Zitzewitz The Wise

Eric Zitzewitz on why the CFTC cracked down on Intrade:

Why prohibit something so harmless? After all, U.S. policy already allows many forms of gambling that have significant social costs and produce no useful information. The real reason may be found in debate surrounding another recent setback for prediction markets, Congress’ last-minute modification of the Dodd-Frank financial reform bill in 2010 to prohibit markets on box-office receipts planned by the Hollywood Stock Exchange (HSX).

The plug was pulled after lobbying by the Motion Picture Association of America, which argued that a prediction market forecasting a poor movie box office could lend an “aura of financial authenticity to gossip.” Translation: Even big-budget failures often have one big weekend before word of mouth kicks in, and information aggregated by prediction markets may deprive them of that.

The bigger threat is to the executives who green-light the flops in the first place. The track record of even the play-money version of HSX is quite good ‑ a real money version would presumably be better. If a prediction market can forecast a poor box office well in advance of production, it raises questions about why an executive cannot. A consultant told me about a software company that ended an internal prediction market that forecast the success of the company’s products. The problem was not that it failed to work ‑ it worked all too well, and that raised awkward questions for the egos involved. (more)

Yes this is a big reason why most firms don’t want prediction markets on internal issues, and why the film industry lobbied to prevent film futures. But I find it harder to see as a big reason the CFTC prevents Intrade, and earlier Nadex, from betting on elections. That seems more simply explained by a general public aversion to what it sees as “gambling.”

GD Star Rating
Tagged as: ,
Trackback URL:
  • TheAudientVoid

    Because elections are a sacred ritual of democracy, by which people are made to feel that they have dramatically more control over the political system than they actually do. if it’s possible to accurately predict how people are going to vote before they actually do, then this weakens the illusion of control.

    •  Then why no effort to ban polls, the analysis of which has shown more accurate results than the prediction markets?

      • Siddharth

        In fact TheAudientVoid is making a prediction. In the future, there will be attempts to ban statistical poll analysis. 

      • VV

         Aggregate poll analysis beats single polls only by a small margin, and there has been no attempt to ban polls.

      • Because no politican in the US would get away with trying to ban polls. The US Supreme court would simply overturn any law that would ban polling.  

        In Germany we do have laws that ban some polls right on election day.

      • That begs the question. If there were severe ideological prejudice against polls, we wouldn’t be apt to have laws that defend them. There are unprotected forms of speech (by legal fiat)–such as “obscenity.” Germany may ban polls on election day, much as the U.S. bans distribution of campaign literature at polling sites. “Time, place, and manner” restriction, as it’s called here. But is there any nominal democracy where polls are prohibited generally? (Like prediction markets may be.)

      • @715a731811a13051cfedfb1a7d614563:disqus srdiamond : No, it doesn’t. The laws we have are the result of the conflict between different actors. They are not the result of an answer to the question: “What laws would maximize the public wellbeing”.
        Given the recent “Money is speech” Citizens United decision you might win a a battle at the US Supreme court to give prediction markets legal protection. 
        There just nobody to fight that battle because losing it would be costly. 

        Prediction market proponents are politically weak. It’s easy for poltician’s to attack them. It’s not easy to attack pollsters.

    •   if it’s possible to accurately predict how people are going to vote before they actually do, then this weakens the illusion of control.

      Not much, if it does at all. Consider this thought experiment. There’s an elected official who is so obviously terrible that everyone predicts he’ll be overturned by landslide, as he is. Wouldn’t it decrease the illusion of control if the candidate everyone predicts will win loses? It would tend to show–this would be my reaction and I think most people’s–that voting-booth behavior bears insufficient relation to what people themselves rationally believe to make the idea of control by voting ludicrous. I don’t know that predictability is the unequivocal ally of illusions of control, but it isn’t the unequivocal enemy.

      Logically, prediction refutes free will. (See my “Another refutation of compatibilism: Newcomb’s paradox shows that free will is entirely illusory”http://tinyurl.com/cdl69lk ) But it’s a deduction people have resisted for millennia..

  •   That seems more simply explained by a general public aversion to what it sees as “gambling.”

    Do you deny that it is gambling? Ah, it’s termed trading, so it’s like investing in the stock market instead of playing poker. It’s kind of a middle case, isn’t it? Like investing in its social productivity; unlike investing in that your investment serves no end besides what derives from the outcome.

    • VV

       Investing i the stock market is also gambling.

      Maybe the difference is that the stock marketing and other forms has higher barrier to entrance than something like intrade, so gambling is considered acceptable as long as it is done by high status people who presumably know what they are doing (actually, they don’t) rather than common folks.

      •  I think the stock market isn’t perceived as being the same as ordinary gambling because stocks represent actual value. The company issues stocks, which allow it to expand production. The fact that stocks trade is secondary to the fact that the stocks are issued to raise money for a productive activity that is external to the gamble.

      • Ilya Shpitser

        Don’t positions in prediction markets represent information, which is something with value also?  Aggregating info has value beyond the money changing hands in the market itself.  That’s sort of the point.

      • Stephen Diamond


        “Positions” in a horse race, too, represent information. Prediction markets produce socially valuable information only incidentally, not in response to the market itself. It is pure positive externality, unlike stocks, which are issued to meet demands for increased production of socially valuable products.

        (Posted yesterday but apparently deleted by an AI with its own agenda.)

      • VV

         Most transactions in the stock market don’t involve newly issued stocks.

        When you buy an already existing share at price X, you are betting that the discounted cumulative dividends that will be payed by that share exceed X (or, more precisely, that the discounted cumulative utility from the dividends payed by that share exceeds the utility of having X $ now). The seller, of course, is making the opposite bet. So, unless you have greatly different risk adversion w.r.t. money or greatly different discount strategies, you can’t be both right:
        The buyer and the seller gamble against each other over the holding value of the share.

      • V.V.::

        The buyer and the seller gamble against each other over the holding value of the share.

        Every resellable good sold contains elements of a gamble. You buy a house, particularly, even a car, and you are potentially gambling on resale value. That doesn’t mean you’re  “gambling.” Some people are, of course. There are turn-over artists in the housing market, etc. But people buy houses because they want a place to live; buy a car because they want  transportation. They are compelled to gamble if they intend eventually to resell, for whatever reason. An element of gambling is necessarily involved whenever there’s a market; house buyers are compelled to gamble, but the excitement of the gamble isn’t their basic motive, because a house is a product with independent value.

        Similarly stocks have independent value; people buy them because it produces a better return than putting the money in the bank or under their mattresses. To get the best value from them, they have to be prepared to gamble, but gambling isn’t necessarily or usually the motive–because they represent value independently of the transaction process, is in gambling.

        Gambling is a zero sum game. Buying stocks or a house isn’t. It’s true that in a sale of stock you are necessarily making a bet. So, when you buy a house, you are (almost) necessarily making a bet. But in neither case is betting necessarily or typically the reason for investing. All home owners could in principle benefit from buying homes; all investors in stocks could in principle benefit from buying stocks, although some in each case benefit more than others. But all gamblers or prediction market participants cannot in principle benefit; one’s loss is necessarily the others gain. That’s what makes it gambling, rather than gambling being incidental to the operation of markets for real value.

        This seems obvious and uncontestable. I wonder if the spin the proponents of prediction markets have given them has successfully obscured these truisms.

      • VV


        When you buy a car or an house at price X from some individual that was previously using it, assuming that both you and the seller are rational and the trade is voluntary, you both gain from the transaction: for you the utility of owning the car or house is greater than the utility of X money, while for the seller it is lower.

        The holding value of a car/house comes from the utility you can get from driving/inhabiting it, and for different people this utility can be different compared to the utility of a certain amount of money.
        It’s true that there are speculators who buy these items for the main purpose of reselling them at an higher price, but in most non-pathological markets these transactions are a minority.

        The holding value of financial assets, on the other hand, comes from the money they can make: when you buy a stock at price X, you are betting that in the future it will make more than X money (corrected for discounting and risk aversion), while the guy who sells that stock to you is betting that it will make less than X money (again, corrected for discounting and risk aversion).In principle, if the seller is much more risk averse and/or discounts much more than the buyer, they could both benefit from the transaction, but typically that’s not the case: either the seller or the buyer, but not both, will make the right prediction, and the gain of one will be approximately the loss of the other one.Up to creation and destruction of stocks, and up to great differences in discounting and risk aversion w.r.t. money, the stock market is essentially a zero sum game.

      • either the seller or the buyer, but not both, will make the right prediction, and the gain of one will be approximately the loss of the other one.Up to creation and destruction of stocks, and up to great differences in discounting and risk aversion w.r.t. money, the stock market is essentially a zero sum game.

        It’s a zero sum game ex post entry but not ex ante, which is the relevant frame. An investor in stocks is strongly incentivized to buy stock if he has the money because the value of capital increases over time with the expansion of production. The value of stock in the stock market rises on average; not so prediction-market positions.

        Once in the market, a trader faces a zero sum game. But he pretty much has to accept that to maintain his assets in a changing economy. If he doesn’t trade he is gambling (by retaining his position) just as much as if he trades (minus transaction fees). Although each transaction is zero sum (at the margin) a failure to trade could produce a large loss in the long-run as the market changes.

        The difference isn’t grasped by looking at the similarities between traders and those invested in the stock market who trade. The difference between the stock market and the prediction markets is that it’s a zero sum game to enter the prediction markets; it’s far from a zero sum game to enter the stock market, where everyone can (in principle) gain.

        To compare buying a house and buying stocks, the use value of a house is for a place to live; the use value of stocks is to provide a likely source of expanding income and wealth; the use value of prediction positions is purely the procurement of gambling excitement (as the general rule).

      • Randersson

        gambling is a “label” that we have applied to a group of activities but from a purely mathmatical perspective, there there is only risk: risk is the possibility of a bad outcome. it is created by uncertainty. uncertainty is created by a lack of information. if one had perfect information, there would be no uncertainty, and thus no risk.

        middle. beginning. and end of discussion.

        the regulatory legal view/definition of gambling is that gambling creates risk … i.e. prior to placing your bet the risk did not exist … but the act of placing the gamble creates the risk …

        trading, speculation, etc is different, these risks already exist in large form, but the act of trading them shares the existing risk amongst a larger audience … in the case of a company, the risk of bad company performance is shared more broadly by markets where a larger group of investors now assume the risk … but if the stock did not exist, the risk of performance is still there …its just centralized for the owners of the company. same as commodity futures, the price of oil going up and down is there, but commodity markets enable the sharing of this risk

        I don’t buy this legal definition, because the volume of risk transfer versus speculation is about 1 to 20 in most commodity markets, far higher in others, so the true result of markets is to create new risks for a large audience that wouldn’t have this risk in the first place if they had not traded but this is the way the regulators tend to look at things …

        ironically, the word risk has a negative bias, it is the possibility of a bad outcome … there is no word in the english language to describe the possibility of a good outcome … somebody here should invent it.

  • From Zitzewitz:

    “A $10 million election contract position, 40 times the position limit that NADEX had proposed, would create only a $1 incentive to vote.”

    I hadn’t thought of that argument: perverse voting incentives. The amount is small, but the direction is salient. It is cognitively dissonant to vote for Obama and predict Romney. (See my “Uncomfortable ideas and disfluent expression affect us similarly” for my theory of cognitive dissonance — http://tinyurl.com/8m65wry )

  • Drewfus

    “aura of financial authenticity to gossip.”

    So understanding how people understand the function and purpose of gossip is the key, rather than say, pointing out that trade is historically low status relative to politics.

    Gossip is often seen as a method of regulating the behavior of individuals who step out of line or aren’t ‘pulling their weight’ – the gossipers create pressure on the gossipee. That’s one interpretation. Mine comes from a work experience i had a few years back. I’d starting seeing one of the women in the office, which of course generated gossip. At one point i was approached by a few of the gossipers, who asked me, hands on hips style, “Are you sleeping with [Ms X]?” Not caring too much about their opinion, i spontaneously answered, “Yes, quite often actually”. This shut them up immediately and left them looking deflated, despondant, almost sad.

    That of course was not what they had wanted to hear or see. They wanted me to be defensive, to deny, or at least try to laugh it off. Not participating in the gossip in this manner had no negative consequences for me whatsoever. In no sense was it as though i had evaded social regulation, for which i should presumably pay a price. In fact the only longer term consequence is that no one in that office ever gossiped about me again, about anything. So i learned to ‘just say yes to gossip’. Admit to everything, even the exagerations – it kills gossip. In a different context, there is an example of ‘just saying yes’ in the this video http://bigthink.com/ideas/30766 (0:43).

    The gossiper(s) wants the gossipee to react to the gossip, not just to listen to it and take it on board. Gossip is an opportunity to show that you care. In the two cases above, the gossipee’s showed that they didn’t care, to their benefit and at the expense of the gossipers.

    So perhaps the problem with election trading – or gossiping by betting – is that the gossipee is in a no-win situation. There is no possibility of nulling the gossip because there is no reasonable equivalent to just saying yes. A major candidate can’t admit they are likely to lose or win – they have to keep up appearances at all times.

    • Stephen Diamond

      A major candidate can’t admit they are likely to lose or win – they have to keep up appearances at all times.

      1. But there’s no pressure against, say, journalists, who state–as at least one major one did–that Romney was a sure win.

      One difference that does exist is that journalists might be expected to present arguments; traders can present none (and as traders, might have negative incentive to). The “gossipee” can respond to an argument, but not in the same way to a market. But then, polling, too, is gossip, isn’t it? And traditional gossip isn’t necessarily accompanied by argument.

      In the end, Romney did have a way to respond to the gossip coming from polls and markets: doing some specious “polling.”

      2. I’m not sure it’s true that a major candidate must keep up appearances of it being a close race. This was true for Obama for specific reasons, not so true of Romney, who tried to create the impression that he would win handily, as did Romney himself.

      • Drewfus

        “But there’s no pressure against, say, journalists, who state–as at least one major one did–that Romney was a sure win.”

        Gossip is lassiez-faire. Who would want limits on what opportunities we can create that allow others to show they care about us? Interesting that we can create these opportunities purely verbally, or through writing, and that gossip is such a high proportion of total talk. Why such a huge demand to be shown care?

        “One difference that does exist is that journalists might be expected to present arguments; traders can present none”

        So one group – journalists – write their arguments down, the other group – traders – do not (usually). As a group, the traders are more accurate in their predictions. What does this suggest? Perhaps that writing stuff down coalesces the mind around certain ideas, that then remain relatively fixed. The original thoughts are in part due to chance associations, available information and internal feedback (the effects that our prior words have on subsequent ones). These fixed ideas can then become powerful motivators.

        Btw, if traders are not, as a group, any more accurate than journalists, then an alternative to prediction markets might be sophiticated word analysis of relevant journalistic output. Google is probably already working on it.

        “I’m not sure it’s true that a major candidate must keep up appearances of it being a close race.”

        Implicit rules are more powerful than explicit rules. There is of course no strict or even obvious penalty for breaking implicit rules, but paying respect to implicit rules is a good look and has positive consequences for ones social reputation. Socially intelligent folks are explicitly aware of implicit rules – other folks only implicitly. Socially intelligent folks respect or transgress implicit rules pragmatically – other folks live them. Explicit rules are applicable to low-status folks – and the visibility of each can therefore define low-status folks as such.

        “…Romney, who tried to create the impression that he would win handily”…

        … and lost. Your point that candidates do not have to keep up the appearance of a close race would have been much stronger if Romney had won (and assuming that he did brazenly transgress the implicit rule not to predict outcomes). Obviously there must be some ‘softness’ to the ‘keeping up appearances’ rule, otherwise a candidate would not even be allowed to sound confident of victory – the flexibility of implicit rules.

      • Drewfus

        “Interesting that we can create these opportunities [to show care] purely verbally, or through writing, and that gossip is such a high proportion of total talk.”

        My idea is that gossip only works when the gossipee responds ‘appropriately’ – when they are seen to be affected by the content of the gossip. So what about non-verbally manufactured opportunities to show care? I think one might be gift giving. What ‘counts’ with gift giving is how the recipient responds to the gift. A strong, positive reaction tells the giver that the recipient cares about the givers response to their reaction. Given that, it would not make much sense to buy something the recipient actually wanted, because then their reaction would probably be genuine. A genuine reaction would invalidate the recipients opportunity to show care for the giver – a waste of money. Instead, the giver wants to buy something that is of known low utility for the recipient, precisely so the recipient is forced into faking a positive reaction – or risk embarrasment and awkwardness if they do not bother making that effort.

        So all the economic analysis about gifts being an inefficient way to transfer utility are completely missing the point, and all the cultural criticism of the economic ‘scrooges’ and praise of “thoughts that count” is just a convenient cover for what is really going on with gift giving;

        The ceremony of gift giving is an opportunity for the recipient to show they care about the giver, by faking a positive emotional reaction to and gratitude for the givers gift. The faking is understood by both parties as being a fundamental part of the ceremony, as is the relative worthlessness of the gift to the recipient, which precludes the recipients reaction from being a genuine one.

  • Ansis Malins

    Jeez, just run the prediction market behind Tor with Bitcoin.

    • Robin Hanson

      If you think that is so easy, maybe you should do it.

    • Explodicle

      The problem there is that you need to trust that the anonymous prediction market admin won’t run away with everyone’s bitcoins, as has happened with several anonymous bitcoin sites already.

      Before bitcoin prediction markets can see mass adoption, we need to reduce this risk for average people. I’ve been watching two solutions:
      A) Lock up the coins in an m-of-n transaction, so multiple Open Transactions servers vote on where the funds go.
      B) User-made smart contracts within the blockchain itself. The person deciding “who won?” doesn’t need to be anonymous because there’s nothing illegal about signing a factual statement or releasing a private key. (IANAL)

      In the long term B is probably better, but in the short term A is almost ready. I’ve only heard about one guy actually working on B, and his primary focus is on Freicoin right now.

    • VV

      The legality of Bitcoin is already questionable, adding more illegal activities to the list of those which are performed using Bitcoin seems a sure strategy to get it banned.

  • Curt Adams

    There needs to be some more evidence that Intrade was actually interfering with somebody taking advantage of an uninformed public before you can say this was the reason. In politics Intrade can’t even match the information from public polling – you certainly can’t blame the ban on the political bets. I’m inclined to take the ban at face value – there really are rules for commodity trading in the US; they really get enforced; CFTC was just doing its job by demanding Intrade stop. It has very different motives from Congress, which banned the HSX, and whose job really is supposed to be managing conflicts between different interests (albeit it’s supposed to be more just than the HSX case).

    • Robin Hanson

      But Nadex did everything right according to the rules, and was still told no.

      • Curt Adams

        That’s not what the news says:

        From September 2007 to June 25, 2012, Intrade continued to offer options betting on future prices of gold, crude oil and changes in U.S. economic data after it had pledged to stop doing so under the 2005 order, the CFTC said in a lawsuit filed yesterday in federal court in Washington.

        5 years of a defying a court order is pretty serious.

      • Curt, that news story is about Intrade, not Nadex.

      • Curt Adams

        Sorry, lost track of the acronyms. I agree the NADEX decision (taken by Congress) was political and driven by makers of bad movies using Congress to protect unfair profits. I stick by my contention that the CFTC decision to ban Intrade commodities contracts (which led Intrade to shut down, presumably for policing costs) was *not* political but driven by reasonable long-standing laws they were bound to enforce.

      • Robin Hanson

        Curt, the Nadex decision was by the CFTC, not congress.

  • Stephen Diamond

    Anyway, Robin, condolences on the death of Intrade.

    • I smell ‘merica centrism in the air =) Intrade may be back for the US next year. Just “gambled” 25 on the US getting the debt limit raised by dec 31 =)

  • Who cares

    Having co-founded Nadex, designed the box office revenue products that were banned by congress, and worked closely with the CFTC on over 50 new product approvals for the past decade, i can assure you that the issue here is very simple: the CFTC staff sincerely want to do the right thing and are competent but the commissioners are politicians along with all that entails.

    Commissioners are politicians … they  do what politicians do, which is whatever they deem is most in their self interest AND that they believe they can get away with … its that simple.

    Until leadership positions at regulatory agencies are not political appointments, we should expect to suffer through more of the same nonsense … Its sad, but I can’t see this scenario changing anytime soon.

    • Robin Hanson

      This fits my experience. You can’t look at the words in the relevant regulations, go look up their definitions somewhere, and then from that predict these regulatory choices. You instead have to understand that the regulator’s job politically is to allow the kinds of trading that powerful businesses want, and prohibit the kids of trading that they expect the public would object to if they heard. So they make up rules and interpretations as needed to produce that outcome.

      • You can’t look at the words in the relevant regulations, go look up their definitions somewhere, and then from that predict these regulatory choices.

        Textualism versus purposivism is a live issue in jurisprudence. (See Scalia and Garner’s new book ( http://tinyurl.com/d3zv9ns ); see also my “Textualism = American Constitutionalism” ( http://tinyurl.com/d3qzvzg  )

      • So if big business starts to want prediction markets they will lobby for them, and we will profit from them?

  • Why do you need real money? Why not take advantage of the fact that the ability to predict accurately is a high-status skill?

    Have some billionaire create an Internet game. Each participant gets a certain initial wealth–or maybe even an income stream–composed of nominal money. The players who have accumulated the most faux dollars have their names in an ordered public list–kind of like LW’s high-karma list, except the ability to predict, unlike the ability to accumulate LW karma, is a highly transferable skill, worth it to become renown for. Longer though, so everyone has a place. Substitute status for money.

    (Probably a bad idea–I’m not good at this sort of thing.)

    • Robert Koslover

      “Have some billionaire…”  
      I gotta admit, that is an incredibly powerful tool!
      Armed with a billionaire who I could simple “have” do something I wanted, why… I’d have the purchasing power of a billionaire!!   This reminds me of an old reliable (and famous) method for becoming a millionaire:
      Step 1:  Go to the bank, open a savings account, and deposit $10,000.
      Step 2:  Deposit another $10,000, every week.
      Result:  In less than two years, you will have a million dollars!!
      Anyway, thanks, srdiamond, (and I do mean that sincerely) for adding a little light to my day.

      • Stephen Diamond

        Yeah, I probably underestimate the difficulties of garnering a billionaire. (I guess I figure that if Eliezer Yudkowsky can do it, anyone can.)

        Maybe the state should do it. A prediction market produces a public good independent of market demand. (It uses a market mechanism, which is another matter that I think people confuse with the other—see my exchange with Ilya). Perhaps “traders” (who are essentially gamblers) shouldn’t be counted on to find that socially valuable information is the information most interesting to bet on.

      • Creating such a game doesn’t cost that much resources. You could do it in your spare time at home. There’s no need for someone with a high amount of resources to get the project started. 

        Once you have your first version of the game you can go out and seek someone to pay you to expand the game. 

      • dmytryl

        You don’t need billionaire to create a game, but given that it would be a quite non interesting game overall (likely ranking below 5 out of 10 on any game review site at best and probably getting no mention at all), you may need to do a lot of advertising, at least buy the ads at the sites that do reviews so you get some score. For that you may need a billionaire.

  • Pingback: Government kills Intrade to save movie exec jobs | Leon Atkinson()

  • Pingback: Overcoming Bias : Combo Markets, Live!()

  • Isn’t politician’s don’t want the public to bet on results of political election quite similar to hollywood doesn’t want the public to bet on the results of film results?

  • Pingback: Overcoming Bias : Elite Evaluator Rents()