« One good thing about Bitcoin | Main | Economic Inequality, Saving and Economic Growth »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Slightly off topic.... Devaluing your own currency is a doddle compared to attempting internal devaluation when you're a Euro periphery country. Anyone who can think of a way of making internal devaluation as quick and painless as devaluing your own currency deserves ten Nobel Prizes.

Ralph: more or less on-topic. Yep. That's why I think targeting inflation is better than targeting the exchange rate. Recent Eurozone experience confirms that. But sometimes we need to think about the same thing with the opposite sign. It is easier to revalue your currency than to get the same change in the real exchange rate via inflation. More generally, it is easier to have fluctuations in your nominal exchange rate than fluctuations in your price level.

Targeting an exchange rate is a form of price control. Exchange rates are just assets and we should use asset markets to determine asset prices. If we use market to determine exchange rates, a falling or rising exchange rate is neither bad or good news.

An inflation target is fine, but we need rules, like a Taylor rule. The price of an independent central bank is rules. Central bank discretion opens up the bank to every political lobby group in the country looking to tip monetary policy in their favour.

At this time, it looks to me that even an approximate application of Taylor rule would put us on the track to increases in the overnight lending rate. But, the signal that the government gave to the market by hiring the form head of the EDC, who on many occasions suggested that the Canadian economy is flagging from “expensive” exports, is that the BoC is in full discretion mode. The market will not fight the BoC if the political winds in Ottawa are blowing in the direction of devalued dollar.

Let's just use markets to determine what our exports are worth - which means a dollar completely determined by market forces with a central bank that operates on well defined rules. What I find the most ridiculous is that economists fail to point out that exports are what we pay for imports. When our dollar falls, we must generate more exports to get the same level of imports. Only in the twisted logic of “policy advice” does mirco folly turn into macro wisdom.

Avon: "Targeting an exchange rate is a form of price control."

Spoken like a finance guy!

I disagree. The government produces money. The government monetary policy must target *something*. It cannot "do nothing". It can target the CPI, or the exchange rate, or the price of gold, or....

What the Bank of Canada is really targeting when it targets the exchange rate is one price of money. Money is an asset. It can target the price of money against the USD, against the CPI, against gold, or whatever.

My own research (now out of date) suggests things like Taylor rules would do worse at targeting inflation.

Fixed exchange rates are bad, but monetary policy that uses FX operations as its main instrument seems to work in Singapore:
Not sure it will work in Canada, though.

What I find hilarious is that when a first year student takes a course in economics, she is taught how great markets are at solving allocation problems. Once the class is over, the professor returns to her professional life expounding how much better discretionary central control is than using markets.

Chun: one of the main problems with Canada using the exchange rate as an instrument (rather than a target) is political. US senators would get upset if they saw that Canada was "manipulating" its exchange rate to hit the 2% inflation target, even though the exchange rate would do much the same thing as it does now, where the Bank of Canada "manipulates" the interest rate. A Martian, using annual data, would find it difficult to tell the difference.

Avon: suppose an individual firm were producing some unique good. How much should it produce? What price should it charge? It must make some sort of decision. The Bank of Canada is that firm.

Now, given an inflation target (or an NGDP target) for monetary policy, we *might* want to use prediction markets to set the instrument, rather than use the Bank's discretion, to better hit that target. Scott Sumner is keen on that question.

I am not sure there is nothing to be done, so 1 is still an interesting question. Does it mean the economy is growing stronger resulting in the rise, and/or this is a visitation of the resource curse that risks unbalancing the economy. In the latter case, perhaps sterilization of the flows through a sovereign wealth fund are warranted. I agree these are difficult questions though.

Chun - I've raised Singapore in a lot of comments (I live in sg). One argument I've seen for MAS to target FX rate is that SG is an extraordinarily open economy - both import and export trade are > GDP. My guess is that the result is that FX rates have a quicker/more direct impact on inflation in SG than it would in US. I'm not sure where Canada falls in this spectrum. MAS also has the advantage of a single mandate - target inflation only. The government has been fairly reliable in using fiscal and immigration policy to manage full employment - or at least trying.

Hi Professor Rowe!
Even if a country's exchange rate could be one of many target variables of the central bank, is it still not a legit question to ask what are the economic impacts of a lower/higher Canadian dollar on the sectors/agents/businesses/consumers etc?
Without explaining the comparative effectiveness or efficacy of the central bank policy stance on a particular target variable (i.e. inflation or exchange rate etc), I think, an economist should still be able to identify/quantify the exchange rate's impacts on the economy. And if you are asked that question, what would your answer be?

I think the problem with the BOC targeting the exchange rate is that it can far too volatile an indicator to follow. Better to let the exchange rate changes work their way through their economy and show up into the CPI before deciding what kind of an impact the exchange rate is going to have on the price level.

And while exporters like to tout the benefits of lower exchange rates its really a lot more complicated than what has been passing as analysis on the CBC. Trade is not autarkic, countries dont simply produce within their borders and then swap their goods with other countries. In the era of globalization, there is far more trade in intermediate goods to muddy the simplistic analysis of Canadian dollar down therefore exports up.

Take cross border shopping as an example, people often assert that a lower dollar with disincentivize canadians to travel to the US to shop, keeping valuable canadian dollars within our borders. The problem with that view is that a lot of our merchandise comes from abroad and therefore a lower dollar means higher costs for Canadian retailers. While they might sell more, their margin shrinks and their profits are unlikely to be positively affected to any significant degree.

You have three groups that are effected by changes in the exchange rates, consumers, exporters who rely on foreign intermediate inputs and exporters who rely on input produced within Canada. Guess which group is the most vocal and interested in lower exchange rates.

The question I would ask is how economic theory explains a move of 10 per cent or more (some now predicting 85 cents) in the context of interest policy forecasts that have been relatively benign in any event and for a long time now.

Why the hypersensitivity of markets in this context?

I'd ask the same question about market responses to the second derivative of QE - i.e. tapering.

Because of the way the Canadian exports are priced internationally (some significant commodities are priced in the US dollar), would a lower Lonnie have the same impact across the exports spectrum? I suppose, not. Then, some sectors would face different outcome than others due to the lower value of the Canadian dollar. Just an aside, almost in all provincial fiscal forecasts, there is a statement like 'risks to assumption' (in a somewhat elasticity term) which says that 'one cent reduction in the Canadian dollar will lead to an X amount of fiscal impact'. And this ad hoc assumptions vary significantly across Canadian provinces depending to some extent on, I guess, the structure of that province's exports.
A question could still be asked: what are the likely consequences of a lower Canadian dollar on Western Canada's vs Ontario's exports as they are significantly different in terms of structure and import intensity?

Lord:"perhaps sterilization of the flows through a sovereign wealth fund are warranted. I agree these are difficult questions though."
A majority share of the oil wealth acrue to the provinces. For let's say AB, a sovereign fund might consist of canadian assets in Ontario.For a province , canadian assets are foreign, and C$ is a foreign currency. A sovereign fund in C$ sterilize the situation from a rovincial standpoint but does nothing from the BoC perspective.

Nick Rowe: "US senators would get upset if they saw that Canada was "manipulating" its exchange rate to hit the 2% inflation target"

If you don't tell them, they'll never figure it out. ;)

Let's just use markets to determine what our exports are worth - which means a dollar completely determined by market forces with a central bank that operates on well defined rules. What I find the most ridiculous is that economists fail to point out that exports are what we pay for imports. When our dollar falls, we must generate more exports to get the same level of imports. Only in the twisted logic of “policy advice” does mirco folly turn into macro wisdom.

The reason economists don't point it out is because it's not true. We live in a monetary exchange economy. We do not trade coconuts for sugarcane, we trade goods and services for cash, and the cash can and does come from anywhere. In a monetary exchange economy, the exchange between two currencies can adjust or the real volume of goods can adjust. There are far more free parameters in a monetary exchange economy than a real goods exchange economy.

Do exports pay for imports? NO!

"The exchange rate didn't just fall. Something, call it X, caused it to fall. So when we ask "Is the fall in the exchange rate good news or bad news?", what are we really asking? You can't give a good answer if you are unclear on the question."

Actually, as economists, we're mostly dismal at figuring out what caused exchange rates to fall (or rise), aren't we? What fraction of the daily or weekly movements in the value of the CAD do we think we can reliably explain on average? Who has a reliable quantitative model that explains the movements over the past few days or weeks?

Nick, how do you suggest we answer questions like that when we're highly uncertain what caused the event in the first place?

Simon: "Who has a reliable quantitative model that explains the movements over the past few days or weeks?"

I thought you and Bob Amano did? No? ;-)

JKH: Here's the way I think of it.

Very loosely: Define X as "demand minus supply for Canadian goods relative to demand minus supply for US goods".

Decompose all shocks to X into a permanent component and a transitory component.

If the Bank of Canada is getting everything just right:

The exchange rate adjusts in response to the permanent component of shocks to X.

The interest rate differential adjusts in response to the transitory component of shocks to X.

Osman: suppose Steve Poloz suddenly announced: "we have decided to depreciate the exchange rate by 10%, even though this will cause inflation to rise above target". What would happen? I don't have a clue. Nor would anyone. Because he hasn't said what the new policy would be for the future. Expectations, and markets, would be all over the place. We simply would not know what the new monetary regime would be. If instead he said he was going to fix the exchange rate at 80 cents US, and keep it there, and people believed him, then we could try to answer the question.

Funny coincidence: One of my students came to my office hour a few minutes back. He asked me "Is the fall in the exchange rate good news or bad news?" I was sure he had read this post, and was teasing me. But no!

Nick, I think it's a bad policy. They want to support Canadian export through the exchange rate. How has it helped BlackBerry or Nortel? Not much. And the impact of Canadian industrial destruction on Canada as a whole is much larger than a couple of percentage drop due to the rise in the exchange rate. Plus, it makes Canadian consumers poorer.

Personally, I am for higher exchange rate (it helps consumers) and targeted supply side policies (it's supposed to help exporters).

Alex: remember monetary neutrality. The permanent *level* of the nominal exchange rate has no real effects; it just affects the level of prices, and not the real exchange rate. But fluctuations in the nominal exchange rate, and how it fluctuates in response to shocks, can matter a lot, because prices are sticky in the short run.

Nick, I understand your "theoretical" point. From a practical point of view, the short run can last pretty long. And it will make you poorer. Your wealth as accumulated asset will diminish. Your income, I agree with you here, will come back "eventually."

Determinant: in the short run and financial sense, exports pay for imports. In the long run and economic sense, imports, by promoting efficiency, enables us to export.

Nick, and your consumption will decrease too in the short run.

Jacqes: That is why there is taxation though that may be more difficult in Canada.

Jacques: It's almost as bad as S=I. As a monetary transaction, that money can come from other places in the national accounts. Not to say that imports and exports don't have important economic roles, but it there clearly isn't an export more y, get more x imported direct relationship.

Witness China.

This discussion on exchange rates continues to highlight the difficulties I have in reconciling what goes on in economics. My background is theoretical physics, but I work as a quant with a deep interest in financial economics and modern macroeconomic theory (and I mean the technical literature).

I cannot understand how all economists don’t fully internalize the message of Kydland and Prescott nor the nuanced advances in recursive formulations that allow us to make progress in light of Kydland and Prescott’s original objections. In almost any scenario you want to dream up, discretion - especially of the kind we see today - is harmful. If we read Bernanke, and Friedman and Schwartz, we see that it was the discretion of supposed great men who caused the Great Depression. They really thought they knew better. Given the impossible task of developing some fantastic dynamic program with all sorts of levers for central bankers to pull, a much simpler and robust approach is to use something close to a Taylor rule. We simply don’t understand the system well enough to finesse monetary policy - and anything else is just pretending. Unfortunately, we are moving more toward a model that relies more and more on central bank discretion, especially in its growing role as a closet macroprudential oversight body.

In my split career - one as a theoretical physicist and the other as a quant, I can tell you that economics is a beautiful discipline that is as rich as anything in Nature. I can also tell you that economics is much, much harder than anything in physics. When you write down a new idea in physics, you work very hard to find all the ways you can be wrong. We can almost always get to the root of the problem this way - and when we can’t, we do an experiment. In economics it is much harder to know when you’re wrong and the stakes are so much higher when policy advisors are in actuality engineering the trajectory of people’s lives, sometimes for generations. We should be extremely cautious in implementing discretionary policies or trusting the judgment of experts in central control.

Determinant: Do exports pay for imports? NO!

Whoops, my bad. I forgot that I actually eat US dollars.




That case of currency appreciation is the opposite of what we’re now facing.

But sticking with that case, you said at one point:

“But the US economy won't stay depressed relative to Canada's forever. D will eventually begin to fall. And that means the weighted (discounted at rate b) sum of expected future Ds must also begin to fall at some time. It is at that point that the exchange rate begins to depreciate again. And (if my head is straight on this), at some point before the exchange rate begins to depreciate, current D will be above permanent D, so transitory D will be positive. It's at that point the Bank of Canada should raise interest rates above the US Fed.”

Here’s my interpretation of that - I’d be interested if this makes sense to you.

Suppose the starting point is D = 0 and Canadian rates equal US rates.

Then D starts to move up in favor of Canada, so Canada needs to tighten policy somehow.

The spot exchange rate does all of the work initially.

Spot FX follows the change in current D based on the expectation that current D = permanent D. So no change in rates yet.

At some point current D and spot FX both overshoot – indirectly – in the sense that permanent D is suddenly revised back downward so that current D suddenly exceeds permanent D and spot FX has overshot the new permanent D. This is where transitory D is suddenly created.

So spot FX starts to depreciate.

But the forward FX markets also build in an expectation of a spot FX path that is consistent with converging to the new permanent D.

So the FX forward market starts to reflect an expected depreciation.

For example, the 1 year forward FX rate reflects an expected deprecation.

That expected depreciation forces up one year Canadian interest rates. That’s because US investors demand a higher one year interest rate in order to offset their cost of forward hedging back into US dollars.

The Bank of Canada then follows the lead of the term markets by a formal tightening of policy in the overnight rate.

If permanent D completely reverses, then the spot FX rate should return to where it was, the Bank should be able to unwind its interest rate tightening, and forward FX will be flat with spot.

Other variations depend on where permanent D settles in relative to its starting point and to starting interest rate levels. (I assumed D = 0 and interest rates were the same at the start for simplicity.)

Then it’s a matter of reversing all of this for the current situation of depreciation today.

JKH: I think that's right, but it's too hard, and too early in the morning, for me to be sure. Except, at this point, I think you need to say:

"If permanent D completely reverses, then the spot FX rate should return to where it was, the Bank should be able to unwind its interest rate [**differential**] tightening, and forward FX will be flat with spot."

I think that is what we are seeing now: the interest rate **differential** is going to be disappearing soon, but that could come from the Fed raising as much as the BoC lowering interest rates.

(I'm impressed you managed to get your head around that piece. It's one of the few times my own math has run ahead of my ability to understand it. I don't really understand my own model. Normally, with me, it's the other way around.)

"the Bank should be able to unwind its interest rate [**differential**] tightening"


I assumed the opening interest rate differential was zero for simplicity

So the unwinding of the tightening was the unwinding of the differential

Jacques René Giguère: "in the short run and financial sense, exports pay for imports"

No comprende. To me that means that imports also pay for exports, which means that imports equal exports, which is true globally, but seldom true for any specific country. Apparently you mean something else?

“But the US economy won't stay depressed relative to Canada's forever. D will eventually begin to fall....

“So the FX forward market starts to reflect an expected depreciation.”

Exchange rates don't work this way. The first part of this discussion is trying to implement some kind of dynamic program that Kydland and Prescott warned us about more than 30 years ago. Moreover, if someone could find a signal that told us when depreciation was going to occur, she should be running a hedge fund!

It is impossible to understand exchanges rates without thinking about asset pricing. In asset pricing you price off a stochastic discount factor which encodes the market's correction for risk - which is time varying. What we find is that when prices are high relative to payoffs, (e.g., high price-dividend ratios, high price-rent, etc.), expected returns are low. This is the time series predictability we see in stock returns and other asset markets (and yes, it is still efficient). Dividends don't forecast returns which implies that there is some return predictability in the system since the dividend-price ratio (slowly) mean reverts. The same effect appears in the FX market. If we see a currency with a higher than average interest rate spread relative to our domestic rate, an investor will enjoy higher returns, over the long run, by investing in the foreign currency. A naive calculation will say that you should expect depreciation (and there is a substantial risk of it), and so you should not expect to do any better over the long run. That is false. Over the long run, systematically holding currencies with higher than average interest rate spreads actually earns higher expected returns than holding domestic debt. Why? Because of risk premia. Everyone sees the possibility of a depreciation and being risk averse, they don't want to hold the currency, but anyone who does hold the risk will see a reward. This is entirely rational and efficient, not some overshoot, undershoot silliness. It is not enough to say that interest rate differentials suggests currency depreciation, etc. We must build an understanding of the stochastic discount factor. Remember that the forward rate is only the expectation under the risk neutral measure. In any asset pricing problem - which includes exchange rates -you can't get anywhere without explicitly thinking about risk premia.

Avon: "The first part of this discussion is trying to implement some kind of dynamic program that Kydland and Prescott warned us about more than 30 years ago."

I am assuming the central bank can make a credible commitment to target 2% inflation.

"Moreover, if someone could find a signal that told us when depreciation was going to occur, she should be running a hedge fund!"

My model assumes uncovered interest rate parity, so the assumption that nobody can make profits is already built into the model. In other words, if the model were true, you could not make profits from knowing the model. Standard RE stuff.

But yes, it's a simple model that ignores risk-premia.

Would US senators really complain? It seems to me the curency manipulator charge is put on whoevers policy they dont like at the moment. No bad word ever about Singapore or Switzerland for example.

As I had thought back in 2008/2009, the US and pretty well all other countries were trying to devalue their currency, racing to the bottom during the financial crisis, and that Canada was trying to get ahead of the US as this occurred. My thought was that it was pretty well impossible for Canada to overpower the US as they powered down, and so our currency rose against the US.

And now, rather suddenly we see the Canadian currency losing significant ground against the US dollar - so how does the BOC cause this to happen now? Conditions must have changed significantly lately if my thoughts have any validity.

Managing a 2% or whatever inflation target would be near impossible unless the BOC new the end point of the devaluation and also had support from the US government and/or the markets.

I still do not understand how we can run 5 years at or near par with the worlds reserve currency and survive fairly well overall, and then need to take fairly drastic action now. Assuming it is just the BOC taking this action

Determinant,Min: in the short run and financially, you must export to get the foreign exhange needed to pay for exports. In the long run, when you stop being autarkic, you no longer produce your own bananas but import them. Increased efficiency rise your apple production. The excess is exported.Same for the other country. What remains is to find the terms of trade that will distribute the surplus between the market players.

Jacques René Giguère: "in the short run and financially, you must export to get the foreign exhange needed to pay for exports"

So, by your definitions, if I buy Korean Won for US Dollars at the airport I am importing Won and exporting Dollars? And there is no such thing as a trade deficit or surplus?

Chun: As Nick mentioned, it makes sense for SG to target the exchange rate, because most consumption goods have to be imported. An FX appreciation target then has a more direct impact on consumer inflation than trying to manage aggregate demand via interest rates. The weakness of this system is that, with SG's commitment to an open capital account, any influence over the monetary supply and domestic interest rates has to be abrogated. Basically, Singapore's interest rates are determined by the Fed.

I'd also say the reason why figuring out exchange rate movements is so tricky is that not only are exchange rates really prices, but they are also relative prices. Saying "the Canadian Dollar depreciated against the US Dollar" is NOT the same thing as saying "the US Dollar appreciated against the Canadian Dollar". There's more than a semantic difference there, and the economic effects (and how to explain it to laymen) will differ based on which of the two is actually happening. Which, I think, is basically the point Nick is trying to get across.

Strikes me Nick Rowe’s point is just an example of a more general point which, roughly speaking is: “simple targets like inflation, unemployment, the exchange rate, are all defective in that one gets a fuller picture by looking at what lies behind an X% inflation rate or a Y% unemployment level, etc.”


The exchange rate of Singapore (and therefore indirectly your Malaysia) is aligned 60% to the Euro and 30% to China and 10% to the Fed, see


Applause on your blog, btw !


Don't know if you saw this, but Philip Cross had a piece in yesterday's Globe & Mail that takes (on my reading, at least) a very much "non X" approach to the dollar's recent fall, concluding that it is unambiguously bad. He argues in fact that "A falling exchange rate almost always lowers living standards". (Problem for upper-year open-economy macro students: "Identify an exogenous shock that might raise home-economy standard-of-living while producing a depreciation of the real exchange rate.")

Min: Yes you are.
Current account surplus and deficits are possible when you use money and set up a bank account where you stash your surplus and draw down your deficits. And the current account imbalance is possible because you are either lending (capital account deficit meaning you are not collecting your due in goods and services for now)or borrowing (you promise the pay later in goods and services). Of course, the original movement may come from the capital account: you wish to invest abroad and so send your goods there (that is you send currency which the foreigner use to buy your goods) or the foreigner invest here , gives you currency and you buy foreign goods with foreign currency.
Basic international trade and balance of payments stuff.

Jacques René Giguère: "Basic international trade and balance of payments stuff."

Thanks. :)

But there are different definitions out there.


"Who gains, and who loses, from the fall in the exchange rate? For Canada as a whole, is the fall in the exchange rate good news or bad news?"

From an asset / liability standpoint (capital account), people that have assets denominated in U. S. dollars and liabilities denominated in Canadian dollars win, people in the opposite situation lose.

Reporters often forget that for every current account deficit, there is often a capital account surplus.

But let's say there are no capital markets (no debt / no equity). This would mean that a government is literally printing / destroying money - spending it into existence and taxing it out of existence (presuming there is a "money authority").

A change in exchange rate would mean that either a change in the relative money policies of two governments has occurred or a change in the relative liquidity preference or productivity of the two populations has occurred.

Rather than go through all of the possibilities I will list two that may not be obvious. I will refer to two countries that begin with a balance of trade as country A and country B.

Scenario #1: Country A experiences a sudden increase in productivity - some smart young engineer from that country has invents the teleportation device (See Star Trek). This puts country A at a significant advantage in its ability to deliver goods / services at a lower cost, and absent money policy change, this will put upward pressure on the relative value of country A's currency relative to country B's. Over time this effect may wear off as the technology is assimilated, patents run out, and it is duplicated in both countries.

Does an increase in productivity by one country hurt another country? I would say no - just my opinion - rising tide lifts all boats.

Scenario #2: Country A experiences a sudden increase in liquidity preference (higher demand for liquid assets relative to illiquid goods) relative to another country. A change in relative liquidity preference will put downward pressure on Country A's currency relative to country B's.

Does a change in relative liquidity preference by country A hurt country B? Again I would say no - in the long run it hurts both countries equally because a demand for liquid assets by country A can be satisfied by either the currency of country A or country B (assuming no capital controls).

Current account surplus and deficits are possible when you use money and set up a bank account where you stash your surplus and draw down your deficits.

That was my point. All international transactions are in money. In point of fact, that international exchange between CAD and Won is actually a trade of CAD for USD and then USD for Won. The USD is the only currency that has widespread markets with every other currency, and that's why it's called the world's reserve currency.

There's no neutrality of money in international trade.


One more scenario:

Scenario #3: Country A experiences a sudden increase in both productivity and liquidity preference relative to country B. The effects on the exchange rate between country A and country B totally offset each other. Does this hurt either / both / neither countries?

Frank 08.56: The PSST is altered. Everybody should benefit in the end. But there might be unequal gains. Since people do not think in a solely Pareto way but also take into account relative gains, you can have opposition from the ones who gains least.

Nick: Even the Amano-van Norden model didn't claim much success at explaining weekly or daily movements. (Did we ever claim monthly R-squareds close to 50%? If we did, please shoot me.) So the point remains....no one who does the numbers thinks that they can explain the bulk of these short-term movements.


Thanks for the numbers (and the applause!). I'm not sure if the numbers are really correct though, because the SGD-CNY exchange rate is orders of magnitude more volatile than for the SGD against the EUR or the MYR or even the USD. A look at the charts shows SGD-CNY as almost a stationary (if volatile) series, while the EUR, MYR and USD charts look consistent with MAS stated policy instrument of using real exchange rate appreciation. Then again, maybe they're just pulling a fast one and anchoring on the CNY - but that leaves unexplained the suspiciously steady appreciation against the rest over a longer sample period.

I'd also say that there's little evidence that the SGD cross rates are affecting the Ringgit exchange rates either. I'd hazard to say the same factors are impinging on both currencies rather than MAS influencing Bank Negara, who have forsworn large scale forex intervention.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad