Why Scott Sumner’s NGDP Level Futures Will Give Goldman Sachs $500bn in One Day, Part I

I never wanted to write this post.

I might not be the world’s biggest fan of monetary policy, but I like NGDP level targets. A NGDP level target would be superior to our current dual mandate structure.

But NGDP level futures contracts are a very bad, even terrible way to moderate the level of NGDP. I thought the fascination with NGDP level futures would have played out by now, without me doing a full on smackdown. But that hasn’t happened.

Instead, we get Noah Smith arguing trading is volatile. It’s not a great reason to avoid NGDP futures. It’s a bit thin.

So I am compelled to write a post I did not want to write.

NGDP level futures would almost certainly hand Goldman Sachs and hedge funds a payday worth over $500bn, while giving almost nothing else to the rest of the economy. Either that, or NGDP level futures would never be traded by anyone.

There are no other outcomes for NGDP level futures. It’s between some dude pulling down a multi-billion dollar bonus, or nobody trades them. There is no in between.

NGDP level futures are such a bad idea I can’t even stand to hear about them – they are offensive to everything I know about how futures markets work. I’ll show NGDP level futures have a host of extremely serious problems, and even worse, one of these serious problems cannot be overcome by any possible futures contract design.

Then, I’ll show the only reasonable way to structure an NGDP contract to have the proper economic impact. Only, it won’t be a futures contract, and people won’t trade it. It won’t be –can’t be – anything like what Scott Sumner has in mind.

What are NGDP level futures contracts?

Mayor Bill Woosley explains:

“Sumner proposes having the central bank buy and sell the futures contract at a target price.   There would be zero volatility in the price of the futures contract.

I believe that Sumner has in mind a system where the target price increases at a 5% annual rate.  However, it would also be possible to define an index by dividing the actual value of nominal GDP by the target, which would mean that the price of the future contract would never change.

Of course, what this means is that the central bank always takes a position opposite to the net position of market traders.   If there are more bulls than bears on the market, the Fed must be a bear too.   If there are more bears than bulls, the Fed must be a bull.     Smith’s argument would then be that because stock prices vary more than actual dividends, the Fed’s position on the futures contracts would vary more than actual fluctuations of nominal GDP.   Since nominal GDP varies so much, then the Fed’s position on the contracts would vary a tremendous amount.”

Let’s draw out the major components of this idea:

  1. The U.S. Federal Reserve is the market maker.
  2. The price of the futures contract never changes due to the Fed buying and selling contracts at a set price as the market maker.
  3. The contracts settle to the NGDP level as given by the government, presumably on the release date of the NGDP number by the BEA
  4. There is unlimited liquidity in the contract.
  5. The idea is to put more money into the economy when the economy is very bad, and less into the economy when the economy is only mildly bad. The worse the economy, the more money pumped into the economy (and vice versa)
  6. Open interest and its direction would be used as an indication to what the fed needs to do for the economy.

I haven’t read every Scott Sumner post on NGDP level futures, but I’ve read a bunch. Bill W is giving a fair, if truncated version, of NGDP level futures.

Noah Smith gives us an update from Scott Sumner, and this description is very fair to Scott’s version of the NGDP level futures.  Here’s the nutshell response from Noah:

In an email exchange, Scott Sumner has clarified the nature of his NGDP futures market proposal. In a nutshell, he proposes that the Fed act as a market maker, buying and selling infinite quantities of NGDP futures at the target price. Demand for NGDP futures would then be used to determine Fed policy; if NGDP futures demand increased, the Fed would commence open-market operations to bring down expected NGDP. The price of NGDP futures would not move, but demand would swing from positive to negative, moving Fed policy as it swung.” (bold mine)

See? I was very fair.

Why you should take this anti-NGDP level futures screed very seriously

I know how to design futures contracts, and why they work.

I designed and created futures contracts for several years. It’s called product development within the futures industry. I have a patent application on a futures contract design. The patent application number is 20090210336 and you can find the text of the application here.

My patent application contains a pretty cool idea. It stipulates 2 entirely separate daily cash flows for a single futures contract. All current futures contracts (that I know) have only 1 cash flow. Two cash flows in a contract is a useful idea because it allows for transparent contract pricing separate from financing costs. On all existing futures contract, financing costs are rolled into the price of the contract. This is why the value of Australian Dollar futures contracts are different than the price of the AUDUSD spot rate. My patent application would make spot contracts easy to price and possible to commonly clear, and allow leverage for these contracts.

The original application of the idea was to have spot FX contracts listed on a futures exchange and commonly cleared.

The clearing and daily settlement process described within the patent application would result in a string of extremely interesting cleared derivative and leveraged products, like spot IRS, cleared and standardized CDS, and a range of other cool products too.

Additionally, I tried to design a GDP futures contract for my job in 2007 and 2008. This was prior to even hearing about Scott Sumner.

I thought initially GDP contracts could be huge contracts for the exchange, with large daily volumes. After I looked into GDP contracts, I concluded GDP contracts were fatally flawed, even using the Federal Government as the market maker.

This was before I heard of Scott Sumner and his NGDP level futures. The final recommendation was to drop the idea of GDP futures entirely.

Inflation futures were also considered during this time. Because as soon as you think about GDP, you think about nominal GDP and the GDP deflator which then creates real GDP.  The GDP deflator is a measure of inflation, which ends up being a bit different than CPI or PPI.

Inflation futures also failed a few basic tests.

How to decide if something might be a viable futures contract

A few basic questions should be answered in a high level of detail when considering any futures contract.

  1. Who will trade this contract?
  2. When will these users trade this contract?
  3. How do the margins interact with the price movement and commissions?
It is impossible to guarantee people will trade a futures contract. But the contracts can be so flawed people won’t trade them. It’s a filtering process, where a good contract passes a series of tests, and bad contracts/products fail the same tests. When bad contracts fail these tests, it’s certain nobody will trade them. Good contracts pass the tests, and it’s then possible the contracts might be traded by someone.

NGDP level futures fail the basic tests very badly.

So let’s go down the line and answer these questions. Once they are answered, you’ll see how any viable NGDP level contracts will end up handing Goldman a $500bn payday while doing almost nothing for the real economy.

#1: Who will trade NGDP level futures? Answer: Only the Goldmans of the world and Hedge Funds

The major participant groups for futures market are market makers, large speculators, hedgers, and small speculators. It’s possible to put every trader of futures contracts into one of these groups.

Markets revolve around hedgers, or “dumb” money. Profits for market makers and speculators in actively traded futures markets come about because hedgers are largely price indifferent. A hedger doesn’t care about the exact price received on the trade. Hedgers care about eliminating the adverse price change risk. I don’t want to go into more detail on this, but please take my word for it, or better yet, think it through yourself why hedgers are the source of profits.

The first question to ask is: who needs to trade this contract? Who will call on the phone and say “Thank god you listed this contract – I placed my first trade already!” Who needs to hedge NGDP level risk?

You’d think almost everyone. That’s what I originally thought when I investigated GDP contracts. I was wrong.

Corporations? Major Corporations will not trade NGDP level futures. There is a huge problem for corporations – being wrong with NGDP level futures will put them out of business. Imagine someone like Target staring at a bad quarter due to a sluggish economy. Target attempts to make up for lost profits by selling NGDP futures – but surprise! NGDP comes in better than expected and Target loses money on its sale of NGDP level futures in addition to losing money in their day to day operations!

All of a sudden a bad quarter becomes an awful quarter which threatens to put Target out of business.

Then, many major corporations have a very slippery relationship between NGDP levels and corporate profits. If you look at corporate profits vs. GDP growth, you’ll see this is the case. Our best GDP growth in the last 50 years – the late 1990’s – had ok corporate profits. We’re hitting quarter after quarter of record corporate profits during a period of sluggish NGDP growth.

Here’s a chart of NGDP percent change YoY *3 and Corporate Profits percent change. The relationship is not very strong at all, and not consistent in sign. It’s not a hedge worthy relationship for corporations in aggregate.

For example, what is the relationship between NGDP level and profits for Walmart? Figure Walmart probably has the best feel for actual levels of economic activity of any consumer corporation in the nation. Yet, would Walmart actually trade NGDP level futures? I think not.

The non-participation of corporations alone is a huge strike against NGDP level futures. This problem is enough to sink the contract just by itself.

Municipalities will not trade NGDP contracts. Orange County, Jefferson County – it’s not going to happen in sufficient size to make a difference to the economy.

I’ll spare you the discussion of small speculators, of which 90% lose money, and can’t be a big factor in this market due to obvious reasons. Imagine Granny losing money on NGDP futures, sponsored by the U.S. government, and tell me what politician in would consider proposing the fed doing this.

Pension funds have no direct interest in NGDP level futures.

So what firms do have the pieces in place to trade these contracts? What kind of firms have:

  1. A huge team of economists and analysts already on staff looking at economic data from every possible source – and also provides extensive and detailed real-time economic forecasts; and
  2. A vast network of contacts across multiple industries for boots on the ground information from the real world; and
  3. Vast trading and risk management expertise?

HOLY HELL! Goldman Sachs has all of this infrastructure already in place!

I think we’re looking at Goldman Sachs and other IBs, plus hedge funds and professional traders as the primary users of NGDP level futures. This was my conclusion when I first looked at GDP futures.

Basically, are perfectly designed for places like Goldman Sachs and other IB and hedge funds to trade.

No, I didn’t forget – the title of this post is: “Why NGDP level futures contracts will give Goldman Sachs $500bn in one day, part 1”

#2: When will users trade this contract? (Answer: The biggest MOC order in history)

If you’ve traded before, you know the term “Market on Close”. It’s where you place an order to go to the market and buy or sell at the close of business. It’s known as MOC by us “big shot” traders.

Here’s a few questions: Why would anyone trade NGDP level futures early in the life of the contract? Why not wait until the last possible minute to place any trade in the NGDP level futures contract!

According to Scott Sumners specification, the Fed makes a market at 5% NGDP level growth. The Fed is good at the 5% price today, tomorrow, and at the very last minute, in unlimited size and liquidity.

There is no reason to “trade” any market if you know the price you’ll get. If you know the price, wait until you have the most information possible, then make a decision to trade or not to trade. If price is never the problem, time becomes the scarce and valuable resource.

Any trader with a few brain cells can easily figure out it’s better to wait when you know the price you’ll get on your trade. Those Goldman people are extremely smart, and so are the hedge fund people. They will wait to place a trade until they have all the information possible. It’s what I would do. It’s what you’d do too – now that I told you this obvious flaw in having the fed make a market at a set price.

This means the only people who will trade the contract – Goldman, IBs and hedge funds –  will wait until the last day of trading to place their orders.

Give me 1 good reason to trade this market before the last possible moment. One good reason. I’ll take on all comers in the comments. After I’ve pointed this out, would you personally trade on anything but the last day? No.

Scott Sumner envisions a world where the market gives information to the fed based on the level and direction of open interest in the futures contract. This is a horribly misguided idea. Nobody would trade this contract except possibly on the day prior to expiration. Any other choice of trade time would be stupid.

I didn’t forget – the title of this post is: “Why NGDP level futures contracts will give Goldman Sachs $500bn in one day, part 1”

I have not tackled this question yet, but I will in Part II.

  • How do the margins interact with the price movement and commissions?

We know who will trade this contract (Goldman), and when they will trade it (on the last day, at the close). The answer to the question about price and margin ends up being a stake through the heart of NGDP level futures, unless you think giving Goldman $500bn in one day is a good idea.

If you want a hint as to how this works and why it’s the case, look up David Beckworth’s post on the success of the U.K. in hitting a 5.3% NGDP level target, and think about the implications for how margins and price would interact given that history.


(Update 1-24-2013: More on exactly how goldman will make billions here.  )


Expert in business development, product development, and direct marketing. Developed strategic sales plans, product innovations, and business plans for multiple companies. Conceived the patent pending Spot Equivalent Futures (SEF) mechanism, which allows true replication of spot and swap like products in the futures space.

View all posts by

Leave a Reply

91 Comments on "Why Scott Sumner’s NGDP Level Futures Will Give Goldman Sachs $500bn in One Day, Part I"

Notify of

bill woolsey
4 years 10 months ago

The reason to trade the future is because you believe that nominal
GDP will be away from the target even after the Fed makes whatever adjustment in monetary policy it believes is best.

This would happen if you disagree with others willing to take positions on the future, includng the Fed.

Now, after the Fed stops trading, the price of the contract is free to change. It has no policy relevance. Whether people choose to trade up to the last minute, is up to them. But the Fed would not be committed to fixing the price up to the contract is settled.

By the way, none of us assume that the contract will be settled based upon the prelimary value of nominal GDP. The final one, at the end of the subsequent quarter would be better, I think.

Anyway, if people trade them away until settlement time, that is OK.

If people quit trading when the Fed stops, and wait for the cash settlement, that is fine.

It doesn’ t make any differrence.

I certainly don’t favor this reform so that traders can make profits on short term price movements. Or for brokers to make money on commissions. Why would anyone favor a market for such reasons?

4 years 10 months ago

Hedge funds would trade this a lot to hedge their VIX positions.

4 years 10 months ago

Personally, I don’t see why we need such a convoluted mechanism to get new money into the economy. I don’t see (other than for obvious scaremongering reasons) why we don’t just have a committee at the treasury, entirely independent of politicians, decide how much expenditure of the government is to be financed by new money, rather than through taxation or borrowing. Whereby the quantity of expenditure financed by new money is adjusted so as to reach an inflation or NGDP target. This would leave the central bank with the role only of being a LOLR and preventing interest rates getting excessively high, perhaps interest rates would still be the main driver of AD in normal times, but with this new money mechanism being implemented only when in a severe recession/stagnation with rates near zero.

Can someone, ANYONE, please critique this for me. Whenever I suggest this idea it’s just ignored, is it really that stupid?

4 years 10 months ago

The purpose of the futures contract is to get people to reveal their divergent views about what will happen to nominal GDP. It isn’t used to get money into the economy.

As for your committee approach–what happens if the government wants to reduce its debt, but the quantity of money needs to rise. The solution is simple–create money and buy existing debt. By my reading, you would insist that the government either spend more or cut taxes.

4 years 10 months ago

If the purpose is to get a market forecast of NGDP, surely this can be done by other methods; for instance the BoE provides quarterly reports (used to anyway, don’t know if it still does) of various predictions of inflation by private institutions (consultancies, banks, research groups etc..); it’s not a market forecast exactly, but its a forecast from the private sector. I suppose the market is more effective at revealing peoples true forecast whether they like to or not; an NGDP futures scheme might work, but then what purpose does it serve to have the fed manipulate the price of the contracts directly? That makes no sense.

As for debt monetization, I agree that’s a good approach. But again it should be done directly (currently illegal), rather than the fed buying debt from member banks on the secondary market in exchange for excess reserves; which is far less direct and much more convoluted because that money doesn’t really circulate. What would be wrong with a more direct approach, so long as politicians get nowhere near it?

4 years 10 months ago

My point is – if the way to affect nominal GDP is still OMOs, then just do the OMOs. Is your claim simply that NGDP futures are the most efficient/ effective way for ‘people’ to reveal their expectations of NGDP?

4 years 10 months ago

Why not just take an opinion poll?

4 years 10 months ago

Alternatively, the committee could decide how much debt to monetize (targeting domestic debt holders only), this would benefit from not having possibly undue influence on the decision of politicians regarding current expenditure, however the mechanism would be probably less powerful.

But yes, criticisms please!

bill woolsey
4 years 10 months ago
The purpose of the contract isn’t for the Fed or anyone else to make lots of money of brokerage fees. The purpose of the contract isn’t for clever traders to make money at the expense of hedgers or small speculators. Frankly, I don’t think any futures market is valuable if that it the goal. The last minute trading issue was brought up by Garrison and White in JMCB 1997. Sumner responded in 2006, “Let a Thousand Models Bloom.” While I am not convinced that trading all occurs at the last minute, the Fed can still adjust monetary policy after trading closes. Trying to stick the Fed with a position it doesn’t like at the last minute would be foolish. For example, the Fed trades the quarter 3 2013 contract all during quarter 3 2012. Goldman Sach sticks the Fed with a 500 billion position on the contract at the last hour of the last day of quarter 3. The Fed understakes open market opeartions in bonds over the next few days. So what that this is during quarter 4 of 2012. What buyers and sellers are looking for is differences in expectations. The trader (maybe Goldman’s Sachs) wants a situation where it thinks nomimal GDP will be below target and someone else has the opposite view. That someone else could be the Fed. Of course, everyone wants to be right and the others wrong. I found your argument about Target hedging and your representation of corporate profits to be wrongheaded. But I must admit that I don’t really see any benefit to creating these futures for the purpose of hedging. I am sure you mispoke when you imagined that Target would sell futures during a quarter when nominal GDP was low. Surely, they would sell futures now on the chance… Read more »
4 years 10 months ago


Let’s say that the CB is indeed abel to peg the price of NGDP futures in this manner. What causes the NGDP to track the NGDP futures?

4 years 10 months ago

The central bank impacts nominal GDP by changing the quantity of money through open market purchases with bonds.

4 years 10 months ago

Which it does/ can do even now. So why the NGDP futures?

4 years 10 months ago

To provide an incentive for people to communicate their views about what will happen to future nominal GDP.

4 years 10 months ago

“The central bank impacts nominal GDP by changing the quantity of money through open market purchases with bonds.”

Japan? Still deflating.

4 years 10 months ago

When Japan has an explicit nominal GDP growth path target, no interest on reserves, and the entire Japanese national debt belongs to the central bank, then come talk to me.

I don’t claim that inflation targeting works well, much less that modest open market operations are sufficient. Get base money equal to the target for nominal GDP–for starters.

Anyway, I have more ideas, but don’t even waste my time with “open market operations don’t work,” until there is nothing left that it is legal to buy.

4 years 10 months ago

Bill – from what I make of your comment, this is like a S&P futures contract with Dec 2013 expiry on which the Fed makes markets only during Dec 2012. Why would anybody trade with the Fed in such a strange and limited market? It is hard enough for us financiers to start up new markets with sufficient liquidity in futures which trade up to the point of expiry of the underlying, let alone a stunted shadow of a real market which is what this seems to be.

Scott Sumner
4 years 10 months ago

BTW, In my proposal each contract is only pegged by the Fed for one day. So of course they’ll all trade the “last day,” there is no other day!

4 years 10 months ago

“So of course they’ll all trade the “last day,” there is no other day!”

So what are you defining as the end of days— I mean the last day, the BEA issues for each quarter’s GDP an advanced estimate, a revised estimate and a third estimate (issued a month apart and sometimes they come back months later with more revisions).
How do you score a game when the points can change months after the fact?

4 years 10 months ago

I think the best approach is to settle based upon the final results.

But Sumner is hear discussing a proposal to use interpolation between quarterly results to determine daily nominal GDP values, and settle contracts traded each day according the the difference between the extrapolated actual value and the target value. And so, the contract for each day is only traded one day.

My own view is that it is better to trade for a longer period of time, the entire quarter.