The Robin Hanson manipulation papers make unrealistic assumptions, but its not like prediction markets are a bad idea…!!…

No Gravatar

In terms of unrealistic assumptions in Robin Hanson&#8217-s series of papers on manipulation, the major ones have been out there since at least 2004.

Despite some limited evidence, the insistence on traders needing to know the direction of manipulation isn&#8217-t too compelling since the direction will be manifest insofar as the price is &#8220-wrong.&#8221- &#8220-Noise trader&#8221- is a politically loaded and misleading term. Misleading because it suggests that the mean effect will be zero, when in reality &#8220-noise trading&#8221- usually takes the form of feedback trading. Lack of feedback trading is a significant assumption in the Hanson manipulation papers. Fortunately, prediction markets have objective settlements at specified times, unlike traditional assets where the meaning of prices is open to interpretation, making them more prone to feedback trading and irrational booms and busts.

With prediction markets, conditions for manipulation are more favorable when the settlement is far off in time, and when there are subjective inputs to the settlement, e.g. in politics. A distant settlement simultaneously makes it less clear what the real price should be, and delays manipulator losses because there is less incentive to correct price. At the limit, a manipulator could introduce a price distortion when a contract is launched, only to reverse position for small liquidity-related loss immediately before settlement, thereby destroying the markets &#8220-integral&#8221- of error over time.

Another big assumption, also identified by Paul Hewitt, is that traders have equal account sizes. But maybe this isn&#8217-t a huge problem if settlement is forthcoming, and maybe the issue could be mitigated with additional exchange disclosures, such as the standard deviation of position sizes in a given market. While this could discourage liquidity as large traders would become paranoid about their positions, it is essentially a &#8220-soft&#8221- position limit, and traders would be forewarned of one-sided markets (which could of course be the result of someone well-informed, but I &#8211- the google-anonymous* writer &#8211- would bet that more concentration comes with more error on average&#8230- this can be tested by someone with the data, of course maintaining trader anonymity)

Even accounting for long-term settlement, feedback trading, semi-subjective settlements, and account size imbalances, it seems one would have to abide to an overly rigid tenet of &#8220-do no harm&#8221- to hold that prediction markets are, on net, a bad idea. (Do no harm is of course abhorrent to libertarians, and even doctors don&#8217-t actually follow such a rule.) Moreover, some pathologies like political self-fulfilling prophecy will only happen if prediction markets have already demonstrated their value and have become more popular. But even if one believes in their long term success, single pathologies can damage one&#8217-s reputation permanently&#8230- if one plans to die at a reasonable age.

[*Given the political climate, many firms have issued directives to employees to not engage in even the slightest appearance of impropriety, which might include blogging on manipulation.]

The New York Times on InTrades US political election prediction markets

No Gravatar

The NYT writers discusses 2 (different?) issues.

#1. There was market arbitrage opportunies in the recent past between InTrade and BetFair &#8212-unlike 4 years ago, and contrary to the laws of economics.

– The price of the Barack Obama event derivative was cheaper on InTrade than on BetFair and the Iowa Electronic Markets. Conversely, the price of the John McCain event derivative was more expensive on InTrade than on BetFair and the Iowa Electronic Markets.

#2. The NYT writer reports (without linking to it) the findings of the InTrade investigation about the behavior of their unnamed &#8220-institutional investor&#8221-.

– InTrade CEO John Delaney suggests that that institutional investor:

  1. might operate on InTrade at specific times where it might not be able to find liquidity on BetFair and/or IEM-
  2. might be a bookmaker willing to hedge its risks on a prediction exchange (a.k.a. betting exchange).

– Justin Wolfers&#8217- PHD student remarks that that institutional investor is not making an effort to shop around for the best prices, within each InTrade political prediction market.

RELATED: See the comments on Midas Oracle here, here, here, and here.

The blogger at Marginal Revolution misinforms the public by repeating the misinterpretation thrown around by liberal hack Paul Krugman about the alleged manipulation on the InTrade prediction markets.

No Gravatar

Alex Tabarrok writes that &#8220-someone was manipulating Intrade to boost John McCain&#8217-s stock price&#8221-.

No&#8230-!!!&#8230-

John Delaney said that that firm has been hedging on InTrade &#8212-a normal and beneficial activity on the other (larger and more liquid) financial markets.

InTrade is not liquid enough to weather (quickly enough) the impact made by the hedging activities, at this time, but will in the future, if growth continues.

Manipulation is bad.

Hedging is good.

InTrade offers an explanation of strange trading.

No Gravatar

Intrade has made a statement on the unusual trading that many have noted and alleged to be manipulative. The statement suggests that the price action is mostly attributable to a single firm, a hedger &#8220-using our markets in good faith and in the ordinary course of their business.&#8221-

The first company that comes to mind is Centrist Messenger. Centrist is an interesting firm that re-sells political ad time and refunds sales to customers whose candidate loses. Centrist has stated publicly that it uses Intrade to hedge this exposure.* If Centrist had something to do with the unusual trading, it suggests that they sold more Obama than McCain ads, creating exposure to a GOP victory, resulting in McCain buys and Obama sales on Intrade. Why such a firm would be such urgent price-takers isn&#8217-t fully explained.

Whether or not it was Centrist isn&#8217-t important, but as these markets mature we should expect them to attract more hedging activity, and this might introduce persistent price distortions. Indeed it makes sense for people in the top tax bracket to be long Obama apart from considerations of his chances of victory. This is another uncomfortable subject that I&#8217-ve warned about in the past. When these markets become deeper and more widely available, the odds of the high-tax candidates might begin to show an upwards bias, a risk premium. Interestingly, Musto and Yilmaz predict that such markets will eventually lead to increased promises of redistribution by candidates. Talk about unintended consequences.

Intrade is doing the right thing here though, dealing with tough issues realistically and with as much transparency as possible. They provide valuable information, for free, even in places where they are not necessarily welcome. The depth of this information helps us to evaluate Intrade prices and have more confidence in them. Here is an example below, based on Obama&#8217-s market over the past two weeks. Some have noted that the purported attacks occurred in hours where the market was unusually thin. This chart measures such price manipulability. The red line represents the ease of a downwards attack. It is the 100 x the amount of margin required to sweep the top fifteen bids divided by the difference between the highest bid and the fifteenth highest bid. (That is, how much the probability of an Obama victory can be moved by risking $100. Commissions are not taken into account but would of course would be vital.) The green line is the ease of an upwards attack. This is a very preliminary study and I will leave it to others to voice initial impressions. The fact that we can gauge to what extent traders are exercising market power is in itself important and encouraging however.

* Technically another firm does the trading. Centrist is incorporated in the US, and the trading firm is incorporated in St. Kitts. Through this arrangement, Centrist cleverly avoids violating UIGEA.

[Cross-posted from Risk Markets and Politics ]

As Justin Wolfers noted, maybe there are today bigger practical obstacles to prediction market arbitrage.

No Gravatar

Legal restrictions for US traders on foreign prediction exchanges (BetFair, etc.)-

Transaction fees (you would need to operate on 2 exchanges)-

Currency risks and cost for hedging on that.

Eric Crampton (a Canadian exiled in New Zealand) says he has managed to turn a buck, though, by arbitraging between InTrade and iPredict New Zealand. He also makes 2 theoretical points. Go read it.

Is InTrade being manipulated? Why does InTrade give a discounted probability for Barack Obama as US president?

No Gravatar

A quick link panorama.

#1. Is InTrade being manipulated?

– Nate Silver shows that there are abrupt downward pressures on the Barack Obama event derivative, while we also see some abrupt upward pressures on the Hillary Clinton event derivative.

However, you can see by yourself that InTrade is resilient enough and does a great job of going back to normal [*], after just a few hours of trading:

– At Portfolio, blogger Zubin Jelveh blows the incidents out of proportion.

– Professor Lance Fortnow has a more careful analysis and notes that the price of the Barack Obama bounces back quickly enough.

– Quick thought: Maybe the media should use an average of event derivate prices for the last 5 work days&#8230- so that the abrupt perturbations would be eliminated.

[*] UPDATE:

Professor Eric Zitzewitz:

I’m not sure you can conclude from Silver’s graphs that the market goes “back to normal.” You can conclude that it moves back in the opposite direction of the impact those large trades. Back when the Hillary for President market looked like it was being manipulated, it appeared that the manipulator was both placing a large purchase and then placing limit orders to provide price support and slow down the reversion of the price.

UPDATE: Are we witnessing manipulation attempts on the &#8220-Florida to vote Republican&#8221- prediction market at InTrade?

#2. Why does InTrade give a discounted probability for Barack Obama as US president?

– As you remember, Emile Servan-Schreiber of NewsFutures believes that it&#8217-s a Republican conspiracy all over.

– Professor Justin Wolfers puts up an hypothesis: it&#8217-s legally impossible for US traders to arbitrage on BetFair.

– InTrade put up a crappy excuse: the industry is still too &#8220-young&#8221-. How lame. How stupid. The industry was younger in the previous elections, where arbitrage opportunities didn&#8217-t exist according to professors Justin Wolfers and Eric Zitzewitz (see their 2004 paper and their other publications).

– Blogger Zubin Jelveh swallows the InTrade P.R. line, and adds another crappy InTrade P.R. line: More arbitrage opportunities are being exposed in open air because much more observers are hunting down arbitrage opportunities in 2008 than in previous elections. That&#8217-s a second blatant cretinery, uncorrected by the Portfolio blogger. Re-read Justin Wolfers&#8217- blog post. Professor Justin Wolfers states that:

The current variation in price is larger than I have ever seen in my years of studying prediction markets. The forces of arbitrage that would typically eliminate these differences have been handicapped by the legal restrictions preventing U.S.-based traders from using overseas markets.

– Finally, professor Lance Fortnow says nothing about the arbitrage opportunities between InTrade and BetFair, but does offer some technical points about the issue of polls versus the prediction markets, centered around the question of state correlations. Read on.

UPDATE: Eric Crampton (a Canadian exiled in New Zealand) says he has managed to turn a buck by arbitraging between InTrade and iPredict New Zealand. He also makes 2 theoretical points. Go read it.

UPDATE: Greg Mankiw just linked to Nate Silver.

The Giuliani manipulator buyer is back

No Gravatar

Judging from his hours, he&#8217-s based in the US.

We see you on the bid again, and your cosmetic offers as well. You are the whale in this market, but it&#8217-s a small pond. Be careful.

More details on this strange trading later&#8230-

Addendum:

Exhibit A: a view of trading known as &#8220-market profile&#8221- from March 1st through the 7th. Price is on the y-axis and volume is on the x-axis, instead of time. What typically develops on these charts are sideways normal-distribution-like patterns, which is unsurprising by the central limit theorem. Often, a jump to a new mean corresponds to an event. The pattern below is unheard of in liquid markets, except in risk-arb and other &#8220-peg&#8221–ish situations.

Giuliani Volume@Price 3/1/07-3/7/07

What first comes to mind is that the exchange is manufacturing volume with bogus &#8220-wash&#8221- trades, but the first time 33.3 printed (which is where half of the volume for March occurred as of yesterday), the price had been in the teens, and 33.3 marked an all-time high for the contract. This doesn&#8217-t make sense as fake volume nor some sort of internal initialization trade relating to TEN&#8217-s restructuring.

Yesterday&#8217-s trading suggests that a single buyer is pushing the market up and is currently successfully holding it at 40 while posting offers to appear as a passive market-maker. 1-2000 buy orders remained near 40 until about 10pm EST yesterday and returned this morning, EST. In these thin markets, this is quite a lot especially considering the high price level &#8212- and it is high since it&#8217-s almost a year before the first primary. Of course the recent buyer might be unrelated to whoever caused the anomaly at 33.3. To be continued..

Addendum:

After taking the weekend off, our buyer was back by 10am EDT this morning. He has to defend 33.3 which shouldn&#8217-t be too difficult considering that there are only 3 major candidates. Truth be told, 40 isn&#8217-t that high for this contract, and the price does more-or-less reflect recent polls, but this guy is awfully confident. There is a fine line between a manipulator and an overconfident trader who is too large for the market.