Prediction Markets = Clear Expiry + Disperse Information + Participation Incentives

No Gravatar

Jed Christiansen at Forbes (just after John Delaney&#8217-s ill-written and pointless comment):

A market effectively aggregates the information from everyone participating. So anything where:

  • there is a clear result
  • information is dispersed between people and/or locations
  • people have an incentive to participate in the market

will likely provide better results than any other forecasting method. Experts just aren&#8217-t as good as they (or anyone else) think they are. It&#8217-s simply better to ask the crowd in these cases.

Missing from Jed Christiansen&#8217-s comment is the emphasis on long series for comparison. Takes time and hundreds of prediction markets to prove the wisdom of crowds.

&#8212-

UPDATE: Jed Christiansen comments&#8230-

Chris, I agree that for probability assessment, a number of measurements are required to assess success. However, for metrics (ie, sales of widget X, rating of product Y) it doesn&#8217-t require a long series at all. Depending on how poor the current forecasting model is performing, a prediction market could prove successful after just a few measurements.

ads, auctions and markets: a proposal

No Gravatar

Ads make the web go round. US internet advertisement expenditure is expected to surpass $21 billions in 2007, while the share of web ads in the UK is already bigger than ad spending in radio and newspapers. More specifically, adwords (+adsense) is maybe the single reason behind growth and sustainability of the biggest business fairytale in our times, returning for example $2.6 (+$1.0) billions in 2006Q3 alone.

Textual ads are auction powered. Speaking of Google, let me repeat the well-known process (which is valid with some variations also in Yahoo&#8217-s Search Engine Marketing and Microsoft&#8217-s ad center -forgive my partial knowledge). Each time you view or click on an ad link, the advertiser rewards the site owner for the use of her screen real estate and the redirection of your attention span. More accessible real estate costs a lot, as the more the advertisers pay the higher their ad is ranked and exposing their message to more visitors. Ad ranking occurs mainly from a dynamic Vickrey auction model, namely cost-per-click bids is the most influential factor defining your ads rank. But, to my eyes, the system&#8217-s current status seems sub-optimal. From the ad platform engine&#8217-s perspective, earnings aren&#8217-t maximized, as the advertiser pays only for actual clicks and not for impressions. From the advertiser&#8217-s view, the ad&#8217-s impact is also sub-optimal, as high rank doesn&#8217-t guarantee more visitors, in the case of low ad relevancy to the user&#8217-s query and actual interests. Finally, and most importantly, the auction-based fundamental model doesn&#8217-t accumulate the collective intelligence of the previous visitors&#8217- behavior and probably results in a poor user experience. (Well, in practice, clickthrough rates are also evaluated and the adrank algorithm is much more complicated, but this doesn&#8217-t reduce the validity of comments on the fundamental choice of an auction model).

While the auction-based approach apparently works, let me propose a more simple, direct and transparent market-based variation. In such a case, the ad-space of the universe of all potential keywords combination consists a gigantic marketplace, while each keywords&#8217- combination will form a market in this marketplace. In each market, an advertiser&#8217-s submission triggers the creation of a stock, which is initially traded at the defined cost-per-click price. But this price is nothing but constant- it goes up each time a user clicks on the ad, or down each time a user clicks on a different ad (this second action could even be omitted). As a result, stock prices, therefore ad ranking, evolves dynamically to enhance previous visitors&#8217- choices and leverage the wisdom of crowds in forming an elegant user experience. Moreover, it maximizes engine&#8217-s gainings and advertisers&#8217- impact, while enables a fully trackable procedure for advantage of both the platform and its users.

I would like to stress my lack of extended or insiders&#8217- knowledge on the topic, but I&#8217-m more than eager to discover relevant attempts, or -even better- participate in some attempt to put this idea into reality.

cross-posted from my blog

The BetFair Starting Prices… explained

No Gravatar

The Guardian (which has the best coverage on BetFair-TradeFair):

The Betfair betting exchange will launch its starting price betting service in the middle of next month, after it received the approval of a number of the exchange&#8217-s biggest-staking clients at a series of demonstrations over the last seven days.

Betfair has spent two years developing a robust system to allow bets to be placed at its own SP. As a result, three extra columns will soon be added to its display for every win market on its British and Irish racing service: one for bets on a horse at the Betfair SP, one for &#8220-lay&#8221- bets against it at SP, and a &#8220-guide&#8221- price in the middle showing what the SP is likely to be.

The SP market and the normal exchange market will be separate, in an attempt to ensure that it is extremely difficult to manipulate the SP. Betfair also hopes to offset any reduction in overall liquidity – which is vital to any exchange – by attracting new clients.

The great majority of off-course bets are still settled at the official SP, derived from on-course betting markets, while even on the internet, around 55% of racing bets are settled at SP. The Betfair SP, derived from a market with a negligible profit margin, will allow punters to make a direct comparison between the bookmakers&#8217- odds and those on offer on the exchange.

SP bets will have some differences from &#8220-normal&#8221- bets on Betfair. Whereas normal bets can be cancelled until the moment when a rival gambler accepts the other side of the bet, SP bets cannot be cancelled once placed. Punters can, however, specify a minimum price at which they are willing to back, and a maximum price for lay bets.

In time, SP betting is likely to extend to the exchange&#8217-s place-only markets, and to many other sports in which the SP concept has never been used before.

&#8220-This has the potential to be the most significant step forward for Betfair since the exchange was launched,&#8221- Stephen Burn, Betfair&#8217-s director of horse racing, said yesterday. &#8220-Racing is scratching the surface. In time, we hope that we will be able to add it across football, tennis and so on. We could even return an SP on the next general election.&#8221-

Niall O&#8217-Connor has more.

NEXT: More on BetFair&#8217-s Starting Prices.

Austan Goolsbee on Iraq and the Collective Wisdom of Bond Markets

No Gravatar

Austan Goolsbee, writing in the New York Times, discusses Michael Greenstone&#8217-s paper (discussed here at Midas Oracle in September) that examines the market for Iraq&#8217-s bonds for an assessment of the long-term future of the Iraq government. Goolsbee&#8217-s quick conclusion: &#8220-But global financial markets have been monitoring the war for months, and with remarkable consistency, they have concluded that the long-term prospects for a stable Iraq are very bleak.&#8221-

It wasn’t until Professor Greenstone began examining the financial markets’ pricing of Iraqi government debt that he had his eureka moment. It was immediately clear that the bond market — which, historically, has often been an early indicator of the demise of a political system — was pessimistic about the Iraqi government’s chances for survival.

First, some background &#8230- the Iraqi government issued about $3 billion of new bonds in January 2006. These dollar-denominated bonds pay 2.9 percent twice a year and mature in 2028, paying the face value of $100.

To say the least, the market for these bonds is not robust: as of last week, a bond with a face value of $100 was trading at around $60. Professor Greenstone calculated that, from the markets’ standpoint, the implied default risk over the life of the bond was about 80 percent.

The important point is that anyone who owns one of these Iraqi bonds has to decide each day whether the Iraqi government is likely to be functional enough to make its debt payments, or will default along the way. All else being equal, if the surge policy is effective, it ought to be raising the market price of these bonds.

Bondholders “aren’t politically motivated,” Professor Greenstone said. “They don’t have to rationalize their previous statements or justify their votes from years past. All they care about is whether there will be a functioning Iraq in the future such that they will receive their payments.” At a certain price, most securities will find a buyer, and there are still buyers for Iraqi bonds. But the price they are willing to pay is very low.

Goolsbee tosses in a few examples which show, in his words, &#8221- the collective wisdom of financial markets has proved remarkably adept at evaluating events and predicting the future, even the turning points of war&#8220-:

During the American Civil War, for example, when Confederate forces lost at Gettysburg, Confederate cotton bonds traded in England dropped by about 14 percent. During World War II, German government bonds fell 7 percent when the Russians started their counterattack at Stalingrad in 1942, and French government bonds rose 16 percent after the Allied invasion at Normandy in 1944. Many such examples of the prescience of financial markets have been documented by economic historians.

Of course a few cherry picked examples, while suggestive, should not be considered conclusive.

Chris Masse, in a post about negative comments on the war by a just-retired high ranking military officer, said:

We can’t rely on retirees to tell us the truth. We need an anonymous information aggregation mechanism that gives an incentive to people who come forward with advanced information: the prediction markets.

While bond markets might be useful as a stand in for prediction markets, presumably well-designed prediction markets could provide a somewhat more articulated position than can be extracted from a twenty-year bond market.

NOTE: Greenstone&#8217-s paper, &#8220-Is the &#8216-Surge Working? Some New Facts,&#8221- is available from the SSRN.

Robert Schillers MacroShares = Flawed Product??

Greg Newton:

The MacroShares are irretrievably broken. They have never performed as advertised. They show no signs of ever working as advertised. They are a disgrace to the ETF market, and have been, pretty much since they were introduced.

External Link: MACROshares


Author Profile&nbsp-Editor and Publisher of Midas Oracle .ORG .NET .COM &#8212- Chris Masse&#8217-s mugshot &#8212- Contact Chris Masse &#8212- Chris Masse&#8217-s LinkedIn profile &#8212- Chris Masse&#8217-s FaceBook profile &#8212- Chris Masse&#8217-s Google profile &#8212- Sophia-Antipolis, France, E.U. Read more from this author&#8230-


Read the previous blog posts by Chris. F. Masse:

  • Comments are now completely open on Midas Oracle.
  • Albert Einstein, Chairman of the Midas Oracle Advisory Board
  • Erratic –but not Stochastic– Charts
  • Barack Obama is the 44th US president.
  • We already have prediction markets in future tax rates. It’s called the municipal bond yield curve.
  • DELEGATES AND SUPERDELEGATES ACCOUNTANCY
  • O’Reilly – Money-Tech Conference

How to sell art short

No GravatarFor a while it seemed like a month wouldn&#8217-t pass without hearing of a new record-breaking art auction, along with the inevitable insinuations of a &#8220-bubble&#8221-. Last night the buyers blinked, as a Van Gogh work expected to command $28-35 million did not get any bids.

People talk about art as an asset class and yet there is no way to sell it short or hedge against declines in value. Clearly, prediction markets are an answer. It seems like this would be a good niche market for a real-money exchange. Markets could either be tied to upcoming auctions or, more likely, an art price index. The latter would allow any art owner a hedge with basis risk.

This, by the way, is not to imply that art prices are about to collapse. There is some evidence that they trail housing prices with a lag of a year or two, but this is mainly anecdotal to the late &#8217-80s / early &#8217-90s period. It is more likely that the location of demand is shifting.

Exchanges that are fortunate enough to be operating in modern legal and regulatory regimes show a somewhat limited imagination in their offerings. There are opportunities in the current re-shaping of how art is priced and artists are rewarded.

Previous blog posts by Jason Ruspini:

  • 2009 tax futures yielding 1.5%
  • Intrade, with carry
  • Talking tax futures on BNN, Canada’s business channel
  • Tax Futures, “In Real Life”
  • YooNew, fears and hopes

If Musharraf goes, should we celebrate?

No Gravatar

Benazir Bhutto made a compelling case in the Sunday New York Times that she would be a better partner for the West than Musharraf. As she noted, $10 billion in U.S. aid since 2001 has not resulted in the elimination of Al Qaeda&#8217-s base there. There seems to be substantial reason to believe that Pakistan and its intelligence service are part of the problem, not the solution.

But would Musharraf&#8217-s fall be good news? It&#8217-s not obvious, since we don&#8217-t know who will replace him (i.e., Bhutto or someone more anti-Western). And even if it&#8217-s Bhutto, we don&#8217-t know how much help she&#8217-ll be once in office.

Intrade has just listed a contract that may help provide an answer. In addition to adding Musharraf to their foreign leader departure series (as Chris noted yesterday), they also added a parlay for Musharraf leaving and Osama being captured (at my suggestion). So if these contracts turn out to be liquid, we&#8217-ll be able to get a market-based estimate of the correlation between Musharraf and a reasonable summary statistic for whether outcomes in Pakistan are good for the West.

Of course, the usual caution about correlation and causality applies here. If Osama is caught tomorrow, it probably helps Musharraf keep his job. The longer term conditional probabilities should be less contaminated by this problem and the difference betweens between the prices of the long and short dated contracts might be used to contract a clean(er) estimate of a causal effect.

This seems like the most interesting geopolitical decision market to come along in a little while. But it&#8217-ll only work if it&#8217-s liquid. So those of you with Intrade accounts, put up an order or two. Your country needs you.

Price for Pakistani National General Election Date at intrade.com

Price for Pervez Musharraf (Pakistan) (Rule 1.8 Applies) at intrade.com

Price for Osama Bin Laden at intrade.com

BetFair Australia is a low-margin betting operator that helps horse racing compete against the other gambling products.

No Gravatar

ABC News:

[…] Betfair can take bets from right around the country [Australia], but needs a change in law in states other than Tasmania – where it is already licensed – to enable it to advertise. Mr Twaits says the laws are invalid and are holding racing back.

&#8220-Thoroughbred racing&#8217-s share of the total gambling pie in Australia has dropped from 26 per cent in 1991-92 down to 8 per cent last year, and it&#8217-s continuing to slide,&#8221- he said. &#8220-And the reason for that is that punters, or gamblers generally, are moving to low-margin products that have nothing to do with racing and racing doesn&#8217-t earn a return from those forms of gambling. As a low-margin operator, what we offer racing is the chance to compete better with those non-racing low-margin operators and return a fair amount to the industry.&#8221-

The NSW racing industry says Betfair does not return a fair amount to the industry, paying 24 cents per $100, compared to $1 from bookmakers and $4.50 from the TAB. But Mr Twaits says that is &#8220-just plain wrong&#8221-. &#8220-We&#8217-ve never put an offer to any racing industry official in the country that has been in those terms,&#8221- he said. &#8220-What we have offered to do is to pay a share of our gross revenue to the racing industry and a share of the gross revenue that we&#8217-ve offered to pay is in line with what the TABs pay here in Australia. What matters is how much punters spend with wagering operators. So what they spend with us and what they spend with the TAB might be markedly different but if we agree to pay the same percentage of that spend to the racing industry then they should be happy with that.&#8221-

Thanks to Niall O&#8217-Connor for this interesting link.

Meet David Jack, the managing director of TradeFair.

No Gravatar

David Jack is on the right&#8230- (here with Andrew &#8220-Bert&#8221- Black, the BetFair co-founder).

David Jack, the managing director of TradeFair

David Jack (Managing Director of TradeFair, a BetFair spin-off)

Previously: Binaries and Spreads: BetFair spins off TradeFair.

NEXT: Why does Tradefair care about Prediction Markets – by TradeFair&#8217-s David Jack – 2007-12-06