Oil prices, the Canadian dollar and the kink at parity

The Canadian dollar has been bouncing around parity with the US dollar recently, and the Fed's adoption of another round of quantitative easing plus the relaxing of the Chinese government's tightening policies has people (including me) thinking  that USD-denominated commodity prices will rise, and with it the Canadian dollar.

But will it? The last time we were in this situation, something odd happened when the CAD-USD exchange rate hit 1.0.

Here is a scatter plot of daily CAD-USD exchanges rate and USD-denominated oil prices since 2002:

(Yes, I know that there are other things affecting the exchange rate other than oil prices. One of the things this graph tells me is that for the present purposes, they don't matter nearly as much as does oil.)

As best as I can make out, the story goes like this. During the 2000's, oil prices drifted up and dragged the CAD with it along what looks to be a fairly straight line. By the time oil prices hit $US 80 in September 2007, the CAD was near parity with the USD. Oil prices continued to rise, and it looks as though the relationship that had held for six years remained stable for a couple of months more: the Canadian dollar continued to appreciate.

But here's where it gets weird. For some reason, the CAD depreciated back down to about USD 1.0 in November 2007, and it fluctuated around parity for the next eight months or so, even though oil prices almost doubled. The explanation for this doesn't seem to be interest rate differentials; Mike Moffatt's small forecasting model breaks down here as well.

After that, the financial crisis hit, and everything reverted back to where we were in the mid-2000's. Since then, oil prices and the Canadian exchange rate have been moving back up the same path they traced out earlier, and we are now roughly where we were back in September 2007.

What happens next? As I mentioned earlier, I'm expecting USD-denominated oil prices to continue increasing, but the linear relationship could break down again at the same point it did last time, for at least two reasons:

• A story based on psychology and that fact that 1.0 is an important focal point. (Someone will no doubt be able to articulate this much better than I can)
• The Bank of Canada would prefer to avoid a sharp appreciation, and might adopt a more expansionary policy as the Fed embarks on QE2.

Or not. It's not as though anyone has a great track record in predicting oil prices and exchange rates.

 

 

Comments

[Just visiting from Macleans]

Q: Suppose you had two exporting sectors - energy and manufacturing. Rising oil prices leads to rising currency - but once to get close to/above parity, the manufacturing sector can no longer compete (as per the graph MM provided earlier here: http://tinyurl.com/36gbsqj ) I'm not suggesting that's necessarily what's happening here, just wonder what it might look like.

[Stephen Gordon]

Perhaps, but it's not clear why being competitive should be treated as a binary phenomenon, nor why the kink would happen be at 1.

[Just visiting from Macleans]

Except the kink only occurs over less than one year before it rescinds if I'm reading the graph right (blue Dec 07- July 08, and then partly into the red).

I don't know enough about oil contracts (what pct is on spot market) but perhaps there is some lag in longer term deliveries if there is a sharp run up in prices - as occurred during this period.

[Phil Koop]

The actual spike in oil prices was pretty narrow; there was a flat section from November 2007 - February 2010, so it was really just nine months. How much physical oil, in heavy grades, was actually transacted at these prices? If I recall correctly, refinery utilization was declining at the same time that crude prices were rising. How many physical supply contracts were signed at these prices? How many futures contracts were left open for physical delivery? Is there any particular reason why purely notional transactions on Nymex or ICE should have boosted CAD? Most physical oil is traded on the forward market, not the spot.

[Just visiting from Macleans]

Suncor - avg sales price ($ Can) crude

Q1 '06 $65.75
Q2 '06 $75.34
Q3 '06 $71.99
Q4 '06 $60.65

Q1 '07 $65.61
Q2 '07 $71.01
Q3 '07 $76.97
Q4 '07 $82.36

Q1 '08 $96.16
Q2 '08 $121.12
Q3 '08 $116.32
Q4 '08 $61.53

Q1 '09 $59.14
Q2 '09 $59.00
Q3 '09 $61.70
Q4 '09 $64.81

[Mike Moffatt]

I think we might be all taking the oil->dollar relationship too literally. Oil and the CAD/USD may both rise together because they're both denominated in USD, so increasing weakness in the USD will cause both to rise.

[Bob]

Factor in the Dow and see what happens.

[Nick Rowe]

Wow! That does look like a kink. If that thing were upside down, I would have said it's a short run wage-Phillips Curve, with a zero lower bound because zero is a magic number.

Yep, it's saying "focal point" to me. A dollar really is a dollar. Well, a little bit, anyway. Because some people think that some people think......that some people think it is a dollar. Where "some people" might include the Bank of Canada, somewhere along the line.

Either that, or forex market traders are math-challenged, and find it easier to multiply and divide by one.

[Just visiting from Macleans]

What about the green line, Oct '07-Nov'07? Something big happened around then to put the whole scatter diagram out of whack. Maybe some new PhD hire devised a new and untested algorithm for currency trading. Or the old program had a Y2K-ish glitch at $1.09 .

[Phil Koop]

JVMC - any volumes to go with those numbers? And any delivery date data? How much new capacity did the high price contracts induce?

[Just visiting from Macleans]

I got them from the Suncor annual reports.

pg 102 (48 on pdf viewer) 2009:
http://www.suncor.com/pdf/fins-suncor_annual_report_2009_en.pdf

pg 92 2007:
http://www.suncor.com/pdf/fins-suncor_annual_report_2009_en.pdf

Volumes etc in tables above or below those pages.

[Mike Moffatt]

"What about the green line, Oct '07-Nov'07? Something big happened around then to put the whole scatter diagram out of whack."

It only lasted 2 weeks, but I remember that time really, really well. Needless to say I was not a happy man.

[Kosta]

Great post Stephen! In general, I think Mike Moffat is right in saying that both WTI and CAD/USD are 'driven' by US Dollar weakness. However, back in 2007/2008, there was definitely a kink in both relationships. When CAD/USD spiked up to $1.09 in November 2007 (the green period in Stephen's chart), it was accompanied by a proportionate increase in WTI, but it wasn't accompanied by an increase in USD weakness (measured by the US Dollar Index, or by EUR/USD strength).

The subsequent blue period where CAD/USD hovered around parity while WTI raced up to $147 was marked by profound US Dollar weakness with the US Dollar Index hitting all-time lows and EUR/USD trading up to $1.60 on multiple occasions. It appears that for the first 8 months of 2008, the Canadian dollar lost its correlation to both WTI and US Dollar weakness.

Thinking back to 2008, why didn't CAD/USD appreciate as the US Dollar weakened against every other currency? Understanding that period should be a clue to what happens next.

[K]

I can tell you that there was no significant backwardation in the futures market (maybe $3) even at the peak of the oil market, so forward transactions don't explain anything.