$110 bills are still lying on the sidewalk

I have a new post on the Hypermind NGDP prediction market, over at Econlog.  I argue that it might be best if the market fails.

Even so, I’m encouraging people to participate.  The prize money for each contract is over $36,000, and it costs nothing to participate.  Where else in life can you win money with no risk of losing?

Only 321 people have participated in the first contract, which ends in April, and even fewer in the second, which ends at the end of April 2019.  At that rate, the average amount of winnings per participant will exceed $110.

I’d also encourage journalists to pay more attention to this market.  What other data point better describes the expected growth in aggregate demand over the next year?  Be specific.





14 Responses to “$110 bills are still lying on the sidewalk”

  1. Gravatar of StatsGuy StatsGuy
    20. February 2018 at 17:17

    Treasury should launch NGDP indexed bonds. Would get you a better prediction. Float them at X% base rate with payoff = X% + NGDP rate. X% probably is negative (right now). Bump up the yield against the price of TIPS, and… NGDP prediction that is backed by the deepest and most liquid market in the world.

    I don’t believe small markets, but I’m open to being convinced.

    What would be really interesting is “conditional NGDP/RGDP” markets. So, you could float a contract which is “predicted RGDP growth rate over 2 years IF the Trump Budget gets passed”, and then force everyone at the OMB & CBO to invest 20% of their salaries in it. 🙂 Let’s see what that does to “dynamic scoring”.

  2. Gravatar of Negation of Ideology Negation of Ideology
    20. February 2018 at 18:41

    Scott – Off topic, but I didn’t know if you saw Ponurru and Beckworth’s column arguing the Fed has been too tight:


  3. Gravatar of Alec Fahrin Alec Fahrin
    20. February 2018 at 21:17

    I agree. Sometimes there is no success like failure, paradoxically enough.

    The Fed has kept NGDP growth much more stable ever since you began blogging. That is good, and yet it hurts the necessity of this blog and your efforts.

    My apologies that the quality of comments here has diminished though.
    Everything comes to an end. I just wish I had arrived here sooner.

    Yet, maybe this is not. Unforeseen events and a Fed regime change can affect much. We all should hope not, but few crises are obvious beforehand.

  4. Gravatar of dtoh dtoh
    20. February 2018 at 23:03

    You’re not including all the transactions costs (our time) in your calculation… or taxes if we win.

  5. Gravatar of Inklet Inklet
    21. February 2018 at 00:50

    How is/are the winner/s decided? Do you have to exactly predict the number and you win it all?

    If so, I don’t think you could get many conclusions or predictive power in the long run. Markets work because there is a selective pressure that removes bad forecasters and rewards good ones. This reward is not “all in”, and the better you get the higher the prize, but even if you are only close to the real value you win something.

    If there is no penalty for losing you don’t remove bad forecasters. If only the exact number wins, it becomes a lottery and the effort of actually researching and come up with a number is not worth it.

    How does Hypermind deal with those problems?

    As a side note, It took me a while until I saw that I had to register (and ask to accept me it seems) to make the prediction, so maybe the web layout and registration process could improve a bit. I tell you all this as someone who just found out about Hypermind (and your blog a week ago. I’m reading it all, comments included, but I’m still in 2010), so many people could have similar reactions to me.

  6. Gravatar of Benjamin Cole Benjamin Cole
    21. February 2018 at 03:06


    The Bank of Japan has been too tight.

    “The jobless rate (in Japan) stood at 2.8 percent in December last year and the job-to-application ratio hit 1.59, the highest level in more than 43 years, meaning there are 159 job openings for every 100 job seekers in Japan.”–Japan Times

    Then there is this:

    Over half of Japan firms do not plan base pay rise this year
    Reuters-Feb 19, 2561 BE
    The firms intending to boost wages cited returning a portion of profits to employees and motivating workers. Some even cited government pressure. “We have no choice but to raise base pay because of policies by the government and Keidanren,” wrote a manager of another transport equipment firm.


    My guess is that the Phillips Curve and NAIRU are not universal.

  7. Gravatar of Inklet Inklet
    21. February 2018 at 05:15


    “My guess is that the Phillips Curve and NAIRU are not universal.”


    Some people think that the Philips curve was due to women entering the work force.

  8. Gravatar of bill bill
    21. February 2018 at 08:03

    Should we, or the Fed, read anything into the fact that the 10 year TIPS rate has been more volatile since 2008?

  9. Gravatar of ssumner ssumner
    21. February 2018 at 09:20

    Statsguy, Yes, in a less stupid world the government would subsidize trading in lots of conditional prediction markets. They don’t do this, because they don’t want to know the truth.

    Negation, Yes, a very good article.

    Alec, Yes, good news for the economy is bad news for me, and vice versa.

    dtoh, Not everyone is as rich as you. 🙂

    Inklet, The winner’s payouts are based on trading profits. Each trader works with play money, and at the end the biggest winners are those with the largest trading profits.

    Yes, there may be uninformed traders, but as a practical matter only well informed people tend to trade in this market. To begin with, how many people even know there is an NGDP futures market? Those who do, usually know what NGDP is.

    Having said that, obviously the market is far from perfect. It’s a demonstration project. The Fed should provide $100,000,000/year in prize money, and you should have to invest some of your own money.

    Bill, No, inflation doesn’t really matter.

  10. Gravatar of sd0000 sd0000
    21. February 2018 at 10:15

    The UI on hypermind is just awful. It looks like something from the mid 90’s. Is there even a way to zoom in on the Y axis? It’s impossible to see the fluctuations which is the actual interesting part.

  11. Gravatar of Christian List Christian List
    21. February 2018 at 12:54


    Tulip mania: the classic story of a Dutch financial bubble is mostly wrong.


    -> Scott was right again.

  12. Gravatar of bill bill
    21. February 2018 at 13:49

    Inflation doesn’t matter, except that it is the Fed’s target and the market has at times predicted as much as a 5% cumulative undershoot over the next 10 years. Which is probably more like an 8% cumulative undershoot when converted to PCE, the real stated target.

  13. Gravatar of dtoh dtoh
    21. February 2018 at 15:14

    Even if you value the punter’s time at the minimum wage transaction costs will still exceed expected after tax returns. 🙂

  14. Gravatar of ssumner ssumner
    22. February 2018 at 15:26

    Christian, I think Gruber discredited the Tulip bubble about 20 years ago.

Leave a Reply