Contract Specifications for NGDP Futures

I know, this is beating an issue into the ground.
Here is something I’ve put together on the contract specs for NGDP futures. It’s clear I am not a fan of NGDP futures, but Scott Sumner complained I didn’t bother reading his papers on NGDP futures.
He’s right. I did not spend the money to go and read these papers. I relied on information from his blog to piece together his vision of NGDP futures. In particular, I relied on a post from May of 2009 to get his basic picture of NGDP futures.
Here are what I assume would be his contract specifications for NGDP futures, based on my reading of the May 2009 post.
This post is mostly a holder of information so we know the specs for these futures.
Contract Specifications:
Settlement Value: Bureau of Economic Analysis NGDP Table 3 times 1/1,000,000,000,000, rounded to the nearest .01  (This contract size is far too low)
Price Quote: Dollars and cents to hundredths of a cent. Example Value: 15,585.6 (from BEA Table 3 Line 1) * 1,000,000,000 = 15,585,600,000,000/1,000,000,000,000 = $15.5856
Tick Size (Minimum Fluctuation): 1/10,000 (one ten thousandth) of a dollar.
Tick Value: $.0001
Contract Months: The first 12 contract months in the quarterly January, April, July, October cycle. Three years of quarterly NGDP futures contracts.

Last Trading Day: Release date of Preliminary last Quarter GDP Report. Trading in expiring contract closes at 7:29am Central Time on the last trading day.
 Special Rules:
Margin: 10% of full contract value (Instead of normal futures margin calculations)
Margin Interest Rate: To be set by U.S. Federal Reserve at a rate high enough to insure sufficient liquidity
Market Maker: U.S. Federal Reserve
Contract Price: Two possibilities
  1. Set by U.S. Federal Reserve on a 5% p.a NGDP per capita growth path. Contract cannot trade at any other price as U.S. federal reserve promises to buy unlimited quantities of contract at this price.
  2. (This was by Bill Woosley in the comments section at MR and not Scott S in his post) Set by U.S. Federal Reserve on a 5% p.a NGDP per capita growth path until 9 months before contract expiration. Contract cannot trade at any other price as U.S. federal reserve promises to buy unlimited quantities of contract at this price. When expiration is closer than 9 months, contract is allowed to float in price, and fed no longer counts this open interest to the net open interest for the related open market operations.
Open Market Operations: Federal Reserve performs open market operations based on net open interest across all active contracts long/short at a 5:1 ratio. Example: Net futures position held by public is short $500m notional value of NGDP futures, so U.S. Fed purchases $2.5bn of T-Bills in response.
Comments
  • David Beckworth November 6, 2012 at 2:06 pm

    Michael,

    Since you see problems with NGDP futures, can you think of way to design and implement them that would over come these problems?

    • Michael Sankowski November 7, 2012 at 1:16 am

      David,

      I have tried. NGDP futures face problems which are very difficult to overcome.

      The problems are not due to targeting NGDP, or any of the reasons Scott addresses in his papers. They are problems which are inherent in every futures contract ever designed. Successful contracts overcome them, but I don’t see ways for NGDP futures to overcome them.

      I do believe using unemployment is a better way to guide policy. It’s way easier to count unemployed than it is to estimate NGDP.

      Who are the hedgers for this contract? Every successful contract has a highly motivated hedging population who uses the contract every day, no matter what the price is. Hedgers are the major participants in every “trading” market, and they are why exchanges exist. I don’t see much or any reason for major corporations to hedge NGDP risk. NGDP risk is too diffuse, too uncorrelated with individual payoffs, to attract hedgers.

      The “long and variable lag” of monetary policy doesn’t really help someone who actually needs a boost for next quarter.

      I don’t see a good way to overcome this problem for NGDP futures.

      When I think about unemployment, it seems to be a real world vote by employers on how much they want to hire.

      I proposed the TC rule for fiscal policy, but it could easily be adapted to monetary policy, or even some combination of fiscal and monetary policy. Instead of using the inflation rate as one of the inputs, we could use the NGDP level vs. the target NGDP level.

  • beowulf November 6, 2012 at 7:13 pm

    The unemployment rate is, by far, the better dashboard gauge to adjust policy (like James Meade said, float payroll tax rate inversely to unemployment rate). GDP is too fuzzy and its lead time is too long.

    The U3 unemployment rate reported on the first Friday after every month with BLS making revisions once a year (for trailing 5 years). GDP is also reported 12 times a year, its quarterly (NOT monthly) numbers are reported 3 times each ,not counting BEA’s annual revisions.
    –First (Advance) Report, issued a month after the end of the target quarter
    –Second (Preliminary) Report, issued two months after quarter
    –Third (Final) Report, issued three months (an entire quarter!) after quarter. A few days after this, BLS would be releasing its THIRD monthly U3 update for the quarter after the target quarter.

    The numbers do jump around. For example in 2Q12, Advance Report says 1.5% growth, Preliminary Report says 1.7%, Final Report says 1.3% growth. Woul you have a GDP futures market for each Report or would Cullen and I have to go offshore to set up markets for the other two Reports? :o)

    • beowulf November 7, 2012 at 12:54 pm

      Just to clarify my last point, I was teasing about going offshore with Cullen to set up our own NGDP futures market; that would be like drummer and the guitarist for The Police deciding that Sting was dead weight and had to go.
      Mike’s has been a Chicago trader for years and knows more about the futures market than anyone you’ve ever met, read or heard of. Relevant to this discussion, he’s even invented and patented a unique futures product before.

      I was piling on with the problems with using GDP, but that was putting the cart before the horse. If Mike is stumped with how to set up a functioning NGDP market, its reasonable to conclude its just not possible to set up a functioning NGDP market. To paraphrase the Zero Dark 30 trailer, there is no secret team on another floor with all the answers, its just Mike. :o)

      • Michael Sankowski November 7, 2012 at 8:25 pm

        lol – that’s right. There are no secret rooms out there, with some secret team of futures contract designers, who are going to come up with something magic to make this all work. If I am not seeing it, I’d say it probably doesn’t work. I could very well be wrong.

        One thing we know – the CME, Eurex, and LIFFE (NYSE) have not listed a GDP futures contract, much less a successful GDP futures contract which is widely traded today. Neither is there an active swaps market, or even any GDP swaps traded.

        Futures markets typically follow what is already happening in informal OTC markets. This is how the first futures exchanges started – they mimicked, formalized, and standardized what was already happening between traders.

        The CME/CBOT is a ruthless company. So is Eurex. If they saw the need for these products, they would list something to try and capitalize on the strength of their clearinghouse.

        John Labuszewski is the MD of product development at the CME. He’s the guy to talk to about NGDP futures. However, I strongly suspect he’s going to echo nearly all of my concerns. I’d be completely shocked if he (really someone on his team) hasn’t already done a formal analysis of GDP futures structures at some point years ago, and he probably included NGDP futures in his analysis.

        http://www.marketswiki.com/mwiki/John_W._Labuszewski