Scott Sumner  

Zachary David on NGDP futures targeting

PRINT
Observations from Europe... Slavery in the Age of Em...

Jim S. directed me to a Zachary David post criticizing my NGDP futures targeting proposal. He links to my Mercatus paper on the plan, but seems to ignore its contents. Most of the objections that he raised are answered in the Mercatus paper. He may not agree with my arguments, but he should have at least acknowledged them.

For instance, he suggests that there might be little interest in trading NGDP futures contracts, but doesn't tell us why that would be a problem. (It isn't.) He wonders why NGDP futures would be such a good idea, given that the private sector hasn't already created such a market. Perhaps that's because the private sector is not legally allowed to do monetary policy.

Breaking down the mechanics you get:

1. You believe NGDP will be lower than the Fed's posted market of x% at expiration
2. So you sell a futures contract to express this view
3. The Fed responds by increasing the monetary base
4. Thus raising expected NGDP against your view

So... by selling that futures contract you have just raised the expected settlement value against yourself. It's like if selling a futures contract in copper triggered a mechanism which automatically lowered the existing quantity of copper. Expressing your view immediately causes your view to be less right. I'm not suggesting that people don't respond to prices all the time -- e.g. a farmer seeing a high price for future grains and subsequently planting more of them. But in general, a well-functioning market shouldn't have an infinitely liquid counter-party triggering automatic causal effects against every position.[2]


Notice that he suggests a way in which this market would differ from other markets, and then explains that his reasoning was actually incorrect---futures purchases in other markets do trigger supply responses---and then ends up telling us how he thinks a "well-functioning market" should behave, without providing any justification.

I think such a mechanism not only defeats the purpose of price discovery, but in Sumner's market it might lead to other hilariously manipulative scenarios. Consider: because the Fed is pegging the price of the futures contract and buying/selling infinite quantities, a smart player could load up on assets whose prices would be affected by a large expansion of the monetary base, then sell a tremendous amount of NGDP futures to trigger open market purchases, then sell the aforementioned assets and subsequently repurchase the NGDP futures at the same price for a tidy profit.
Note that the purpose of my proposal is not "price discovery" any more than a $35/oz. peg of the dollar to gold is aimed at "price discovery" of gold prices. Taking the gold standard analogy one step further, one could argue that a "smart player" could have loaded up on gold during the 1920s, and simultaneously sold short an asset that would be likely to decline in price under a contractionary monetary policy. And then the manipulator could have sold the gold back to the central bank at the same price. Does this seem too good to be true? That's probably because it is. Everyone else would have the same opportunity, so it's very unclear at whose expense this manipulator will profit.

Under NGDP futures targeting, manipulation would be even more difficult. If you sold the asset back immediately, it would not significantly impact monetary policy. Now consider what would happen if you held the NGDP futures until maturity. If you bought NGDP futures to trigger a contractionary monetary policy, and it succeeded, you would lose money on the NGDP futures (unlike the previous gold example where you break even on gold.) In that case, you need even bigger profits on your side bets. The problem is obvious---other "manipulators" will be taking the opposite strategy. If they respond to you by selling NGDP futures, moving monetary policy closer to the correct level, they will breakeven on the NGDP futures and profit (at your expense) on their side bets in other markets.

In any case, his criticism doesn't even apply to Bill Woolsey's version of the policy, one of the four versions I discussed in the Mercatus paper, and the version I have more recently concluded is the best. Under this plan the Fed simply makes dollars convertible into NGDP futures at a fixed price, without any automatic link between NGDP futures purchases and the base. It's analogous to a gold exchange standard where the central bank is not required to follow any sort of "rules of the game". Perhaps David didn't read that far.

There's also a long discussion of NGDP data revisions. David worries that this might reduce interest in an NGDP futures market, as investors wouldn't have confidence that the initial estimate would hold up over time. This is one reason I like the Woolsey version of the plan, it's easier to head off this sort of criticism (which I would argue are not a problem in any of the 4 versions). Under what Woolsey calls "index futures convertibility", it makes no difference if no one buys or sells NGDP futures, just as under a gold standard it makes no different whether anyone takes the central bank up on the offer to exchange unlimited amounts of dollars for gold at $35/oz.

Think of NGDP futures targeting as highway guardrails---does it matter if cars never brush up against them? No, and it also doesn't matter if there is no trading of NGDP futures contracts.

PS. Note that the policy discussion in this post is entirely different from my proposal that the Fed set up and subsidize trading in NGDP futures for research purposes. In that case the price is not fixed, and it matters very much if there is little or no trading volume.

PPS. This made me smile:


As a result, I predict there would be little to no volume in any NGDP futures contract.[3]

In fact, I set up a futures market for one-year NGDP contracts at Hypermind with just $5000 in subsidy, and there was active trading, albeit a modest volume. Imagine the volume of trading if the prize money had been $50,000. Then imagine $500,000. Then imagine $5 million. Then imagine $50 million. Then imagine $500 million. Then think about the fact that even $500 million is chump change to the Fed.


Comments and Sharing






COMMENTS (11 to date)
Silas Barta writes:

What about manipulation of NGDP itself? Couldn't you make NGDP arbitrarily high by having two untaxed institutions do sham purchases between each other?

James writes:

"He wonders why NGDP futures would be such a good idea, given that the private sector hasn't already created such a market. Perhaps that's because the private sector is not legally allowed to do monetary policy."

That (the private sector not being allowed to conduct monetary policy) is a poor explanation. It did not stop the creation of any of the existing futures contracts.

I wish Scott would give specific predictions of the consequences of NGDP targeting. The case for NGDP targeting would be much stronger if there were some reason to believe that the benefits would be quantitatively large relative to other types of central bank mandates.

Scott Sumner writes:

Silas, Purchases don't add to GDP, unless they reflect current year output.

James, I think you are missing the point. Those other futures markets filled a demand for speculation and hedging services. The purpose of NGDP futures would be entirely different, it would be aimed at assisting monetary policy, not providing hedging and speculation services. What possible incentive would a private firm have to assist monetary policy? How do they gain from that?

You asked:

"I wish Scott would give specific predictions of the consequences of NGDP targeting. The case for NGDP targeting would be much stronger if there were some reason to believe that the benefits would be quantitatively large relative to other types of central bank mandates."

I have a Mercatus paper on the gains from NGDP targeting, and another paper on the gains from NGDP futures targeting. Those papers answer your question.

Silas Barta writes:
Silas, Purchases don't add to GDP, unless they reflect current year output.

I'm not following -- how would that scheme not get counted in GDP?

James writes:

Scott,

I read those papers in the past and don't recall finding any specific (and falsifiable) claim such as "Under NGDP futures targeting, the average unemployment rate would decrease by one fifth relative to its average in the 50 year period before NGDP targeting." If there is a passage I have missed you could easily cite the page and paragraph.

"The purpose of NGDP futures would be entirely different, it would be aimed at assisting monetary policy, not providing hedging and speculation services."

That is your preferred purpose for an NGDP futures contract but a private firm could create such a contract for any reason. The fact remains that no private firm has ever created such a futures contract for any purpose.

Silas,

If you and I sell a baseball card back and forth between ourselves that's not GDP, at least in theory. Realistically, there are probably ways to do this and call it GDP. If we call ourselves card dealers, the purchses would be accumulation of inventory and therefore part of the I in the expenditures formula for GDP.

CA writes:

Scott, Z. David left a pretty snarky update responding to this post in his original post.

Also, Noah Smith has weighed in and, not surprisingly, he disagrees with you.
http://noahpinionblog.blogspot.com/2016/07/criticisms-of-ngdp-futures-targeting.html

Charlie writes:

I feel Zack and Noah's pain. I think our generation has just lost the ability to think about commodity standards.

Have you every thought about explaining the proposal through NGDP linked bonds? Like TIPS, the Fed would issue NGDP linked bonds.

You'll probably recall Cochrane did something similar here with TIPS (or CPI futures), I think shortly after you guys "debated."

http://faculty.chicagobooth.edu/john.cochrane/research/papers/big_stick.html

I just find it easier to wrap my head around. Maybe you've already done this somewhere.

rayward writes:

Inherent in the criticism of Sumner's idea for NGDPLT is suspicion of futures markets; indeed, suspicion of finance, suspicion of markets. Futures markets go back to the rice futures market in Japan, an enormous advancement in finance that significantly mitigated swings in rice supply and price. But I look at Sumner's idea not so much as to whether it would be a perfect model for monetary policy, but by comparison to the alternatives, including the models currently used by the Fed, which have been anything but perfect.

Silas Barta writes:

@Scott_Sumner:

If you and I sell a baseball card back and forth between ourselves that's not GDP, at least in theory. Realistically, there are probably ways to do this and call it GDP. If we call ourselves card dealers, the purchses would be accumulation of inventory and therefore part of the I in the expenditures formula for GDP

Well, that still wouldn't matter for purposes of the very NGDP calculation used to base policy on, unless we *specifically* registered as dealers and chose to classify that card as inventory. If we were just two consumers buying things from freelancers, they both count toward consumption.

Second, even if the NGDP targeting world somehow solved this problem, it wouldn't have a way of doing it for services. If Alice's Non-Profit buys $X of A/C repair services from Bob's Non-Profit, and vice versa, that shows up as the purchase of $2X worth of services. But, of course, they can do those transactions, and log them, without any service (A/C or otherwise) actually happening -- say, if they wanted NGDP figures to be higher than they currently are.

All of the above suggests that NGDP is just as vulnerable to Goodhart's Law as every other panacea before it: yeah, it *correlated* with the thing we really want to optimize, but that breaks as soon as you try to optimize it specifically.

James writes:

Silas,

It was I, not Scott, who responded to you and yes, people could "make" NGDP if they really wanted to. When it comes to services, your point makes even more sense. Two federal agencies (the Fed and the SEC, perhaps) could do paid consulting for eachother at equal and opposite dollar amounts to get NGDP to any target by fiscal means.

I could not agree more with your broader point about Goodhart's law. Changing central bank mandates is likely to be about as useful as Soviet glass factories changing production targets from kilograms of glass to square meters of glass when the goal is useful windowpanes. Some people do not want to see this.

Silas Barta writes:

@James: Thanks for the correction and clarifying the point of agreement.

Comments for this entry have been closed
Return to top