Zero (effectively, finally); What Next for the Bank of Canada?

So, the Bank of Canada has finally cut the overnight rate target to 0.25%, which is as low as it feels technically able to go, so is effectively zero, by its reckoning. I'm not going to discuss those technical issues on the problems with 0.00% vs. 0.25%, because it's not my comparative advantage. But this TD Economics report seems useful, if you are interested.

I called for it to cut to zero on December 16 2008, so I feel sort of vindicated. But that was as much chance as anything else. I could give no really convincing arguments why it needed to cut to zero back then, so I can't claim to have really known that it should have cut to zero.

The Bank also added:

Conditional on the outlook for inflation, the target overnight rate can be expected to remain at its current level until the end of the second quarter of 2010 in order to achieve the inflation target.

That's interesting, but less than a commitment, since its future overnight rate target is always conditional on the outlook for inflation.

What next for the Bank of Canada? That, coincidentally, was the title of my February 27, 2009 post. I don't see much reason to change what I wrote then. The Bank of Canada should conduct quantitative easing, by unsterilised purchases of an index of commercial stocks and/or bonds.

But I think I understand a little bit better now how such a policy would work. There is nothing written in stone that central banks must use an interest rate control instrument, and certainly not just one particular interest rate, on settlement balances. For most of history monetary policy has been conducted (often for very different ends, admittedly) using the price of gold, silver, or foreign exchange (and probably many other things, if you search widely enough) as the control instrument. And changes in those control instruments were certainly not ineffective. They mattered, and changed aggregate demand, real output, and prices. The fact that one out of many possible control instruments has hit a wall does not mean that all control instruments have hit similar walls.

And the overnight rate is a very peculiar control instrument anyway. It goes in a different direction in the short run to the way it will end up going in the long run. An expansionary monetary policy would cause an increase in expected inflation, an increase in expected real growth, and a recovery in financial markets and financial institutions. And all of these things would cause the overnight rate of interest, and nominal interest rates on short, safe, liquid paper generally, to rise, not fall. So the zero lower bound is not a systemic problem; it is a problem with one particular instrument.

According to a recent Bloomberg survey, the majority of 27 Canadian economists who responded thought that the Bank will and should conduct monetary policy by other means. (I have the full survey results, but am not sure whether they want me to publish them).

Being closest to the actual workings of monetary policy as it is conducted now, it will be hardest for people at the Bank of Canada especially to imagine themselves in an alternative universe with a different control instrument and hence different transmission mechanism. But some of them are also excellent inuitive economists, so I think they will.

Comments

[Dee]

It said quantitative easing involves creating new money to purchase government or private assets to encourage new bank lending. Credit easing involves purchasing assets in “credit markets which are important to the functioning of the financial system,” and may not involve creating new money, the central bank said.

Encourage new bank lending? Oh yeah, that will work. The banks will lend to businesses/consumers for altruistic reasons. ^o)

[brendon]

I like Dave Altig's metaphor for monetary policy:

"Imagine a long rope. At one of end of the rope are short-term, relatively riskless interest rates. Farther along the rope are yields on longer-term but still relatively safe assets. Off at the other end of the rope are multiple tethers representing mortgage rates, corporate bond rates, and auto loan rates—the sorts of interest rates that drive decisions by businesses and consumers.......imagine that someone had placed a series of bricks at strategic points along the segment of rope connecting short-term interest rates to broader market rates. With these bricks in place, it is simply not enough for a central bank to keep snapping short-term interest rate"

The action yesterday doesn't address those bricks and, given the latest BoC survey of loan officers, monetary policy since the beginning of the crisis hasn't so much as budged the bricks. I'm not sure why the Bank is holding off on QE/CE but perhaps we'll learn more tomorrow when the MP Update is released.

[brendon]

Just a heads up, all comments are appearing in italics.

[Nick Rowe]

i/ Let's see if that works, to turn off italics.

[Patrick]

This : might do it.

[Dee]

/i

I like the rope analogy.

[Dee]

try this /i

[Dee]

I think it's typepad. I'll try this.

[Dee]

checking

[westslope]

Could the italic 'issue' be a metaphor for conducting central bank policy? :-)

Don't wish to exaggerate the importance of it.... but I think Carney's BoC blew it. First he (the figurative "he") was too optimistic; now he is unnecessarily honest, forthcoming and pessimistic. Struck me as bureaucratic butt covering.

Cutting the rate will reduce pressure on the Canadian dollar. Is it a useful or necessary signal of BoC commitment to supply additional liquity as required? Perhaps. I'm not aware of the BoC experiencing difficulties in the overnight market.

[Patrick]

Maybe Nick can set me straight if I'm wrong on this, but I tend to disagree that Carney blew it. The more pessimistic he is the more likely the Canadian dollar is to drop. If the dollar drops, that helps our exports, since our economy is largely export dependent more exports means more employment. Sure, currency devaluation makes imports more expensive but with unemployment headed to intolerable levels it might make sense to trade purchasing power for higher employment. Obviously, not everyone can pursue this policy with global AD in the tank, but Canada is a relatively small country. And since the arrival of the Euro, there aren't that many small advanced countries with independent central banks anymore. We might just be able to pull it off... I dunno. There's probably some macro gotcha I'm not considering...

[Dee]

Pessimism brought the savings rate up in the U.S.

Britain's government is on prozac. All sunshine and fluffy clouds.

http://www.bloomberg.com/apps/news?pid=newsarchive&sid=aeJL2HXp_MFA

Good to see this board hasn't become permanently italicized.;)

Thanks to your friendly neighbourhood admin! - SG

[Nick Rowe]

westslope: I didn't think the BoC had any difficulties doing what it wanted in the overnight rate market. It's just that the overnight rate market seems to have lost its link to all the other markets, and isn't doing what the BoC wants in the output market (or isn't doing it enough). Like in Dave Altig's rope analogy.

Patrick: I'm not sure that would work. Suppose the exchange rate depreciates if there's a weak economy (or expected to be). And the weak economy helps strengthen the economy. But if the exchange rate depreciated enough to more than offset the weakness in the economy that caused it to weaken, then it wouldn't have depreciated in the first place. Does that make sense? Unless you argue some sort of slippage between expectations of weakness and actual weakness.

[Scott Sumner]

Nick, Your call for zero rates in December makes me wonder the following:

Has there ever in all of American, Canadian, European, or Japanese history been a rate cut in the early stages of a severe recession that years later looked foolish? My hunch is that there have been a few such cases when inflation was really high (1974-80) and perhaps a few in the late stages of a recession (2003?, although even that was questionable.) But I don't recall any bad calls early in a recession where inflation was not a problem. Ever. In all of known history. That's why I also felt pretty safe making that call late last year. I also notice that the strongest stock market responses to monetary policy come early in a recession. And I don't think I need tell anyone which direction the market wants interest targets moved. But that's not because the stock market likes inflation, it does poorly during periods of high inflation, even demand side inflation like 1966-72.

[Nick Rowe]

Scott:

I am going to try to answer your question in a very roundabout manner.

Suppose the central bank is controlling an instrument R to try to keep some some target variable P equal to some target level P*. (You can think of P as inflation or NGDP growth, I don't think it matters much).

The bank knows there is a lag in the effect of R on P, so targets at a (say) 24 month horizon. (The Bank of Canada says it targets 2% inflation at an 18-24 month horizon.) So it sets R(t), conditional on all other available information (It), otherwise known as the indicators, so that the rational expectation of P(t+24)=P*. In math:

E[P(t+24)/{R(t),I(t)}=P*

Under rational expectations, the bank's forecast errors [P(t+24)-P*] must be a 24 month Moving Average process, and uncorrelated with the elements in the lagged information set {R(t) and I(t)}.

(This is just Milton Friedman's thermostat metaphor, formalised, and with imperfect information added. A country with a good central bank is like a house with a good thermostat. You see fluctuations in monetary policy (the amount of oil burned in the furnace), fluctuations in the outside temperature (other indicators), and fluctuations in the inside temperature (inflation), but there is no correlation between oil burned and inside temperature. So it looks as though monetary policy has no effect on inflation.)

So looking back at at any data set, you will see fluctuations in inflation (or NGDP growth) as a 24-month MA process, but they ought to be uncorrelated (if the bank is not making systematic mistakes) with 24-month lagged monetary policy (and anything else known to the bank 24 months ago). I think this means that it will always appear that the bank is behind the curve in retrospect, looking back at any recession, severe or otherwise. This misses the recessions that didn't happen because the bank got policy right. And it misses the booms that happened when the bank made the opposite mistake (in hindsight) and offset an impending recession that wasn't really there.

I'm on the CD Howe's Monetary Policy Council (our shadow Fed). http://www.cdhowe.org/display.cfm?page=monetarySynopsis With hindsight, I was less behind the curve than the Bank (Yey!, though I got other things wrong in the past), but the Bank was not more behind the curve than our Council, so the Bank was not worse than "expert" opinion in real time.

BTW, the analysis I did above seems to have some connection to your futures targeting insight. Here's some more, if you are interested:
http://economics.ca/2003/papers/0266.pdf
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/01/why-theres-so-little-good-evidence-that-fiscal-or-monetary-policy-works.html