Re-Writing History - the Canadian Dollar in the 1990s

From a widely cited Paul Krugman piece:

Yep, you can have fiscal austerity without contraction if you have a massive devaluation against your main trading partner. So we can have austerity without a new depression as long as all the world’s major economies devalue against … oh, wait.

And monetary policy, of course, wasn’t up against the zero lower bound, so the Bank of Canada could and did offset fiscal austerity with looser monetary policy (which partly explains the drop in the loonie.)

Unless one defines partly as barely, there are some problems with this version of history:

Paul Martin's famous austerity budget was released on February 27, 1995 so it seems as good a place to start as any. According to the MERT dataset we have the following data:

• Canadian Effective Overnight Rate: 8.12
• U.S. Effective Federal Funds Rate: 5.99
• Difference: 2.13
• WTI Cushing Oil Price: 18.64
• $1 CDN = $0.71613 U.S.

Warning bells should be going off when we look at that last figure. On the date the austerity budet was released the loonie was worth under 0.72 USD. This despite the fact that the Canadian dollar was trading over 0.89 USD in November 1991 - a 17 cent loss in the 3 1/2 years before the austerity budget. In other words, the "massive devaluation" (Krugman's words) largely happened before the austerity movement in Canada. To avoid the valid criticism that I cherry picked dates, I decided to examine the value of the Canadian dollar on February 27 of each year (if the date was a weekend, I picked the closest trading date). Here's the data in convenient bar graph form:

The Canadian dollar fell by 0.2466 USD in the 11 years between 1991 and 2002, before starting to rise again. However of that 0.1523 (61.7%) happened in the four years between 1991 and 1995. In fact, the Canadian dollar was roughly the same value in 1998 as it was in 1995. It's hard to see how fiscal or monetary policy affected the Canadian dollar at all during most of the 1990s.

First consider the period before 1998. Here's the relevant data on February 27, 1996, one year after the budget:

• Canadian Effective Overnight Rate: 5.17 (down 2.95)
• U.S. Effective Federal Funds Rate: 5.20 (down 0.79)
• Difference: -0.03 (down 2.16)
• WTI Cushing Oil Price: 19.65 (up 1.01)
• $1 CDN = $0.72754 U.S. (up 0.01141)

Krugman is correct that Canadian monetary policy is far more aggressive (using nominal interest rates as a measure of the tightness or looseness of monetary policy, Sumnerian argument aside) after the budget than U.S. policy. However the Canadian dollar has moved in the wrong direction (slightly). I can think of two possible non-mutually exclusive explanations:

• Monetary policy is acting with a significant lag, so the drop in the loonie hasn't shown up yet.


• Something else is the driving force for movements in the Canadian dollar.

Fast forward one year later to 1997:

• Canadian Effective Overnight Rate: 3.11 (down 2.14)
• U.S. Effective Federal Funds Rate: 5.29 (up 0.09)
• Difference: -2.18 (down 2.15)
• WTI Cushing Oil Price: 20.80 (up 1.15)
• $1 CDN = $0.73131 U.S. (up 0.00377)

Again, Canadian monetary policy is still more aggresive than U.S. monetary policy. However again the Canadian dollar has moved in the wrong direction (slightly). I can think of two possible non-mutually exclusive explanations:

• Monetary policy is acting with a really significant lag, so the drop in the loonie hasn't shown up yet.


• Something else is the driving force for movements in the Canadian dollar.

On to 1998 and a lower Canadian dollar:

• Canadian Effective Overnight Rate: 4.87 (up 1.76)
• U.S. Effective Federal Funds Rate: 5.59 (up 0.30)
• Difference: -0.70 (up 1.48)
• WTI Cushing Oil Price: 15.44 (down 5.36)
• $1 CDN = $0.70249 U.S. (down 0.02882)

Canadian monetary policy becomes tighter in 1997-1998, while the U.S. effective Federal Funds rate increases only slightly. This causes 2/3rds of the interest rate gap to close. Yet the Canadian dollar has gone down instead of up.

Lastly, let's consider 1999:

• Canadian Effective Overnight Rate: 5.01 (up 0.14)
• U.S. Effective Federal Funds Rate: 4.84 (down 0.75)
• Difference: 0.17 (up 0.87)
• WTI Cushing Oil Price: 12.31 (down 3.13)
• $1 CDN = $0.66339 U.S. (down 0.03910)

The interest rate gap has reversed signs again with the U.S. having lower rates than Canada. Despite all this, the Canadian dollar has fallen.

Unless monetary policy is working with a 3+ year lag, it is very difficult to see how Canadian monetary policy in 1995-1996 caused a drop in the Canadian dollar in 1998-1999. The data suggests that something else may be at play here.

Comments

[Stephen Gordon]

OTOH, I think it's fair to say that the Bank could and did relax monetary policy when Martin started reducing spending. In March 1995, the Bank Rate was 8.47%, and by March 1997 it was 3.25%.

But I think that you're right to say that there were other, possibly more important things driving the exchange rate. My candidates would be oil and/or other commodity prices.

[Mike Moffatt]

"OTOH, I think it's fair to say that the Bank could and did relax monetary policy when Martin started reducing spending."

Absolutely agreed there - that's pretty uncontroversial and fits the data well. My issue is more the middle part of the "fiscal austerity -> relaxed monetary policy -> lower Canadian dollar -> increased net exports" story. It simply does not fit the data well, unless monetary policy works with a longer lag than generally assumed.

[Adam P]

You should have titled the post "re-writing Krugman's point".

First of all: " the "massive devaluation" (Krugman's words) largely happened before the austerity movement in Canada". Krugman's point doesn't in any way depend on the devaluation happening before the austerity program begins.

Secondly: "It's hard to see how fiscal or monetary policy affected the Canadian dollar at all during most of the 1990s". It is completely irrelevant to Krugman's point whether monetary policy caused the drop in the loonie or not. You could delete his paranthetical comment and his argument would still be just as valid.

Further, while nobody thinks that there could be a 3 year lag from a monetary policy change to effect the exchange rate it's perfectly reasonable for the devaluation to occur in anticipation of the easing (thinking of the drop from 1991-94 here). Moreover, there are enough frictions on the real side of the economy that it can take years for a large devaluation to fully show up in the trade balance.

Finally, Krugman's point about Canadian monetary policy having more room to be effective because they were far from the zero bound is independant of the exchange rate. Regardless any effect of monetary policy on the exchange rate, in an economy that is not liquidity trapped monetary policy can induce private expenditures to make up for the reduced government spending. That is much more difficult for todays Fed and ECB.

[Adam P]

In fact, if the monetary easing didn't have anything to do with the devaluation of the loonie then that only makes Krugman's point stronger.

In the quote you cite if we change the second paragraph to:

And monetary policy, of course, wasn’t up against the zero lower bound, so the Bank of Canada could and did offset fiscal austerity with looser monetary policy (in addition to the drop in the loonie.)

then the argument only strengthens.

[Kevin milligan]

What Adam P said. You are conflating two separate arguments that Krugman makes.

A) that doing fiscal contraction in the presense of a massive devaluation (however that devaluation is caused) is a different beast than doing the fiscal contraction without a devaluation.

B) The fiscal contraction contributed to the fall of the loonie.

I have no opinion on (B), but I don't see how the argument in this post addresses (A) at all.

[Mike Moffatt]

"A) that doing fiscal contraction in the presense of a massive devaluation (however that devaluation is caused) is a different beast than doing the fiscal contraction without a devaluation.

I have no opinion on (B), but I don't see how the argument in this post addresses (A) at all."

It doesn't. At all. What on earth made you and Adam think I was trying to?

I specifically hilighted in bold the part I took issue with, I'm not sure how I would have made this any clearer.

[Mike Moffatt]

That being said, it'd be interesting to hear an explanation how a tumble in commodity prices helps Canada on net.

[Kevin milligan]

Check this sentence carefully and note what part of Krugman's quote it comes from.

"In other words, the "massive devaluation" (Krugman's words) largely happened before the austerity movement in Canada"

[Mike Moffatt]

Exactly... so it couldn't have been caused by the monetary/fiscal policy changes! (unless, I suppose, Forex traders saw the changes in monetary/fiscal policy coming several years in advance, but IMO that's a pretty big leap)

[Adam P]

"What on earth made you and Adam think I was trying to?"

The fact that you brought up Krugman at all. Citing his post sets the context for what you're saying. Why dispute something that is completely irrelevant to what he is saying?

By citing Krugman's post you imply that you're talking about the same issue as he is.

If all you wnated to say was that the empirical link between interest rates and exchange rates is weak, well then nobody will argue because everyone already knows that. But if you want to make that argument sensibly then you should collect more than one datapoint, but don't worry plenty of people before you have looked at this properly and yes, the conection between interest rates and exchanges rates is weak.

[Mike Moffatt]

"The fact that you brought up Krugman at all."

I suspect you're right. I bet if, say, Thoma had written the original blog post and I had taken objection to the claim: "which partly explains the drop in the loonie" everyone would have said "yeah, that makes sense, Thoma probably shouldn't have included that paranthetical aside".

But since it's a controversial figure like Krugman, people feel a strong attachment to him either for or against. I suspect at some point someone will automatically agree with me, for the same reason.

"If all you wnated to say was that the empirical link between interest rates and exchange rates is weak"

NO NO NO NO NO. No. :) That is *not* the point of this blog post. Can you please re-read it.

My point was, in *this particular instance* the link is weak. The evidence does not support it in this case. If I wanted to make a general argument, I would have made a general argument. My case was very specific, because the claim was very specific.

[Adam P]

You've completely missed my point. If Thoma had written the post you cited I still would have inferred that you intended to say something relevant to the post you're citing and to the general discussion of fiscal austerity that has been going on in many forums.

Fine, you wnated to say that in this particular instance the link is weak. Agreed.

But this fact is completely irrelavant to the fiscal austerity discussion which Krugman's post was about.

[Mike Moffatt]

You've completely missed my point.

Maybe you should have made it more relevant to *my* blog post, since it was the blog post you were citing. (j/k) :)

[Adam P]

Point taken, then can you explain how this passage:

"the "massive devaluation" (Krugman's words) largely happened before the austerity movement in Canada"

is at all relevant to the point that *in this instance the connection between monetary policy and the exchange rate is weak to non-existent*?

(This was one of the things that made me think you were talking about the fiscal austerity discussion).

[Mike Moffatt]

"is at all relevant to the point that *in this instance the connection between monetary policy and the exchange rate is weak to non-existent*?"

Absolutely! If monetary policy had to do with the Canadian exchange rate devaluation, the devaluation should come at the same time or later. Instead it happened several years *before*. This suggests either:

1. Forex traders knew about the changes in monetary and fiscal policy changes 3+ years in advance (unlikely) OR

2. The 1991-1995 devaluation of the dollar had nothing to do with the policy changes.

Since #1 is highly unlikely, that only leaves #2.

[Mike Moffatt]

For what it's worth, my views on the larger issue, particularly on the "zero bound" problem are Sumnerian.

[Adam P]

But why didn't you say: "the "massive devaluation" largely happened before the monetary easing in Canada"?

and why bring up Krugman again? It all seems to point to you trying to say something relevant to the austerity discussion.

[Mike Moffatt]

But why didn't you say: "the "massive devaluation" largely happened before the monetary easing in Canada"?

Isn't that the same thing, though? Given "so the Bank of Canada could and did offset fiscal austerity with looser monetary policy". Unless the offset happened *before* the fiscal austerity, the timing is still the same.

[Just visiting from Macleans]

As a non-economist, I say "WTF"?

You developed your "toy" MERT based upon data from 1995 onward, admit it is too unsophisticated to handle the last few years, then go back and analyse deviations (noise) against your best fit curve over the same data period.

Man.

[Jon]

Krugman is off on partisan nonsense as usual. I cannot believe he really means what he says. There is an unsettled question about the effectiveness of fiscal policy during a liquidity trap, but its rather shocking to claim that fiscal austerity must be accompanied by devaluation--always, even during the height of the boom.

The fiscal contraction should trade-off 1-1 with private sector growth.

[Weston]

Canada is a Petrol Currency, or a Mining/Mineral thats where our Wealth is. Oil up CND Up so from what I have heard from others, should do a chart on that

[Mike Moffatt]

"You developed your "toy" MERT based upon data from 1995 onward"

No I didn't. The data is from 2001-2006.

"admit it is too unsophisticated to handle the last few years"

No I didn't. Where in this blog post did you read that.

"then go back and analyse deviations (noise) against your best fit curve over the same data period."

No I didn't. See above.

Where did you pull any of that from this blog post?

This blog post seems to be the perfect Rorschach test.

[Just visiting from Macleans]

Yeah, you're right. I'll quit reading/commenting on your blog posts - reminds me too much of Peter Bell.

[Mike Moffatt]

"reminds me too much of Peter Bell."

Peter Bell from Ivey? Peter's a terrific guy but I don't think we're that much alike. :)

[Just visiting from Macleans]

He was good in theory. But some of his cases /applications were kooky. I also gave him a hard time - don't take it personally (and he was entertaining - good at the put downs).

[Mike Moffatt]

No worries... I have great deal of respect for Peter and he's personally helped me a great deal. Being compared to Peter is a tremendous complement!

[Alex Plante]

I remember in the 80s there were a lot of rumours going around that the US would not accept the Free Trade deal unless the Canadian dollar was above a certain level. That's why the Bank of Canada raised interest and pushed up the value of the Canadian dollar in the late 80s. Canada entering into the free-trade agreement with an over-valued dollar is part of the reason the recession of the 90s was so severe in this country. The drop in tax revenue coupled with high interest rates explains why the deficit exploded in the early 90s.

[Simon van Norden]

Mike;

I'm a former Krugman student and I worked at the Bank in 1990s developing exchange rate models. I've read your post and I can't understand your argument. Please help me.

As I read it, Krugman says:
1) the BoC can and did use (relatively) looser monetary policy to help offset the effects of (relatively) tighter fiscal policy.
2) this looser policy partly explains the drop in the loonie.

I don't think you're taking issue with #1 (and I'll humbly apologize if I'm that confused.) Your beef is with #2.

But how do you investigate #2? You use 5 data points (more like 4 since you work in first differences) and show that the interest rate differential over that period doesn't seem to explain the exchange rate. You also conclude (and I and many others agree) that other factors also play a major role in affecting the exchange rate.

But as you point out, Krugman says "partly explains." As a macroeconomist, I'd say he's thinking of a counterfactual experiment without monetary easing in which the exchange rate is not as weak. That's not a testable statement and you don't claim to test it. But as an econometrician, I'd say you're confusing lack of evidence with a rejection of a hypothesis. You didn't find evidence in your 5-4 observations to support Krugman's contention. So what. That doesn't mean he's wrong, and it doesn't mean no such evidence exists.

BTW, did you do a literature search? Has anyone tried to decompose CAD exchange rate movements over this period into the effect of monetary policy and other components? Has anyone found evidence that relatively looser monetary in Canada (all other things equal) makes the CAD weaker? It seems to me that for your argument to make sense you need to claim either

1) Looser Canadian monetary does not generally make the CAD weaker.
or
2) It usually does, but (for reasons that you have not hinted at) it did not then.

Which do you prefer? or did I misunderstand your argument?

[RN]

I'm sure Adam and others of similar mind are well aware that going after Krugman is generally just a ploy to try to get readers.

Usually this ploy works, and people show up and discuss. Only on the rarest of occasions does the blogger make a defensible point where Krugman turns out to have been wrong. Almost always the blogger just ends up looking intellectually silly.

This post is a textbook-perfect example of it.

Mike has only a third the intellect of Krugman, and is probably a tenth the economist on his best day, as he has demonstrated, and been so aptly called out on by astute commentators.

But that won't stop him. He'll try again. Going after Krugman brings in readers. And that's the true goal. Marketing.

[Simon van Norden]

Disagree: people need to be kept honest and the brilliant are no exception. Intellect and ability are no substitute for being correct.

[Nick Rowe]

SvN: Yep. Hunting moose is more worthwhile than hunting mouse.

[Adam P]

True, but simplistic, incorrect (and trivial) critiques do not keep anyone honest.