The Loonie (Canadian dollar) has appreciated against the US dollar recently. The Bank of Canada expressed its concern in Thursday's Announcement.
"In recent weeks, financial conditions and commodity prices have improved significantly, and consumer and business confidence have recovered modestly. If the unprecedentedly rapid rise in the Canadian dollar (which reflects a combination of higher commodity prices and generalized weakness in the U.S. currency) proves persistent, it could fully offset these positive factors."
Can the Bank of Canada "talk down" the Loonie? If exchange rates are determined by fundamentals, it's not clear how talk could matter. But I don't think exchange rates are determined by fundamentals. I think the exchange rate is determined by the forex market's current belief about the Bank of Canada's current and future beliefs about the fundamentals. So talk can matter, because it lets the Bank of Canada let the market know what the Bank of Canada believes.
Take a simple "asset pricing" approach to exchange rate determination.
The Bank of Canada has an instrument rule (or reaction function):
1) R(t) = -mS(t) + E{F(t)}
where R is the nominal interest rate, S the US/CA exchange rate, and E{F(t)} is the Bank of Canada's expectation of the "fundamental" forces of inflationary pressure. This instrument rule is really very general. The Bank of Canada looks at all the relevant information, and sets the interest rate accordingly. All I have done is to separate out the exchange rate from all the other relevant information (and assumed a linear reaction function for simplicity).
Foreign exchange market traders set the exchange rate according to the interest rate parity condition:
2) S(t) = R(t) – R*(t) + E{S(t+1)}
where R*(t) is the foreign nominal interest rate, and E{S(t+1)} is forex traders' expectation of the future exchange rate.
Substitute 1 into 2 repeatedly, and you can solve for the exchange rate as:
3) S(t) = (1/m)"permanent"{F} – (1/m)"permanent"{R*}
And you can solve for the interest rate differential as:
4) R(t) = "transitory"{F} + "permanent"{R*}
[I am using the words "permanent" and "transitory" in almost exactly the same way that Milton Friedman used them to talk about "permanent" and "transitory" income. Friedman defined "permanent" income as a hypothetical constant stream of income that had the same present value as the actual stream of income. And "transitory" income as the difference between current and permanent income. Replace the real interest rate in Friedman's formula with the coefficient "m" from the instrument rule, and you get my definition of "permanent" and "transitory".]
That is a very simple theory of how the exchange rate and interest rate get co-determined in the equilibrium of a game between forex traders and the Bank of Canada, where the Bank of Canada reacts to the exchange rate set by the forex traders, and the forex traders set the exchange rate based on their belief about the Bank of Canada's current and future beliefs about the "fundamentals" F.
It is often said that a change in some fundamental, like the price of oil, will cause the Loonie to appreciate. That is not quite right.
First, if the price of oil is part of the set of fundamentals F, it is only the "permanent" component of a rise in the price of oil that will cause the exchange rate to appreciate (the "transitory" component causes the interest rate to increase).
Second, it is not the rise in the price of oil itself that causes the exchange rate to appreciate; it is forex traders' expectation that the Bank of Canada will react to the price of oil that causes the exchange rate to appreciate. More precisely, if forex traders believe that the Bank of Canada believes that the price of oil is part of the set of "fundamentals" F to which it must react, if it wishes to keep inflation on target, then a rise in the price of oil will cause the Loonie to appreciate.
Fundamentals do not affect the exchange rate. Beliefs about beliefs about fundamentals do affect the exchange rate.
By talking, the Bank of Canada can affect forex traders' beliefs about the Bank of Canada's beliefs. Talk can thereby affect the exchange rate.
The Bank of Canada can talk the Loonie down, if it can convince forex traders to change their beliefs about what the Bank of Canada believes.
Normally the Bank of Canada would not bother trying to talk the Loonie down. If the rise in the Loonie more than offset any increase in the fundamentals F, then the Bank would just cut the rate of interest. But with the overnight rate already at the lower bound, the Bank would need to engage in "Quantitative Easing" if it wanted to loosen monetary policy. But I get the sense that the Bank of Canada would rather not engage in QE, if it can avoid it. So it is not surprising to see the Bank trying to talk the Loonie down instead.
The Bank is saying to the forex market: "Look, we think that the Loonie has appreciated more than fundamentals warrant, and unless the Loonie falls a bit, we will engage in QE (or postpone raising the overnight rate)". It's not a threat, just a statement about the Bank's beliefs.
Worthwhile Canadian Initiative 
“Fundamentals do not affect the exchange rate. Beliefs about beliefs about fundamentals do affect the exchange rate.”
You should tell the
Bank of Canada
“These main elements remain in the current version of the Bank’s exchange rate equation, as derived and tested in Issa, Lafrance, and Murray (2006). Formally, long-run movements in the bilateral real Can$/US$ exchange rate, rfxt, are explained by movements in real non-energy commodity prices (comt), real energy commodity prices (enet), and the nominal bilateral interest rate differential (intt).”
Declan: I have been telling anyone who will listen!
But what I am saying does not contradict econometric studies which find that fundamentals explain the exchange rate. Beliefs about beliefs about fundamentals are often highly correlated with the fundamentals themselves. (And current fundamentals are also correlated with “permanent” fundamentals.)
And those correlations are clear enough from the data. But the errors from those sorts of models are still very large. Which no doubt explains why the Bank is reluctant to use the exchange rate as an instrument of policy – they have only a very vague idea of what the rate “should” be.
Yep. If my theory is correct, some of the errors in the existing equations may come from the distinction between current vs, “permanent” values of those fundamentals. Actually, I’m looking for someone who knows macro econometrics better than I do to test this. You doing anything this Summer, Stephen? 🙂
Estimating the Issa, Lafrance, Murray version of the old SVN exchange rate model provides a value for the Canadian dollar of around 82 cents – the current level of the loonie is more consistent with oil prices around $80-90 and higher non-energy prices. I think traders have gotten ahead of the fundamentals.
Agreed that the traders are ahead of the fundamentals, but the fundamentals (flat supply since 2005, plummeting rig counts, rusting energy infrastructure, even the horrible demographics of the oil industry workforce), suggest they aren’t going to be wrong for long.
The oil price crash was a temporary phenomenon brought on by the collapse of world trade last fall. Now that it looks like we averted the end of civilization, my bet is we’ll be back around $100 by the end of the summer.
Nick, I always thought the term “fundamentals” included the future expected path of monetary policy. If it does, then BOC talk influences exchange rates by changing fundamentals. If the term doesn’t include future monetary policy, then talk works without changing fundamentals. Have I oversimplified this so much that I lost a key point?
Scott: it depends what you mean (as in our other discussions) by “monetary policy”. At the back of my mind was the assumption that the BoC is targeting 2% inflation and will continue to do so, and is expected to continue to do so. So “monetary policy”, understood as the Bank’s 2% inflation target, never changes. The Bank does indeed respond to shocks by changing the overnight rate. But it is those shocks, rather than the Bank’s responses to them, that I think of as the fundamentals.
Or, if you define “monetary policy” as the interest rates set by the Bank of Canada, then I would say that current and expected future monetary policy (by the BoC and the Fed) is the ONLY thing that determines the exchange rate. Other “fundamentals like the price of oil etc. are irrelevant, except insofar as they affect current and expected future monetary policy.
Nick, I think I now see your point. The BOC talk gives the market new information about the BOC’s view of the relationship between fundamentals and monetary instrument settings (and hence exchange rates) consistent with a 2% inflation goal.
I think implicit in this view is that when they talk down the loonie, they are really saying “we don’t think we can hit our 2% inflation target if the loonie is so high, and we intend to hit that target, and we just thought you guys in the forex market might want to know that.”
Maybe I’m just restating what you originally wrote. Does this sound right?
Scott: that sounds exactly right. And you are restating, very simply and clearly, what I was trying to write. Which is never a bad thing to do!
Hi Nick. Interesting post as usual. I’m a little late to the party but I recalled your post when I read somewhere recently about this fellow:
http://www.amazon.ca/Big-Players-Economic-Theory-Expectations/dp/0333678265/
Your later comment really drives the point home:
“current and expected future monetary policy (by the BoC and the Fed) is the ONLY thing that determines the exchange rate. Other “fundamentals like the price of oil etc. are irrelevant, except insofar as they affect current and expected future monetary policy.”
I recognize that it’s unavoidable if there is a central bank but one wonders whether there is something fundamentally unstable (or at least less stable) about a system in which prices are so heavily influenced by market participants’ expectations about policymakers’ responses to policymakers’ view of fundamentals rather than the result simply of market participants’ expectations about fundamentals. I am thinking that market expectations about the former are more likely to be homogeneous than market participants’ expectations about the latter would be in the absence of a central bank. That seems like it might be a bad thing from a business cycle perspective.
Thanks David! And that’s an interesting comment.
Now, one of the central arguments for inflation targeting is that it makes explicit exactly what the central bank is TRYING to do. The question remains, should it go further than this?
Chris Ragan, of McGill, has argued that central banks should go further than this, and explain their theoretical perspective, to let people know, at least in principle, the Bank’s theoretical framework, and how it responds to fundamentals. The Bank of Canada’s Type 1 vs Type 2 exchange rate movements was the first fruit of Chris’s endeavours in this line. And while I disagree with some details, the Type 1 vs Type 2 distinction was a lot clearer than what went before. And I think his general argument that transparency means more than just the target has a lot going for it.
Right now, market traders, as you say, are trying to figure out how the Bank of Canada will respond to the fundamentals. We could replace the Bank of Canada’s discretion with a simple fixed instrument rule, that would tell the Bank exactly how to react to all possible indicators. That would remove one source of uncertainty. But it might make hitting the inflation target less certain. If we all really knew what the simple rule was that could best hit the inflation target, the Bank would know it too, and would presumably follow it, so we wouldn’t have to guess how the Bank would react. But if we can’t figure out any such simple rule, we face a trade-off between predicting the Bank’s behaviour better and predicting inflation better.
I’m not sure whether my answer above is correct. I maybe need to think about this some more.
Nick:
I probably didn’t make myself clear the first time. I am aware of the central bank transparency issue and the desirability of avoiding market participant errors due to “monetary surprise.” What I was trying to convey was something a little different (I think) – that the presence of a central bank (or its targeting of exchange rates in some cases) imposes a greater degree of uniformity on market participants’ exchange rate expectations. The uniformity is in some sense artificial because it doesn’t reflect participants’ best estimate of the fundamentals (which presumably could reflect a wide range of different views on the impact of all sorts of factors) but rather their best estimate of what the central bank’s response will be to its beliefs regarding the fundamentals. Greater central bank transparency in that respect doesn’t help – it just provides for more uniformity.
If you believe that the bust phase of the business cycle is a cluster of previous errors that have come to be revealed as errors, then presumably artificially induced uniformity in prior expectations has something to do with creating the cluster.
David: No, I didn’t understand you properly the first time. Now I think I do understand you. I have no idea if you are right or wrong, but it does sound new. You may well be onto something.
Take the extreme case of fixed exchange rates. That destroys the exchange rate as a useful aggregator of forex traders’ information. So other people, who might otherwise use the exchange rate to learn about something else, may be worse off. Does that capture part of your intuition?
I don’t know. I think maybe not (maybe). Perhaps I’ll do the classic economist thing:
On the one hand, if the exchange rate is fixed and somehow sustainable at that level, I suppose other prices or quantities will have to move instead and the changes in those other (formerly more stable) things may capture the similar information (as long as there was no government agency intervening in those other areas about whose actions people would have to form expectations), although obviously the information would be in different form.
On the other hand, the fixed exchange rate may not be sustainable, in which case it will obviously be destabilizing but for reasons other than those I had been talking about (i.e., the distortion of the signal built into market prices). It would be destabilizing because of an abrupt change in the central bank’s monetary policy rule (i.e., changing the peg) and thus an issue of transparency, or predictability (or something). I am not sure how to charactize if it the peg is varied according to a previously communicated rule.
Not sure if any of the above makes sense but I have been feeling sort of Austrian recently and have thus been on the lookout for things that can “coordinate entrepreneurial error”.