. Partly it depends on projections for demographics, government pensions, etc., and partly it depends on what you think is the optimal long-run debt/GDP target. There's almost no theory of the latter. The Barro/Pigou theory (based on the cost of tax distortions being quadratic in the tax rate) says that the optimal debt/GDP ratio is whatever it is now (cycles aside). Set tax rates at some constant sustainable level (because only a tax rate which is expected to be constant over time will minimise the present value of deadweight losses). But that leaves the long-run debt/GDP level indeterminate. It follows a sort of random walk, if there are shocks.

I agree that if you are certain that monetary policy will work, there is no point in using fiscal policy (as an AD stabilisation tool). Any change in fiscal policy will and should be fully offset by monetary policy. 100% de facto crowding out. The LM curve is vertical, because the Bank adjusts the supply of money to make it vertical. But if we are less than 100% certain that monetary policy will work under current circumstances -- if we are unsure whether the Bank has "run out of ammunition", or is "firing blanks", I want both the Bank and the government to be shooting at the bear.

On the comparison between an exchange rate instrument and an OMO instrument: in normal times I would agree with you. No difference. The whole concept of "manipulating" exchange rates is silly, because any central bank action will affect the exchange rate. But it's not 100% certain (to me) that the same thing is true today. That's what the current argument is about. If the interest rate has failed as an instrument, can some other instrument work?

[Nick Rowe]

Oh, and if the bank of Canada bought pounds sterling, and the Bank of England did nothing, this would expand the Canadian money supply, but contract the UK money supply. If the Bank of England then buys Canadian dollars, both countries end up back where they started (except each has bigger FOREX reserves). That seems beggar thy neighbourish.

[Scott Sumner]

Nick, I'm not sure about your last point. Aren't foreign reserves actually securities (not cash?) If so, then if two countries engage in competitive devaluations against each other, then they fail in terms of exchange rates, but they succeed in the sense of depreciating their currencies against goods and services. Right now we need all the countries of the world to agree to competitively depreciate their currencies against each other until the price of goods and services starts rising at about 2% per annum.

On the 2 monetary instrument question, I agree that interest rates no longer work, I was arguing that the other two following instruments are basically one:

1. QE to boost the price level
2. Currency depreciation to boost the price level

Regarding the "we might be out of ammunition" issue,
I have found no economist who disagrees with the following proposition:

If the Fed (or BOC) bought up the entire stock of all assets on planet Earth, and paid in Federal Reserve notes, prices of goods and services would rise at an above target rate. If we assume prices are currently rising at a below target rate, there is some intermediate monetary base increase that is expected to produce on target inflation. Ammunition isn't the issue.

What is the issue then? One theory is the so-called "long tails problem." Because of policy lags, and the liquidity trap, the monetary base setting that leads to a 50-50 chance of inflation overshooting 2%, also leads to a much higher expected rate of inflation (due to a small risk of hyperinflation.) I think this is a phony concern (as TIPS/conventional bond yield spreads are observable in real time), and thus the "policy lags" problem is not worrisome, but lots of people seem to fear it, if only subconsciously.

Patrick, Yes, you are right. I knew about Andorra and Macao, but the latter isn't a country, and I thought Andorra was too small to bother with. You are right about Japan, I (wrongly) assumed that Singapore had caught up by now as Singapore's life expectancy has been rising much faster than Japan's. I also forget how impressive Canada's numbers are. (By the way, when one looks at Macao and HongKong it makes one wonder whether air pollution and stress are really all that bad for one's health. Must be that Asian diet.

To me, there is a huge difference between taxes and forced saving. I think that is why Singapore was the first sizable (more than 1,000,000 people), non-oil country to surpass the U.S. in PPP GDP per capita in 100 years. And in 20 years they will be far ahead of us. Forced saving is much more efficient, much more pro-growth. Here is my source:

http://en.wikipedia.org/wiki/List_of_countries_by_GDP_(PPP)_per_capita

[Nick Rowe]

Scott: "Nick, I'm not sure about your last point. Aren't foreign reserves actually securities (not cash?) If so, then if two countries engage in competitive devaluations against each other, then they fail in terms of exchange rates, but they succeed in the sense of depreciating their currencies against goods and services. Right now we need all the countries of the world to agree to competitively depreciate their currencies against each other until the price of goods and services starts rising at about 2% per annum."

Agreed. Yes, I suppose forex reserves are mostly held in bonds. In which case two countries doing unsterilised forex operations, buying each other's bonds, is equivalent to each doing an OMO, except they get more forex reserves.

On "ammunition", I can't rule out the possibility that the central bank might have to buy a very large amount of assets to get expectations to change. And then if expectations did change, they might suddenly change too much, requiring a quick reversal. That's why the control instrument seems to be so important, and the precise commitment the bank would make, and if credibility could be achieved quickly and smoothly.

Again, I am not as clear on this as I would like to be. More a worry at the back of my mind.

BTW, check the recent comments on the mechanical metaphors post. You are definitely making some headway here!

[Scott Sumner]

Nick, I appreciate your willingness to spend so much time with my proposal. I agree with your first point above, and half agree with the second. I think central banks are concerned about overshooting, and that holds them back. That's a big flaw with QE. But I think central banks lose sight of the fact that there is so much uncertainly about how much cash is needed to escape a trap, precisely because cash hoarding rises sharply in a liquidity trap. I've planned a paper on this for some time, but just haven't had time. The advantage of NGDP targeting is this:

It reduces credibility uncertainty. Under QE you don't know how much base money to inject, because you don't know how much credibility the policy will have. By credibility uncertainty, I mean NGDP growth uncertainty. People hoard cash when NGDP is falling fast, not when it is rising 5%. On the other hand with 5% NGDP futures targeting, you pin down expectations, and so there is no credibility uncertainty. So there is no excuse for fear of monetary stimulus.

Let me add that even in the current environment, I am not willing to give the Fed a pass. By monitoring the TIPS market, they can come pretty close to NGDP targeting, and they can certainly prevent any gross overshooting toward high inflation. Focusing on the TIPS market will prevent expected inflation from rising, and it is expected inflation, not actual inflation, that is so hard to unwind once unleashed.

[Nick Rowe]

Scott: It reminds me of the 1970s/80s overshooting, when central banks were trying to reduce inflation by slowly reducing the growth rate of the money supply. When they eventually succeed in bringing inflation down, and expected inflation and fell, the demand for money increased. So it was almost impossible to bring inflation down smoothly, even with monetary gradualism. The result was high real interest rates and a recession. So there was a definite trade-off between credibility and overshooting.

I think we are seeing the same problem now, only in reverse.

Yes, switching to a better control instrument, and anchoring expectations and credibility around that instrument, would seem important.

[Scott Sumner]

Nick, In the 1970s and 1980s we faced the problem of bring down inflation expectations. That's really hard. As long as we watch TIPS spreads, We should never have a repeat of the 1970s. 5-year inflation expectations should never be allowed to rise above 3% under any circumstances. If they do, there will be no "policy lags" excuses allowed this time. They didn't have a TIPS market in the 1970s.

By the way, I completely agree with your view of what went wrong with gradualism. I hope we never go back to that monetarist approach (although I like many aspects of monetarist theory.)