Scott Sumner  

Flying blind

There's No Such Thing as a Fre... The Ultimate Incivility...

There is a sort of zero sum aspect to exchange rates. Any change in exchange rates means one currency is weaker and one is stronger, relative to the other. But exchange rates tell us nothing about whether currencies are weaker or stronger in any useful absolute sense. Think about that fact when reading the following:

Expectations that the Federal Reserve is on course to start tightening policy has spurred fears of a return of last year's emerging market turmoil, but Societe Generale (Euronext Paris: GLE-FR) tips a strong dollar as a bigger risk.

"A strong dollar tantrum could be a more worrying scenario than a Fed tightening tantrum," Michala Marcussen, global head of economics at Societe Generale, said in a note dated Sunday.

The U.S. dollar index (New York Board of Trade (Futures): =USD) has climbed around 7 percent this year, with the Fed now nearly completing the tapering of its asset purchases, with markets widely expecting interest rate increases to begin sometime next year.

Some analysts are concerned this will spur a repeat of the "taper tantrum," when concerns about the Fed's move to begin tapering caused a brutal selloff in emerging market assets earlier this year and last year.

"Hope today is that a strong dollar will cap U.S. inflation, delay Fed tightening and boost exports to the U.S.," Marcussen noted, but she believes for that to happen, the U.S. dollar would need to strengthen so much that it would signal much weaker growth in the rest of the world.

To delay Fed rate hikes, the euro (Unknown:EURBA=) would need to fall to $1.10, while the U.S. dollar would need to fetch around 120 yen (Exchange:JPYUSD=) and 6.50 yuan , she said. Early Tuesday, the euro was around $1.2690 and the dollar was fetching 109.40 yen and 6.1495 yuan.

"In such a scenario, [a strong] dollar would equate to further capital outflows, placing further pressure on already vulnerable economies," she said. "A 'dollar tantrum' scenario could well prove more painful than a 'Fed tightening tantrum,' assuming the latter comes with better growth in the rest of the world."

This is a messy and very confusing discussion, but it's unfair to simply blame the reporter. We are all hamstrung by a lack of relevant real time data. Look at the last few paragraphs. This indirectly gets at the fact that a strong dollar due to monetary stimulus in other countries is bullish for the world economy, whereas a strong dollar due to tight money in the US is bearish for the world economy. Unfortunately, the dollar's exchange rate by itself doesn't tell us which one is occurring. Hence the discussion is muddled. We need data on whether the dollar is getting stronger in some sort of absolute sense, not relative to other currencies.

We can get some extra information by looking at other asset prices; TIPS spreads, stock prices, commodity prices, etc. But those are also affected by other factors. Thus if China shifts from heavy industry to services, it will reduce global commodity prices and also reduce US TIPS spreads. But it's not necessarily contractionary for the US. It need not reflect slower expected NGDP growth in the US.

What we really need is a real time indicator of the stance of US monetary policy, a deep and liquid NGDP futures market subsidized by the US Treasury or Fed. It's a scandal that we don't have that market, and thus I am currently in discussion with some private individuals who are trying to set up such a market. I hope to be able to announce something soon. But these markets are incredibly cheap to set up, so economists really ought to be asking the US government why this market hasn't already been created. The government could set up a much better market than I can.

Unfortunately, economists seem remarkably uninterested in real time data on market NGDP expectations. They seem happy to keep on discussing ambiguous indicators like "the dollar", and then pointing out that dollar moves up and down could mean one thing or another, depending on other factors. So they put no pressure on the government to provide this sort of useful real time data on the most important cyclical indicator in the world---US NGDP growth expectations. We are flying blind.

PS. Does this have anything to do with "liberty," the theme of this blog? Consider that unexpected sharp declines in NGDP growth expectations, such as 2008-09, lead to all sorts of mischievous government policies that make us less free.

Comments and Sharing

COMMENTS (12 to date)
BC writes:

"But these markets are incredibly cheap to set up..."

Do you mean the exchange is easy to set up or that it is also easy to recruit market makers that will ensure meaningful liquidity? I thought that it might be challenging to find enough people to participate --- liquidity attracts more liquidity, etc. I'm glad to hear that you are trying to set something up though.

Another reason NGDP markets might be useful beyond providing information is that, precisely because NGDP impacts economic cycles, NGDP futures may allow people to hedge "macro cycle" risk.

"Unfortunately, economists seem remarkably uninterested in real time data on market NGDP expectations."

I am still surprised by how many discussions about future inflation don't even acknowledge that market inflation expectations are directly observable. There are many people that seem stuck in a pre-1997 (pre-TIPS) mindset.

Hunter writes:

Ugh, Just bring back the gold standard already.

Andrew_FL writes:

Cap inflation and delay Fed tightening? Does this sentence make no sense or is it just me?


economists really ought to be asking the US government why this market hasn't already been created. The government could set up a much better market than I can.

This seems like a decidedly unlibertarian or unconservative judgement-whichever you prefer. The government could do something better than individuals? Isn't this contrary to the entire set of principles on which the idea that such a market would be a good idea is founded?

Shouldn't the question that needs to be asked, be asked by you, to yourself:

"If NGDP futures markets are such a good idea, why don't they already exist?"

If they need to be subsidized to exist it's because it's not a good (ie profitable) idea for anyone.

Unless, you would tell me that there are Government created obstacles to this kind of thing?

Gordon writes:

Scott, I've noticed that when you write about NGDP level targeting here at Econlog, you discuss it in the same way that you might discuss it over at your normal blog. The problem is the audience over at your normal blog has already been through all the arguments and rationale as to why an NGDP level target serves as a better target for the U.S. rather than other macroeconomic variables. Here at Econlog, the audience is unfamiliar with your rationale for this so it generates a great deal of push back. Maybe a post here about why you came to believe that an NGDP level target would be the best approach would be useful.

Also, given the political view of much of the audience here, you may want to go into a bit of history of how when NGDP expectations drop, the political response of the U.S. government is to curtail freedom. Just try to keep in mind that much of the audience here are very new to your views so you need to explain how you came about them. Afterwards, they may agree with those views or they may disagree with them. But it's important for the discussion that they understand the basis for those views.

ThomasH writes:

I do not know the quote but I think Milton Friedman once said something like that the main cost of inflation was the many misguided ways of trying to suppress it (like fixing exchange rates and other prices).

Scott Sumner writes:

BC, You attract traders with trading subsidies.

I don't believe many people are interested in hedging macro risk, otherwise the market would already exist.

Andrew, NGDP market expectations are a near perfect example of a pure public good. That's probably why the private sector hasn't done it. By the way, I am not calling for the government to "do more", I want them to "do less and do different." Spend millions of dollars less on predicting NGDP with government employees and spend $100,000s more on identifying the market expectation of NGDP growth. That's clearly a move in the libertarian direction.

Keep in mind that whether something "sounds like" a libertarian idea has no necessary relationship to whether something is a libertarian idea.

Gordon, Thanks, that's good advice.

Thomas, Good point.

Brian Donohue writes:


Since rates bottomed out on August 28, CPI expectations have come down about 0.15%, and real Treasury (TIPs) yields have jumped more than 0.30%. I interpret this as the market sensing tightening.


I don't presume to speak for Scott, but I found this 1968 paper from Milton Friedman to be a useful way of understanding Scott's vision:

Andrew_FL writes:

For what it's worth, I did think you would say it was a public good a while after I said that. Although that does raise the question whether what you are attempting to create privately is futile or not.

Anyway I'm sort of a laissez faire absolutist so how it sounds is important to me. I like what that which you want is trying to achieve (stability of the trajectory of nominal income) I'm just skeptical about the method itself.

Well, that, and I think you still favor too much inflation in the long run. But then again I'm more in line with Selgin's view (productivity norm) which would be deflationary in the long run. Even price level stabilizationists are too inflationary in the long run for my taste.

Ken from Ohio writes:

I am really interested in this NGDP futures market.

I recently went back and listened to all of Prof Sumner's discussions on the EconTalk archives.

He discussed this NGDP futures market with Russ Roberts - but only briefly.

I got the impression that the plan would be that the Federal Reserve would sell an asset linked to NGDP measurement. The value of that asset would then rise and fall with actual NGDP.

But I am not really sure how it would work. I am interested in the overall concept and how it could be implemented

Yancey Ward writes:
What we really need is a real time indicator of the stance of US monetary policy, a deep and liquid NGDP futures market subsidized by the US Treasury or Fed.

I bust out laughing every single time I read this.

jj writes:

When the NGDP futures market is up, I volunteer to put in some "dumb money". Actually it's going to be as smart as I can make it, but if I lose it all I'll tell myself "at least I subsidized a really useful market".

Please just don't repeat the various bitcoin market disasters. I don't want to subsidize that.

Gordon writes:

Brian, thanks for the link to the Friedman paper. I love the comparison of money to a machine that is so pervasive that when it gets out of order, it throws all sorts of other machinery out of order. Friedman may have chosen the wrong nominal variable to maintain monetary stability but that takes nothing away from the rest of his analysis.

Comments for this entry have been closed
Return to top