Attention iPredict donors

A couple of years ago I raised funds for two NGDP prediction markets, Hypermind and iPredict.  The money for Hypermind all came from Valve CEO Gabe Newell. A larger amount of money was raised for iPredict, from about 15 people. Unfortunately, iPredict had to end its experiment after a brief run. For the past year I’ve being making inquiries about a refund, and it’s finally paid off.  Here is the information we received from iPredict:

Here is how the donors can go about requesting a refund.

They should contact Iain Devon, Viclink Senior Account Manager, at Their donation has been held in NZ$ and will be refunded at the current USD/NZD exchange rate, which means they may receive less than their original donation due to change in exchange rates (assuming the donor wants to be refunded in US$). They should provide the details of the initial donation made, including date and value. Refunds will be issued via international bank transfer, so they should also be prepared to provide their bank account information.

I want to thank the people at Wellington Victoria University in Wellington, New Zealand.  Their willingness to return the funds further cements New Zealand’s reputation as one of the least corrupt countries on Earth. If you forgot how much you donated, you might check your old emails to me.  I believe all the donors emailed me and informed me of their intentions.  I also probably have that info, if you need it.

During the period after iPredict failed, I vowed not to try to raise additional funds until the issue of refunds could be resolved.  Now that a resolution seems imminent, it’s time to think about future plans for NGDP prediction markets.  My inclination would be to go back to Hypermind, but with a bigger donation this time. I also feel like the annual market is the most macroeconomically useful, even though it is a long time to wait for a payoff.  (Say a 2018:Q1 over 2017:Q1 contract).  I believe we already have about $10,000 to work with, which is double what the annual market had back in 2015.  More money could be raised. With Trump in office, there might be some interesting policy shocks which could impact the market (although it’s also quite possible that NGDP expectations are not greatly affected—either result would be interesting.)  I’m also open to other markets, if someone has a suggestion.

PS.  Thanks to my colleague Ben Klutsey for working with the iPredict people to arrange this refund.



25 Responses to “Attention iPredict donors”

  1. Gravatar of Major-Freedom Major-Freedom
    13. February 2017 at 14:26

    Being serious…how about requesting a presentation for the TV show Shark Tank. I often see investment ideas of questionable value that fail traditional financing attempts getting the green light there. Long shot though…

    If you frame it as a Trump hedge, you may get Mark Cuban on board.

  2. Gravatar of Mark Thomson Mark Thomson
    13. February 2017 at 15:50

    “Wellington University” – No one in NZ would know it by that name. I presume you’re referring to Victoria U. (of Wellington).

  3. Gravatar of dtoh dtoh
    13. February 2017 at 16:04


    1. As I have said many, many times (and I won’t repeat the arguments) but NGDP linked notes (GAINS) would be a much better instrument.

    2. Why do you need a prediction market? So what if the Fed economists are a little off and have to make slight policy adjustments when the real numbers are announced.

  4. Gravatar of ssumner ssumner
    13. February 2017 at 16:54

    Mark, Thanks I corrected it.

    dtoh, That would have all the flaws of TIPS spreads. And it’s far more costly to implement.

    I have many posts on why we need prediction markets, there are many reasons. The most important reason is that we need better ways to identify monetary policy shocks.

  5. Gravatar of Ray Lopez Ray Lopez
    13. February 2017 at 17:29

    @ssumner – why don’t you just go to the well once more, Daddy Warbucks Ken Duda? He’ll just sign a blank check? No need to answer, I assume the answer is you know better. Duda is a businessman and won’t throw good money after bad.

    PS–I did open a iPredict account, and made no money and few trades since the volume was too thin. NZ, AU are in the ends of the earth (EOTE) (anywhere you can see the Southern Cross counts as EOTE = 100 years of solitude). Like to tilt at windmills Dr Sumner?

  6. Gravatar of dtoh dtoh
    13. February 2017 at 17:57

    Why would you possibly recreate the same issues with TIPS if you were trying to create a new instrument.

    A real futures market is neither easy nor cheap to set up.

    Hypermind and iPredict are pure amateur hour, and you’re wasting valuable credibility and mind share bullets on this Quixotic distraction.

  7. Gravatar of Sam Sam
    13. February 2017 at 19:32

    Scott, why not have four annual competitions for the quarterly YoY changes? $2500 a pop would be decent incentive.

    Also iPredict was a total disaster, from a user-interface POV. And its implementation of automated liquidity provision (a bot that put in extremely thin scale orders) was likewise laughable. Hypermind doesn’t do automated liquidity provision (just robotic order matching on alternate outcomes to ensure no-arbitrage). I’d encourage you to use whatever clout you have with Emile Servan-Schreiber to try to convince him to implement an LMSR market maker along Robin Hanson’s design. I have a pretty strong prior that this improves liquidity in its own right, and also that providing opportunity to harvest from a subsidized market maker is a more efficient way to incentivize trading than sponsoring block grants for play-money performance. (In general Hypermind has a somewhat whacky scoring/compensation system, although for the NGDP markets this was mitigated by each one being in a separate “contest” of its own, so the funny averaging over performance on different questions within a contest did not come into play.)

  8. Gravatar of Sam Sam
    13. February 2017 at 19:39

    Semi-off-topic, now that Goldman Sachs people (Cohn, Bannon) are back in an influential policy position, I wonder if we’ll see any interest within the government in reviving Goldman’s economic derivatives markets from the early ’00s. IDK if either Cohn or Bannon were involved in that at GS — I’d guess not. Those products were pretty dismal failures in terms of market uptake. But even an implicit hint that the FRB would take them into consideration in determining policy — let alone an explicitly policy rule, mandated by Congress or otherwise — would probably change that. You’d immediately have constituencies of traders (banks) seeking to manipulate the rates by moving the futures markets, and hedge funds seeking to profit from that manipulation of the futures (by more accurately forecasting true rates). That’s a recipe for liquidity success, provided you can convince the CFTC to remain hands-off.

  9. Gravatar of H_WASSHOI H_WASSHOI
    13. February 2017 at 23:09

    yay. I will buy a MSI Vortex by the prize money

  10. Gravatar of H_WASSHOI H_WASSHOI
    13. February 2017 at 23:20

    I think +-0.1 prediction is difficult about NGDP of 2018:Q1 over 2017:Q1.
    Because future’s fiscal policy plans are vague.

    Maybe baseline is plentiful AD with sober fed.

  11. Gravatar of Edward Edward
    13. February 2017 at 23:26

    If you need an investor, email me, and we can work out a deal. For a good software program, 10,000 is not enough. It’s lunch money.

  12. Gravatar of Student Student
    14. February 2017 at 07:14

    Not sure if MF was joking but that’s actually a good idea. That show is stupid but Cuban may actually be interested and the free publicity would probably help get the ball rolling.

  13. Gravatar of Keenan Keenan
    14. February 2017 at 08:25


    Have you thought at all about the dangers of how a market like this could be easily manipulated? i.e. imagine if someone just sold as many NGDP futures as he could (in early stages, it might cost him 50k), keeping it around 1%. The Fed will then be accommodating to try to boost this. You will wind up with a situation where the Fed is targeting a “bad” estimate of NGDP futures. Thoughts?

    Or imagine this trade: The guy who is offering NGDP futures pulls his offer, the market goes from 1% to 5% in a week. The Fed (who is being very accommodating) quickly tightens and SPX drops 4%. The guy on the offer in NGDP (before pulling his offer) buys SPX puts. If this market is actually used for policy decisions, it will have to be uber liquid so these types of arbs aren’t available.

  14. Gravatar of Ray Lopez Ray Lopez
    14. February 2017 at 08:39

    Keenan is right: an NGDP futures market can be easily manipulated by traders, just like they do now with (illegal) fake buy orders in stocks. Happily, money is neutral so it doesn’t mean much.

  15. Gravatar of Daniel Daniel
    14. February 2017 at 13:12

    I assume you’ve heard of Augur ( since Robin Hanson is an adviser for this project. It hasn’t been completed yet, but it’s aiming to be a fully decentralized prediction market. I think the project will be completed this year.

  16. Gravatar of ssumner ssumner
    14. February 2017 at 13:54

    Ray, Too bad you didn’t make any money.

    I did.

    dtoh, OK, if it’s easy then why don’t you set up the market you proposed? If it’s hard, I’m not interested.

    Sam, Thanks for the info on Hypermind.

    Edward, The $10,000 is not to set up the market, it’s for prize money.

    Student, This isn’t an investment, it’s a non-profit, expected to lose money.

    Sam and Keenan, Yes, I’ve thought about market manipulation. I’ve restructured the proposal so that manipulation would not be a problem. I have a new article in the Journal of Macroeconomics with my latest version of the proposal.

    Daniel. Thanks, I’ll keep that in mind.

  17. Gravatar of ssumner ssumner
    14. February 2017 at 13:55

    dtoh, Sorry I misread your comment. I am not interested in setting up a “real futures market”. I’d like a prediction market.

  18. Gravatar of Major-Freedom Major-Freedom
    14. February 2017 at 15:12


    I was serious.

    Sumner wrote:

    “This isn’t an investment, it’s a non-profit, expected to lose money.”

    Expected to lose money indefinitely? Or just the first few years?

    If the expectation is that it will lose money indefinitely, then that is the market test showing us that the resources are more valued elsewhere in the economy.

    What gain will there be to people who give money to this? If it is not material gain, then is this venture something we would more likely find in a socialist economy where the rewards are pats on the back and silly medals?

  19. Gravatar of Student Student
    15. February 2017 at 04:05

    Scott, I understand that.

  20. Gravatar of Vaidas Urba Vaidas Urba
    15. February 2017 at 09:48


    you were mentioned in this post:

  21. Gravatar of Ray Lopez Ray Lopez
    15. February 2017 at 11:33

    @Vaidas Urba – that brief mention of Sumner in the streetwise professor site was trivial. The guy is railing about how the Fed got it wrong in 2008, but doesn’t mention Sumner’s “crowdsourcing” argument at all. Saying the Fed was too tight in 2008 is simply failing to see the forest (crowdsourcing) for the trees (trying to fix one bad call). Anyway, money is largely neutral so this whole debate is sterile, but I wouldn’t call the post you cite as worthy of mention.

  22. Gravatar of Justin Justin
    15. February 2017 at 14:17

    Scott, You could try to get the folks to set up a market. Then you wouldn’t have to raise $. I emailed them about it, tried to explain that there’d be traders, but they never got back to me. Maybe you’d have better luck.

  23. Gravatar of Bryan Bryan
    15. February 2017 at 14:45

    I agree with Justin and am maybe I’m out of the loop, but why won’t “predictit” run it for you?

  24. Gravatar of Plucky Plucky
    16. February 2017 at 13:19

    Idea: Instead of 1, annual market, you could do each quarter as a contract (listed as vs the year-ago quarter), which would implicitly also create a series of q/q expectations.

    People who are just punting-speculating would likely be attracted to the upcoming quarter, which would (despite being paid out as vs year-ago quarter) effectively become a q/q wager since the preceding 3 quarters would be known.

  25. Gravatar of ssumner ssumner
    17. February 2017 at 17:10

    Student, Do they do non-profits on that show?

    Thanks Vaidas, I did a post.

    Thanks Justin.

    Plucky, I realize the betting interest would be greater, but the macroeconomic usefulness of those contracts is pretty low.

Leave a Reply