A new NGDP prediction market

I finally limped home from a long trip, with a bad cold.  The trip actually started well, as I taught a few classes at Alternative Money University from July 15-18, at the Cato Institute in DC.  Thanks to George Selgin, Lydia Mashburn and the other people at Cato for organizing an outstanding program.  And a special thanks to the students, who restored my faltering faith in humanity.  A number of them seemed very interested in pursuing market monetarist research ideas.  Given that there were students from schools like MIT, Chicago and Stanford, that bodes well for the future.  AMU was probably the most personally rewarding experience I’ve had since I started blogging, far more so than the flurry of publicity I received back around 2012.

Basil Halperin was one of the students at AMU, and he recently did a blog post describing a new NGDP futures market at Augur:

[A]n NGDP futures market is now live on the Augur blockchain. The specific contract is simply a binary option: will the growth rate in NGDP from 2018Q1 to 2019Q1 be greater than 4.5%?

The current price/probability implied by this contract can be viewed on the Augur aggregator website predictions.global: just search “NGDP’, or the permalink is here.

. . .

For those unfamiliar, Augur is a new cryptocurrency project – it launched just last Thursday – built on the Ethereum platform that allows holders of its currency, “REP”, to create prediction markets. To speculate on such markets, an investor must use the Ethereum cryptocurrency (ETH).

The platform is decentralized: for everyone who wants to bet that NGDP growth will exceed 4.5%, there must be a counterparty who takes the other side of the bet. That is, the creators of Augur are not acting as market makers for the contract. The price of the contract will move to equilibrate supply and demand in a decentralized market: if the price is 0.7 ETH, that indicates that the market gives a 70% (risk-neutral) probability that NGDP will exceed 4.5%.

Read the entire post, it is quite interesting.

BTW, about 18 months ago I did a few posts (here and here) discussing Basil’s research on NGDP targeting.  He has a bright future.

I spent this past weekend going back and forth between an old folks home and an IHOP in Arizona.  Because the AC at the IHOP was giving me a fever and chills, I dressed up with three layers of shirts and long pants before walking in 105 degree heat to the restaurant (fortunately just a block away.) With nothing else to do, I read Eliezer Yudkowsky’s excellent book Inadequate Equilibria.

It’s hard to summarize the book in a single sentence, but here are a few themes:

1. Whereas financial markets are highly efficient, many of our other institutions are poorly designed—inadequate equilibria.

2.  While it’s generally unwise to believe that one can beat the stock market, we are often too modest in deferring to existing institutions, or conventional wisdom on a given issue.

3.  We need to rely on both theory and data.  Be a hedgehog and a fox.

The book explores when we should be willing to believe that we have an idea that others have overlooked.  It might be a public policy idea, a start-up company idea, a new medical treatment, or a new academic theory.

One example cited by Yudkowsky is treatments for Seasonal Affective Disorder (SAD), which afflicts millions of people.  He experimented with stringing up 130 LED lights in his home as a way of helping his wife, and it seemed to be sort of successful.  Then he discussed all the reasons why the market might not be expected to produce this treatment.

While reading the book, I could not stop thinking about NGDP predictions markets.  In the past I’ve argued that the failure of the government to set up and subsidize such a market is an example of near criminal negligence.  In Yudkowsky I’ve found a kindred spirit—someone who is outraged by things that the vast majority of people couldn’t care less about, like the ingredients that go into hospital formula for infants, or the fact that many doctors are too lazy to wash their hands as often as they should.

I’ve talked to many economists, and I have yet to hear a single plausible excuse for the lack of a federally subsidized NGDP prediction market.  Not one.  People just sort of shrug, because the cause doesn’t pull on our heartstrings like those kids separated from their parents on the border.

Yudkowsky is similarly exasperated.  Where others see nice shiny hospitals that “help people”, Eliezer sees a monstrous medical industrial complex that needlessly destroys lives.  And he sees these sorts of failures all over the place:

If you truly perceived the world through the eyes of a conventional cynical economist, then the horrors, the abominations, the low- hanging fruits you saw unpicked would annihilate your very soul.

Of course all this refers to the “normal” state of affairs in America, pre-Trump.

In any case, I highly recommend the book. It’s one of those books where the question of whether the author is “right” or “wrong” is almost beside the point; what’s interesting is how Yudkowsky approaches questions.

Here’s a review by Robin Hanson, another by Scott Alexanderand another by Scott AaronsonIf I were made king of the world, I’d probably just turn it over to those four bloggers.  I’m not sure what they’d do, but I pretty sure that none of them would put their ego ahead of the well-being of billions of people.



8 Responses to “A new NGDP prediction market”

  1. Gravatar of H_WASSHOI (Maekawa Miku-nyan lover) H_WASSHOI (Maekawa Miku-nyan lover)
    23. July 2018 at 20:42

    (memorial comment)

  2. Gravatar of Becky Hargrove Becky Hargrove
    24. July 2018 at 05:33

    Thanks for the reminder re Yudkowski’s book. I’ve been meaning to read it since those earlier reviews came out.

  3. Gravatar of Randomize Randomize
    24. July 2018 at 09:40

    I hate to sound like a broken record but…

    Having the treasury issue NGDP-Adjusted Bonds (NIPS or GIPS?) with the Fed pegging the treasury spreads would be the ultimate way to mesh market forecasts with market intervention.

  4. Gravatar of Don Geddis Don Geddis
    24. July 2018 at 09:47

    You are definitely a kindred spirit to the other four bloggers you mentioned. It’s a shame that so few people are able to think and act that way. (It’s almost amusing that, when you’re one of them, it seems so blindingly obvious, and it’s confusing and frustrating why others seem to find that mode of thinking so rare.)

    But at least the five of you are out there, publicly, doing both concrete work, and also providing an example of what is possible.

    (I can throw out a couple more in the “same space” — at least, a similar feeling — to the group you’re describing: In podcasts: Sam Harris. In martial arts: Brazilian Jiu-Jitsu.)

  5. Gravatar of Brian Donohue Brian Donohue
    24. July 2018 at 10:00

    Yudkowsky’s understanding of and ability to communicate ideas around monetary policy is extraordinary. A natural ally for you, since you strike me as someone possessed of very deep understanding here but without Yudkowksy’s flair for communication.

  6. Gravatar of ssumner ssumner
    24. July 2018 at 12:04

    Randomize, That would work too.

    Thanks Don. And yes, Sam Harris is excellent. I could have added many others, such as the GMU bloggers.

    Brian, My forthcoming (hopefully?) book based on my blog would be 10 times better if Yudkowsky wrote it instead.

  7. Gravatar of Michael Sandifer Michael Sandifer
    25. July 2018 at 11:53


    Given that the correlation between US GDP and the S&P 500 is greater than .97 since 1950, what do you have against level-targeting the S&P 500? While an NGDP futures market would make more sense in most ways, it would not be as liquid, for example, as the stock market.

    It’s not hard at all to model S&P 500 prices.

  8. Gravatar of ssumner ssumner
    25. July 2018 at 17:22

    Michael, Those sorts of correlations occur for lots of variables in time series analysis, and are meaningless. Look at the correlation in first differences of logs.

Leave a Reply