The world needs more currency wars
“Currency wars” are one of those topics about which almost everything you read is wrong.
1. The media often confuses changes in nominal exchanges rates (determined by monetary policy) with changes in real exchange rates (determined by national saving/investment policies).
2. The media often draws meaningless distinctions between policies explicitly aimed at reducing exchange rates, with other policies (such as QE and negative IOR) that have the effect of reducing exchange rates. For instance, the US dollar price of euros rose from $1.31 to 1.37 on the day QE1 was announced in March 2009. Is that currency manipulation?
3. The media tends to assume a zero sum game, but currency depreciation by one country will often boost stock markets all over the world, by boosting global AD.
4. The media has a “pro-international agreement” bias. International agreements made by highly educated public servants seem like responsible policymaking. If the media knows little about the issue, they’ll defer to the experts. And “war” sounds like a bad thing. But recall that these are the same experts who pressured the Japanese NOT to devalue in the 1990s, as they were sliding into deflation. How’d that advice work out?
In my view, things get even worse if central banks are perceived to have run out of ammo. (Put aside the question of whether they have actually run out of ammo, which seems impossible.) Suppose the Fed has room to raise and lower interest rates, but the BOJ and ECB have no room. Of course, they can still devalue if they wish. Now assume an international agreement to avoid “currency wars”. So they can’t devalue either. In that case, you’ve resurrected Bretton Woods, at the worst possible time. The US becomes central banker to the world. (David Beckworth has shown that the Fed is already a “monetary superpower”, and a currency agreement under these circumstances would make them even more of a superpower—controlling the fate of not just China, but also Japan and the eurozone.
I hope you can see the obvious problem here. The Fed insists that its mandate is to control American inflation, not Japanese and eurozone inflation. But if we have an agreement to ban currency wars, then the Fed has a moral obligation to stabilize the average inflation rate in the US, Japan, and Europe. I think it goes without saying that current policy is too tight, under that mandate.
And this blog post suggests that there already may be an implicit currency stabilization agreement:
More so, what if central banks think they’re at the limits of monetary policy anyway? If you combine that thought with pressure from the G20 you get to de facto agreement anyway, right? Or at least the behaviour of central banks can be explained, and predicted to an extent, by assuming the existence of a de facto agreement.
Here’s Citi’s Steven Englander saying just that:
It is probably better to think of G20 as similar to Friedman and Savages’s billiards player (link, page 12,13), G20 may not have made a deal but they are behaving as if they did, so we may as well analyze the consequences from that perspective. One question to ask is why they would agree to setting aside currency as a macroeconomic weapon the likely answer is that the winners of the currency war battles may have decided that they were not benefitting enough to offset the negative impact of the ancillary asset market volatility that emerged. Basically they were acknowledging policy ineffectiveness or at least monetary policy ineffectiveness, and the Statement pretty much admitted that.
So we now find ourselves in a situation where G3 has trouble cutting rates. The ECB and BoJ are reluctant because of the strains it is putting on their financial institutions and political unpopularity. The Fed is reluctant for similar reasons and because it seems unlikely that one cut would do great things for the US economy and it would certainly raise a suspicion that negative rates were beckoning. An already unpopular institution would become politically toxic. Bottom line, easing is hard to do. An ECB or BoJ hike is unlikely, needless to say. The Fed has already indicated its reluctance to hike and is very unlikely to hike to defend the currency if anything they seem to be cheering any weakness the USD encounters.
Put this all together and you have an extremely high bar for any G3 central bank cutting rates and an extremely high bar to them raising (and an even higher bar to any of them raising rates in response to currency weakness).
The point being that the FX market is now discounting the chances of aggressive central bank reaction in opposition to short term directional moves. Like, you know, they would if a deal had actually been agreed.
The new reality is one in which FX is being dictated by market forces rather than central banks. Or more so, we suppose, one in which some market participants can push exchange rates around without coming up against a central bank pushing back.
Of course, this could all go out the window when Japan pulls the trigger at its next monetary policy meeting on April 27th. Or they could opt for intervention (since mon pol might be exhausted, goes the theory while another theory says any intervention would have to be short and sharp considering the G20 once again) and surprise our inboxes. Inboxes that have already been surprised by this move in JPY, of course.
I would add that it is not “market forces” pushing the yen much higher; it’s a dramatically tighter monetary policy out of Japan. You might wonder, “How did that happen?” Expectations fairies—people have stopped believing that Kuroda will ease as needed to hit the BOJ’s inflation target. Even though the BOJ did nothing “concrete” in the past few weeks, the future expected path of BOJ policy has become much tighter.
Pay attention to the April 27 BOJ meeting, and then a few days later to the Q1 GDP for the US, which might be negative.
Tags:
12. April 2016 at 12:19
I never got why “curreny wars” are supposed to be a bad thing according to politicians and the media.
Ryan is out by the way. Kasich, Trump and Cruz are in:
“Let me be clear,” Mr. Ryan said. “I do not want nor will I accept the nomination of our party.” He added that he had a message for convention delegates: “If no candidate has the majority on the first ballot, I believe you should only turn to a person who has participated in the primary. Count me out.”
http://www.nytimes.com/2016/04/13/us/politics/paul-ryan-to-ruleout-run-for-president-aide-says.html
12. April 2016 at 12:56
The January meeting notes for the BOJ showed no discussion of increasing its asset purchases.
http://www.bloomberg.com/news/articles/2016-03-18/boj-minutes-show-no-talk-of-more-qqe-before-adopting-minus-rate
I’m assuming this undermined the BOJ’s credibility in hitting its inflation target and is implicit tightening.
12. April 2016 at 15:05
“3. The media tends to assume a zero sum game, but currency depreciation by one country will often boost stock markets all over the world, by boosting global AD.”
I know this is a general comment but just wanted to point out that if China were do significantly depreciate its currency it would be deflationary by boosting the dollar and euro. depends on how much it depreciates of course.
12. April 2016 at 15:58
Gordon, Good find.
James, You said:
“depends on how much it depreciates of course.”
Much more importantly, it depends HOW they did it.
12. April 2016 at 16:00
Excellent blogging.
There is something odd about central bankers. They appear to become physically and emotionally exhausted by conducting QE.
In contrast, central bankers appear to be exhilarated and invigorated by rate hikes and elimination of all monetary stimulus. Real zeal is shown, and iron will.
I wish I was joking, but read central bank literature and that of affiliated parties.
12. April 2016 at 16:20
@Scott,
Nice post. I agree with Benjamim Cole. Much better to replace central banks with an iPhone app which would do a much better job. Seriously though, I think the problems is that many central bankers don’t really understand the transmission mechanism.
12. April 2016 at 16:24
One other thing, a Central Bank depreciates it’s currency by buying foreign assets. If it does this with new money, it’s expansionary in aggregate.
12. April 2016 at 16:59
“The media tends to assume a zero sum game, but currency depreciation by one country will often boost stock markets all over the world, by boosting global AD.”
But boosting AD by inflation is a zero sum game. The zero sum game aspect of inflation is precisely why fiat money is backed by coercion and not consent. It is how there is even such a thing as “currency wars”. The disagreements among central bankers is over the redistribution effects of their respective inflations.
You have the causation direction backwards. Higher AD is not the cause of higher standards of living. Higher AD by inflation redistributes real wealth, since the inflationary system is coercive. Coercion always results in redistribution.
You’re stuck in the fallacious Keynesian box.
12. April 2016 at 18:42
Central banks aren’t out of ammo now, but in the long run they could run out, and that might end up hurting the economy even more.
If you had a 5% NGDP growth target, in the long run you would either need a 5% structural deficit to sustain it, or the government must needlessly interfere with the private sector by having an ever-growing “sovereign wealth fund”.
Without an ever-growing stock of government bonds or “debt”, in the long run the central bank is forced to accumulate private assets to maintain its targets.
12. April 2016 at 19:16
Sumner, I was wondering why you thought the term “global inflation” nonsense in your Econlog comments, but think “global NGDP” is not a nonsensical concept.
Christian List, I responded to you re: history of the Democratic Party. Take a look. NYC was more or less consistently Democratic since Jackson.
12. April 2016 at 19:25
@majorfreedom
But boosting AD by inflation is a zero sum game.
Correct. But the point of monetary policy is to un-stick real prices and wages so you get real some real AD growth in addition to inflation. If prices and wages weren’t sticky, any increase in the money supply (above real the real GDP growth rates) would (as you say) just be inflationary, and monetary policy would be both unnecessary and ineffective.
12. April 2016 at 20:16
Scott Sumner: Q1 USA real GDP negative? Seems like slow growth…
12. April 2016 at 20:44
Government bonds are just like plain old “money”, except that they shift inflation from the present into the future. The job of a central bank is to maintain price stability by smoothing out inflation between the present and the future; one of the main tools they have is to swap bonds for money and vice versa.
Central banks are running into trouble at the moment because, due to government deficits generally being too low, they have been gradually running out of future inflation to shift back into the present.
12. April 2016 at 20:48
In short: central banks cannot create inflation; they can only redistribute it between the future and the present.
13. April 2016 at 01:57
I don’t see why the currency wars are not a zero sum game. Boosting AD in the short run in terms of exports of one country will be an issue for the currency that gets stronger. Also a stronger currency will send a selling signal to investors pushing down stock prices.
13. April 2016 at 03:47
“Much better to replace central banks with an iPhone app which would do a much better job.”
Argentina would…. great stuff… http://www.bubblear.com/controversy-over-bcra-president-vanolis-future-dollar-contracts/
Argentine CB: can you show me a bid on the dollar
Trading Desk: 14.50 bid
Argentine CB: Sold at 10.65
Trading Desk: you mean you are offering at 14.65?
Argentine CB: No!! SOLD at 10.65!!! Where are you now???
Trading Desk: Uh, 14.00 bid???????
Argentine CB: Sold at 10.65!!
I can’t wait to hear who was the buyer of these contracts…
13. April 2016 at 04:44
Back when I worked on a trading desk, we had a word for someone we didn’t know who wanted to sell at below market prices. “Criminal”
13. April 2016 at 05:00
Dimitri, You said:
“If you had a 5% NGDP growth target, in the long run you would either need a 5% structural deficit to sustain it, or the government must needlessly interfere with the private sector by having an ever-growing “sovereign wealth fund”.”
Not at all. The entire demand for base money is barely over 5% of GDP, when NGDP trends upwards at 5%. A structural deficit of even 1% would be plenty. And of course the central bank could also buy alternative assets.
Nicholas, Your mistake is assuming that currency depreciation is equivalent to boosting GDP though exports. When the US sharply devalued in 1933, the trade surplus actually shrank, as the income effect overwhelmed the substitution effect. Zero sum people just focus on the substitution effect.
13. April 2016 at 06:56
Offtopic:
There is an interesting story here: California unions push for an exemption to $15 min. wage: http://www.theguardian.com/us-news/2016/apr/12/los-angeles-15-dollar-minimum-wage-unions
“Union leaders argue the amendment would give businesses and unions the freedom to negotiate better agreements, which might include lower wages but could make up the difference in other benefits such as healthcare. They argue that such exemptions might make businesses more open to unionization.”
13. April 2016 at 07:15
Off topic:
Q4 2005 to Q4 2015 PCE inflation annually compounded growth rate: 1.6%
In other words, the Fed was 25% under its target.
By simply achieving its target, the inflation part of NGDP would today be about 4% higher. Accelerated employment linked to this inlfation-NGDP boost (say 2.5M extra employees) would add about 2% to real GDP. In turn, this supplementary 2% in NGDP corresponds another 1.25M extra employees, and so on. In total, about 5M more people would be employed. That’s a total of 148M employees, which corresponds to a 62% employment-population ratio. In other words, full employment.
Inflation level targeting might actually have done the job.
13. April 2016 at 14:19
Scott,
It seems to me that you’re making a mistake by leaving government bonds out of the scenario.
Let’s say, for simplicity’s sake, that there is only a single kind of interest-bearing government liability, called “T-bills”, and everything else is ordinary non-interest-bearing government liability (“money”). The only way I can see 5% NGDP growth being a long-run equilibrium of sorts is if the government maintains the relative quantities of T-bills and money steady.
Consider, for example, an economy in which we posit two “steady-state” equilibria of government liabilities relative to GDP, both involving 5% NGDP growth:
(1) 2% of NGDP in the form of money, without any bonds.
(2) 1% of NGDP in the form of money, and 1% of NGDP in the form of six-month T-bills.
In a perfect steady-state scenario with no NGDP volatility risk it’s pretty clear that case (2) would make money and bonds essentially interchangeable, except that each $ of bonds would be sold for roughly $0.976 of money at issue, and that value would gradually increase over time, reaching $1.00 at maturity.
I posit that if the government chooses to “pay back” its debt and transition from (2) to (1), there will be *no* inflationary impact whatsoever. Equivalently, a central bank undertaking a “monetary expansion” and switching from (2) to (1) would have no impact at all.
13. April 2016 at 16:15
Anand, Thanks, I did a post.
LK, Interesting.
Dimitri, You said:
“The only way I can see 5% NGDP growth being a long-run equilibrium of sorts is if the government maintains the relative quantities of T-bills and money steady.”
Sure, in the ultra long run everything must rise roughly in proportion. But that doesn’t require a deficit of 5% of GDP, just a 5% annual increase in the national debt. If the national debt starts at 20% of GDP, as in a place like Australia, the required annual deficit is only 1% of GDP. The US could gradually let the debt ratio fall to 20% of GDP, and then rise from there.
I don’t understand the rest of your point, T-bills and base money aren’t even close to being perfect substitutes when interest rates are positive—that’s why it’s called high-powered money.
13. April 2016 at 17:41
Yes, starting at 20% debt/GDP would mean that a 1% deficit can maintain the relative proportions of financial assets, although both the 1% and 5% figure strike me as arbitrary—why does it even matter, as long as the long-term trajectory is stable?
You’re right to say that they’re not close to being perfect substitutes, and so a reasonable economic model should show *some* kind of impact from the switch.
At the same time, though, the plain QTM end of the spectrum strikes me as incorrect as well, because it also cannot be the case that removing the government bonds altogether would leave everything the same.
Consider an extreme case, in which “bonds” and “money” are just two different kinds of accounts a bank can have at a central bank, with the first paying interest and the second not paying interest. What happens now?
13. April 2016 at 19:28
OK, here’s some thoughts on a better theory of prices.
First, let’s start with this:
Sum over all current and future individuals of expected net present value of future consumption = expected net present value of all future final output.
IOW “every final product of the economy is consumed by one or more people”
Now, let’s say we have a fixed quantity of “money” that never changes. Conceptually, it makes sense to say that starting off with 1/10th of the money corresponds to owning 1/10th of all future production of the whole economy.
Obviously, you can also pay for / be paid for doing productive work, which increases the RHS of the equation and partly compensates you for that increase.
So, let’s try something similar, but with a variable amount of money. Thus, if you hold on to your money, the relative value of your money to the value of the entire economy declines over time.
This is a rough guess, but I think it might engender some insight.
For example, if we assume that bonds don’t exist and you are neutral to output over time, and government can only tax you via inflation, the net present value of of $1 is equal to
E[Integral from t=0 to t=inf of {GDP at time t / amt of $ at time t}]
On the other hand, the net present value of a $1 1-year T-bill is equal to
E[Integral from t=1 to t=inf of {GDP at time t / amt of $ at time t}]
Note that the model would be different if we took into account the government’s powers to tax and spend and how exactly they impact on you.
13. April 2016 at 19:44
Also, I think a model like the above gives a coherent idea of what exactly the “expectations channel” could mean: the value of current fiscal assets depends on your beliefs about future government policy.
It seems to me that this means that, relative to the bonds not being there, bonds are inflationary unless you believe they will be offset by taxation (i.e. an alternative way to reduce your purchasing power).
On the other hand, if the bonds will be rolled over forever, or will be replaced by “printed money”, then they are inflationary – but not particularly so.
14. April 2016 at 00:51
[…] but Jefferies do think the Japanese authorities are in a corner, painted in by a strengthening yen, tighter monetary conditions and a drop in inflation […]
14. April 2016 at 00:56
[…] but Jefferies do think the Japanese authorities are in a corner, painted in by a strengthening yen, tighter monetary conditions and a drop in inflation […]
14. April 2016 at 04:45
“T-bills and base money aren’t even close to being perfect substitutes when interest rates are positive”
By your definition, you can can do yourself a favor and stop hedging yourself by saying “when rates are positive”, at zero they are no better (if not worse – due to convexity issues). You are confusing yourself by thinking only in terms of initiation to expry p+l.
14. April 2016 at 04:47
So, here’s a quick summary thesis:
The long-term trajectory of inflation is set by *fiscal*, not *monetary* policy. On the other hand, monetary policy has the power to shift inflation between the present and the future.
Thus the effect of QE is that it can shift inflation from the future into the present, thus increasing current inflation at the cost of future inflation.
Part of the reason for the current situation with the “zero lower bound” is that there isn’t enough future inflation to go around, because governments have been overly committed to being “responsible”, and have been attempting to obtain long-term balanced budgets rather than aiming for structural deficits.
Europe’s case is particularly bad, because the Eurozone is effectively committed to having a long-term balanced budget. The problem is that *no one* in the Eurozone actually has the political authority to commit to long-term inflation.
14. April 2016 at 05:11
Dimitri, You said:
“Consider an extreme case, in which “bonds” and “money” are just two different kinds of accounts a bank can have at a central bank, with the first paying interest and the second not paying interest. What happens now?”
That’s not an extreme case, it’s close to reality. Consider vault cash and interest bearing reserves. In that case, it’s the total quantity of both assets that matter.
On your second point, yes, the existence of bonds is slightly inflationary. But marginal changes in the stock of bonds have an only trivial effect on the price level, which can be easily offset by the Fed. Thus when Reagan started running massive deficits in the early 1980s, inflation plunged because it took just a very slight tightening by the Fed to offset the inflationary impact of the growth in bonds. The Fed didn’t even have to reduce the money supply, just slow the rate of increase.
The US case shows that the monetary authority sets the inflation rate. It was the Fed that decided to reduce inflation from 8% to 2%, not Congress. Can you even imagine Congress trying to target inflation at 2%, the mind boggles.
14. April 2016 at 05:36
Scott, You said:
“That’s not an extreme case, it’s close to reality. Consider vault cash and interest bearing reserves. In that case, it’s the total quantity of both assets that matter.”
OK, good. That’s a start. But it seems ludicrous to me to think that there would be some kind of discontinuity at which there is a switch from “both assets matter” to “only one matters”, even when we move away from the zero lower bound.
I would argue that they *always* both matter, but there are differences between them that cause them to have slightly different effects. Here’s two key ones:
(1) T-bills have higher “frictional costs” than money and this dampens demand somewhat.
(2) Because they pay interest, T-bills serve to discourage some kinds of investments, and increase the extent to which people “sit on” their money.
However, (1) and (2) are still much less powerful than the core fact that both money and T-bills have (either directly or indirectly) the power to *buy stuff*. So, unless the interest rates on T-bills are unusually high, the net quantity of money and T-bills ought to be of greater importance than differences between them.
Scott, You said:
“On your second point, yes, the existence of bonds is slightly inflationary. But marginal changes in the stock of bonds have an only trivial effect on the price level, which can be easily offset by the Fed. Thus when Reagan started running massive deficits in the early 1980s, inflation plunged because it took just a very slight tightening by the Fed to offset the inflationary impact of the growth in bonds. The Fed didn’t even have to reduce the money supply, just slow the rate of increase.”
The problem here is that you’re neglecting the effects of the expectations channel, and thus your explanation is flawed. I would be completely justified in arguing that Reagan’s deficits were not particularly inflationary because of the expectation that they would be moderated in the future, perhaps even with surpluses.
Scott, You said:
“The US case shows that the monetary authority sets the inflation rate. It was the Fed that decided to reduce inflation from 8% to 2%, not Congress. Can you even imagine Congress trying to target inflation at 2%, the mind boggles.”
In fact, inflation rates are set by *both* Congress and the Fed. Yes, the Fed can set the inflation rate in the short run, but the long-run case must necessarily depend on fiscal policy and expectations about future fiscal policy.
My proposal isn’t for Congress to keep inflation at 2%, it’s that they commit to a sufficient structural deficit to keep it at 2% in the *long run* and leave short-run management to the Fed.
14. April 2016 at 05:44
Oh, also: The Reagan deficits weren’t all that big as a percentage of GDP. If anything, those deficits were overcompensated by Clinton, and this overcompensation has been reversed by Obama’s much larger deficits.
However, under Obama, the long-term deficit has gone down *too much*, and hence the central bank has lost a lot of its ability to pull future inflation into the present.
14. April 2016 at 07:40
Actually, here’s a really simple way to state my proposed theory of the price level:
price ~ (expected future quantity of “money”) / (NPV of expected future economic activity denominated in that “money”).
This is also a much better explanation for hyperinflation; hyperinflation doesn’t happen because of the numerator going to infinity, it happens because the denominator goes to zero.
16. April 2016 at 07:04
Dimitri, You said:
“But it seems ludicrous to me to think that there would be some kind of discontinuity at which there is a switch from “both assets matter” to “only one matters”, even when we move away from the zero lower bound.”
It may seem ludicrous, but there’s a massive money demand literature that says it’s true. Personally, I think it’s ludicrous to imagine people substituting T-bills for cash when shopping at Walmart.
You said:
“The problem here is that you’re neglecting the effects of the expectations channel, and thus your explanation is flawed. I would be completely justified in arguing that Reagan’s deficits were not particularly inflationary because of the expectation that they would be moderated in the future, perhaps even with surpluses.”
Since that’s an irrefutable theory, I won’t comment.
You said:
“In fact, inflation rates are set by *both* Congress and the Fed.”
No, Congress have never even tried to target inflation, it’s 100% the Fed (in the US, obviously, not Zimbabwe)
You said:
“The Reagan deficits weren’t all that big as a percentage of GDP. If anything, those deficits were overcompensated by Clinton, and this overcompensation has been reversed by Obama’s much larger deficits.”
This is just handwaving. I don’t buy the theory, and there is zero evidence, as even you seem to admit. So I have no interest in any sort of fiscal theory of the price level. No one in the 1980s had any expectation that Clinton was going to run surpluses in 1999-2001, and so it had no impact on the dramatic slowdown in inflation, as deficits grew much larger.
You are wasting your time on fiscal policy, monetary policy tells us all we need to know to understand the price level.
16. April 2016 at 14:00
Scott, you said:
“This is just handwaving. I don’t buy the theory, and there is zero evidence, as even you seem to admit.”
See, I’d be happy to make this a discussion about evidence, and do away with handwaving. I would very much like to see *your* evidence for the claims you’re making.
For example, your claim about QE is that it doesn’t work because people think it isn’t permanent, but where is your evidence to support that? That claim is equally irrefutable, and (if anything) If anything, it seems to me that massive central bank balance sheets are the new normal, and QE-infinity is here to stay.
The thing is, your claims about monetary policy are deeply flawed in the absence of an actual mechanism by which they’re supposed to work. The “expectations channel” simply isn’t good enough unless there is a concrete underlying mechanism for why people ought to expect what you say they expect.
Scott, you said:
“Personally, I think it’s ludicrous to imagine people substituting T-bills for cash when shopping at Walmart.”
Yes, people can’t pay for things at Walmart by using T-bills. They can, however, pay for things at Walmart out of a bank account that receives interest which is in part (when taken as part of the economy as a whole) supplied by interest paid by the Treasury on the stock of T-bills. Thus, indirectly, people can and do pay for things at Walmart by using T-bills. As you yourself have noted, bank reserves that pay interest are not really much different to T-bills.
Now, there are of course short-term effects due to frictions and the “portfolio balance channel” that Bernanke likes to talk about, but fundamentally I don’t think you’re correct. It may even be true that your basic “hot potato” theory used to be true before the days of ubiquitous (and much-closer-to-frictionless) electronic money, but the fact is that those are no longer the times we live in.
Besides, even if it’s true that such frictions remain a “big deal”, they can only have a *short-term* effect. In the long run, arbitrage between bonds and money still leads to them being essentially interchangeable.
Scott, you said:
“No, Congress have never even tried to target inflation, it’s 100% the Fed (in the US, obviously, not Zimbabwe)”
I’m not saying Congress directly responds to inflation. They do, however, respond to the interest rates that they expect to pay on the debt in the long run, and they do this by changing fiscal policy. Thus the long-term effects of monetary policy work only insofar as they are channeled through long-term fiscal policy.