prediction markets

All posts tagged prediction markets

I have implemented subsidies to encourage trading of some conditional prediction market contracts that may provide useful information about the consequences of the 2008 presidential election, via a simple automated market maker (using an algorithm described near the end of http://hanson.gmu.edu/ifextropy.html). The subsidized market maker ought to provide incentives for traders to devote more thought to these contracts than they would if the liquidity was less predictable.
Intrade has agreed not to charge any trading or expiry fees on these contracts.
Some places to look for extensive description of the motivations behind these subsidies are here and here.

The contracts are:

Please read the detailed specifications at Intrade before trading them, as one-line descriptions are not sufficient for you to fully understand them.
For the first two of those contracts, the market maker will enter bids and asks of 38 contracts, and can lose a maximum of $5187.76 on each contract. For the other four contracts, the market maker will enter bids and asks of 115 contracts, and can lose a maximum of $7906.25 on each contract.
I will maintain a web page here devoted to these contracts.
See also this more eloquent description on Overcoming Bias.

Book review: Information Markets: A New Way of Making Decisions, edited by Robert Hahn and Paul Tetlock
This book contains some good discussions of current issues in the design of prediction markets (aka idea futures).
Since it’s the result of a conference for experts, it is mainly directed toward experts. It shouldn’t be overly hard for laymen to understand, but it probably focuses on issues that are somewhat different from what most laymen would find interesting, so I’d probably recommend reading Surowiecki’s Wisdom of Crowds or some of Robin Hanson’s earlier papers on the subject first.
One surprising result reported here is that the Iowa Electronic Markets show no longshot bias, in contrast to similar markets on Tradesports/Intrade and to widespread types of sports betting. This looks like an important area for research, although that would probably require setting up many variations on those markets (varying things such as the user interface, commissions on trades, limits on how much money can be invested, etc.), which would be expensive and hindered by regulatory uncertainty.
Michael Abramowicz presents an interesting proposal to create incentives to counteract the likely tendency of markets such as prediction markets to discourage people from making public the knowledge that goes into making market prices efficient. I don’t have much of a guess about how well his solution will work. It needs some more thought about how vulnerable it is to manipulation of the intermediate prices used to reward traders who convince others to follow their reasoning (averaging prices over a week or two would be a simple start at deterring manipulation). But I think he understates the importance of the problems he’s trying to solve. He says “while they are endemic to all securities markets, they apparently cause little harm. They are likely to be much more severe, however, in markets with very few active participants.”. I suspect they are significant in most securities markets, and are underestimated because they are very hard to measure. As someone who trades stocks for a living, I’d say that the amount and quality of knowledge that is shared among traders is quite low compared to most professions, although it’s hard to say how much of this is due to desire to keep valuable information secret and how much is due to the difficulty of distinguishing valuable information from misleading information.

This book does an excellent job of reporting important evidence showing that group decisions can be wiser than those of any one individual. He makes some good attempts to describe what conditions cause groups to be wiser than individuals, but when he goes beyond reporting academic research, the quality of the book declines. He exaggerates enough to give critics excuses to reject the valuable parts of the book.
He lists four conditions that he claims determine whether groups are wiser than their individual members. I’m uncertain whether the conditions he lists are sufficient. I would have added something explicit about the need to minimize biases. It’s unclear whether that condition follows from his independence condition, partly because he’s a bit vague about whether he uses independence in the strong sense that statisticians do or whether he’s speaking more colloquially.
Sometimes he ignores those conditions and makes unconvincing blanket statements that larger groups will produce wiser decisions.
He makes exaggerated claims for the idea that crowds are wise due to information possessed by lots of average people rather than the influence of a few wise people. For instance, he disputes a Forsythe et al. paper which argues that a small number of “marginal traders” in a market to predict the 1988 presidential vote were responsible for the price accuracy. Surowiecki’s rejection of this argument depends on a claim that “two investors with the same amount of capital have the same influence on market prices”. But that looks false. For example, if the nonmarginal traders make all their trades on the first day and then blindly hold for a year, and the marginal traders trade with each other over that year in response to new information, prices on most days will be determined by the marginal traders.
It’s not designed to be an investment advice book, but if judged solely as a book on investment, I’d say it ranks in the top ten. It does a very good job of explaining both what’s right and what’s wrong with the random walk theory of the stock market.
He does a good job of ridiculing the “cult of the CEO” whereby most of a company’s value is attributed to its CEO (at least in the U.S.). I was surprised by his report that 95% of investors said they would buy stocks based on their opinion of the CEO. They certainly didn’t get that attitude from successful investors (who seem to do that only in rare cases where they are able to talk at length with the CEO). But his claim that “Corporate profit margins did not increase over the course of the 1990s, even as executive compensation was soaring” looks false, as well as being of questionable relevance to his points about executives being overvalued. And I wish he had also applied his argument to beliefs of the form “if we could just elect a good person to lead the nation”.
Chapter 6 does a good job of combining the best ideas from Wright’s book Nonzero and Fukuyama’s Trust (oddly, he doesn’t cite Trust).
He exaggerates reports that the stock market responded accurately to the Challenger explosion before any public reports indicated the cause. He claims “within a half hour of the shuttle blowing up, the stock market knew what company was responsible.” I don’t know where he gets the “half hour” time period. The paper he cites as the source says the market “pinpointed” Thiokol as the culprit “within an hour”, but it exaggerates a bit. If the percent decline in stock price is the best criterion, then the market provided strong evidence within an hour. If the dollar value of the loss of market capitalization is the best criterion, then the evidence was weak after one hour but strong within four hours.
He also claims “Savvy insiders alone did not cause that first-day drop in Thiokol’s price.”, but shows no sign that he could know whether this is true. He seems to base on the absence of reported selling by executives whom the law requires to report such selling, but he appears to overestimate how reliably that law is obeyed, and to ignore a large number of non-executive insiders (e.g. engineers). He does pass on a nice quote which better illustrates our understanding of these issues: “While markets appear to work in practice, we are not sure how they work in theory.”

Chris Hibbert writes (in a post that is partly about the mess resulting from Tradesports’ contract on North Korean missile launches):

The fact that pay-outs are limited to the amount spent to purchase claims is integral to the institution of prediction markets. If market operators ever pay off both sides of a claim, that is likely to encourage investors to protest many more close calls.

I disagree. Having pay-outs equal to claim purchases is integral to the normal function of well-written claims, but there’s little reason to stick to that rule with a claim written as poorly as the North Korean missile claim was.
Paying off both sides was the most reasonable suggestion I’ve heard for what Tradesports should have done to limit the damage to their reputation. Experience with similar disputes (such as those on FX) suggests that traders already have sufficient motive to protest questionable decisions that it’s hard to see how disputes produced only by additional incentives could bear much resemblance to reasonable disputes. The increased incentive on Tradesports to word their claims so that fewer people misunderstand how they will be judged is likely to have some desirable effects on how Tradesports explains the meaning of their contracts.
Disputed judgments might be inevitable for exchanges that cover subjects as ambitious as Tradesports does, but there’s nothing inevitable about confusion about whether a contract was about DoD confirmation of where the missiles landed, or whether it was about what the missiles did, with DoD statements merely being used if needed to resolve any uncertainty.
(I didn’t trade any of the North Korean missile contracts).

I’ve been slowly working my way through a book by Richard Zacks called An Underground Education. I’ve found one section that deserves a blog entry of it’s own (I’ll discuss the rest of the book when I’ve finished reading it).
It describes a fairly popular betting-style market that ran from 1771 to 1776 in London about whether a diplomat named Chevalier D’Eon was male or female. D’Eon apparently acted and dressed at times as a man, and at other times as a woman, and refused to help the bettors settle their bets (even when D’Eon was offered a large amount of money to provide evidence of his/her sex). Eventually bettors got tired of waiting for an outcome and resorted to at least one lawsuit. The judge decided that D’Eon was a woman based on testimony from both sides of the lawsuit. Why did the side who had bet D’Eon was a man produce testimony that D’Eon was a woman? It was part of an accusation that the other side traded on what the SEC would call inside information. After D’Eon’s death, mortuary attendants said D’Eon was a man.
Aside from the obvious implications for how idea future style markets need to word claims so as to assure a practical means of observing whether a claim is true or not within a practical time period, this report also says some odd things about gender stereotyping. Lots of people probably think prior generations mostly had Victorian attitudes toward gender confusion, but it seems that D’Eon was sufficiently respectable in 1792 to have a dinner party thrown to honor both Thomas Paine and D’Eon.

Here are a few comments from Friday’s Prediction Markets Summit.
Chris Hibbert described a way that a market with multiple outcomes (such as for supreme court nominees, which list contracts for a number of people, plus one for the rest of the field) could improve liquidity with a modest software change. The system could generate bids and asks for a given contract by aggregating the opposite side of all the other contracts in that market, and generate a synthetic order which would sometimes be within the bid/ask range of regular orders.
Google’s Bo Cowgill reported that one legal problem that Google faces in implementing internal prediction markets is that they might sometimes spread information around the company that might make the people with that information insiders in a way that would complicate those employees’ ability to trade in Google stock. It sounds like the insider trading laws are onerous enough to create a nontrivial barrier to spreading information.
Newsfutures’ Emile Servan-Schreiber said that Newsfutures was supporting the use of internal corporate markets for some fairly big corporate decisions.
Mike Knesevitch of Intrade / Tradesports seems to have lots of experience in traditional financial markets. He said a good deal about the liquidity and accuracy of his exchange’s political markets. The Howard Dean contract went from 33 to 9 within about 5 minutes following Dean’s “implosion” speech. Part of Tradesports’ reluctance to make it’s prices easy to link to is the possibility that they will charge money for those prices much like traditional stock and futures exchanges do. I wish there were a good way for users to persuade new exchanges to commit to keeping information free, but I don’t see a practical way to do that.
HedgeStreet’s Russell Andersson reported that CFTC approval for new contracts was easier than I expected. It takes about 24 hours for contracts dealing with things that the CFTC normally deals with, but weeks to months for other topics. (Why so much variation??)
Microsoft Todd Proebsting is surprisingly enthusiastic about most (all?) of Robin Hanson’s vision of what prediction markets might do. He reported a case where a market did a much better job of predicting when a product would ship than the manager did, although there was some confusion when the market fluctuated when it’s prediction created some uncertainty over whether features would be cut to make the deadline (the contract didn’t adequately specify how it would be judged if the product changed significantly). He reported that there was no resistance to prediction markets from upper management (middle management resisted, since the markets are designed to indicate failings of middle management). He mentioned that laws regulating sweepstakes created some obstacles to implementing internal corporate markets (they’re designed to prevent a sweepstakes game from rewarding the people who control the sweepstakes). He inadvertently(?) promoted the use of open source licenses by mentioning that legal concerns deterred him from looking at Robin’s market scoring Lisp code, which has no license granting any permission to use it.
Eric Zitzewitz responded to Manski’s theoretical criticisms of prediction market accuracy (see his paper on Interpreting Prediction Market Prices as Probabilities), and described some problems with inferring causality from markets and how to structure markets to minimize those problems (see his paper on Five Open Questions About Prediction Markets). He showed an amusing graph indicating that Tradesports prices implied Osama was twice as likely to be captured in October 2004 as in November 2004 (implying some connection with the U.S. elections).

The CFTC has reacted to Tradesports‘ futures-like contracts that many U.S. residents have been trading without CFTC regulation.
It is surprising how closely the contracts that they objected to coincide with contracts traded under CFTC regulation – they apparently have prohibited Tradesports from offering to U.S. residents contracts on the results of the next Fed meeting (which Hedgestreet trades under CFTC regulation; Tradesports stopped offering these in May, possibly due to negotiations with the CFTC) but as far as I can tell Tradesports is still able to offer contracts on where the Fed Funds rate will be at the end of the year.
I am also surprised that the CFTC classified the contracts as futures options rather than futures. They do have something resembling as strike price, but otherwise resemble a futures contract more than they resemble an option.

Two years ago a DARPA project was canceled after some demagogues attacked a straw man which bore a superficial resemblance to the actual project. Now Robin Hanson (who had some involvement with the project) has written a defense of the straw man, i.e. an argument that futures markets might be of some value a predicting specific features of terrorist attacks (although not nearly as valuable as more natural uses of futures markets such as predicting the effects of changes in Homeland Security budgets on the harm done by terrorism).
He has a somewhat plausible argument that there is useful information out there that might be elicited by markets, particularly concerning the terrorist choice of method and targets. An important part of his argument is that in order to be useful, the markets might only need to distinguish one-in-a-thousand risks from one-in-a-million risks. One weakness in this argument is that it makes mildly optimistic assumptions about how reasonably people will respond to the information. There is clear evidence much spending that is advertised as defense against terrorism is spent on pork instead. Markets that provide a few bits of information about which targets need defending will raise the cost of that pork-barrel spending, but I can’t tell whether the effect will be enough to meet whatever threshold is needed to have some effect.
The section on moral hazard seems to contain a rather strange assumption about the default level of trader anonymity. The “reduced” level he talks about seems to be about as much as the U.S. government would allow. It isn’t clear to me whether any anonymity helps make the prices more informative (does anyone know of empirical tests of this?). The optimal level of anonymity might vary from issue to issue according to what kind of trader has the best information.
The proposal to hide some prices is more difficult than it sounds (not to mention that it’s far from clear that the problems it would solve are real). Not only would the exchange need to delay notifying traders of the relevant trades, but it would need to delay notifying them of how the trades affected the traders’ cash/credit available for trading other futures. Which would often deter traders from trying to trade when prices are hidden (it’s also unclear whether the trades that would be deterred would add useful information). In addition, I expect many of the futures that would be traded would be about targets and/or methods covering some broad range of time; it’s unclear how to apply a condition about “attacks to occur within the next week” to those.
The proposals to deal with decision selection bias sound politically difficult to implement (unless maybe Futarchy has been substantially implemented). But there isn’t much risk to experimenting with them, and elected officials probably don’t have the attention span to understand the problem, so there probably isn’t much reason to worry about this.

In the latest issue of Econ Journal Watch, Bryan Caplan and Donald Wittman hold an inconclusive debate on whether democracy produces results that are sensibly related to voters’ interests. They come much closer than most such discussions to using the right criteria for answering that question.
But they fail because they implicitly assume that inaccuracies in voters’ beliefs are random mistakes. If that were the case, Wittman’s replies to Caplan would convince me that Caplan’s evidence of voter irrationality is as weak as the arguments that consumer irrationality prevents markets from working, and that Wittman’s proposed experiments might tell us a good deal about how well democracy works.
On the other hand, if you ask whether voters have incentives to hold beliefs that differ from the truth in nonrandom ways, you will see a fairly strong argument that voters’ inadequate incentive to hold accurate beliefs causes systematic problems with democracy.
Imagine that you live near a steel mill. This means that believing that steel import restrictions are bad will increase the risk that your acquaintances will dislike you (because you views endanger their jobs or their friends’ jobs), and will probably bias you toward supporting protectionism.
Or take the issue of how gun control affects crime rates. There are some obvious patterns of beliefs about this which the random-mistake hypothesis fails to predict. Whereas the theory that people adopt beliefs in order to indicate that they think like their friends and neighbors (combined with some regional variations in gun ownership that created some bias before people started thinking about the issue) does a much better job of predicting the observed patterns of belief.
Because this seems to be a widespread problem with democracy, I’m fairly certain democracy works poorly compared to markets and compared to forms of government such as Futarchy which improve the incentives for policies to be based on accurate beliefs.

For a long time I’ve thought that the weakest part of the argument for Futarchy is the problem of choosing a good measure of national welfare. Democracy is sufficiently subject to manipulation by demagogues (e.g. convincing voters that Saddam had some responsibility for Sept. 11 or that confiscating guns will make us safer) that turning politicized disputes about factual questions over to an institution that maximizes GDP would probably be an improvement even if the flaws in using GDP cause it to do foolish things like preferring Microsoft’s commercial crap to free alternatives. But it would be hard to convince people addicted to having democratic processes decide questions of fact to ignore such flaws.
I want to propose that such a system should be designed to maximize life expectancy instead. That measure seems to correlate fairly well with a number of things we value such as wealth and happiness. I’m not sure how it correlates with equality, and suspect it is imperfect at putting the optimal weight on increasing that, so I’m not claiming it’s a utopian solution. But I doubt there’s much risk that maximizing lifespans would increase inequality.
It would create pressures to keep peoples’ hearts beating when they’re brain-dead, and to put undesirable restrictions on activities such as skiing and rock-climbing. But it’s not obvious why that would be significantly different from the biases of existing governments.