1. Blog of the Year 2005:Jason Ruspini
(a quantitative analyst at a NYC hedge fund, and a new entrant in the
prediction market commentariat) — I give the 2005 award to Jason Ruspini's
blog, in spite of the fact that only his postings made during the last
quarter of 2005 are newsworthy. (And what I read from him in the first quarter of 2006
confirms that he is on a good rail track.) There
prediction market blogs out there, and in my judgment, Jason Ruspini
displays seven important characteristics: intellectual curiosity, academic approach (including acceptable knowledge
of the papers on prediction markets), Wall Street know-how, mind independence,
(overly inflated) professional ambition in the field, verbal prudence
(bordering paranoia, I'll get to that in a moment), and a knack for
(ultra-light but thoughtful) blogging. Of all the pajamas bloggers that I index,
Jason Ruspini is probably the most obsessive with prediction markets. (And on a lighter note, as I said, I know for a fact that he's midly paranoid —he'll probably take this Blog Of The Year award as ultimate proof that the
entire field of prediction markets is conspiring against him. If you're a
shrink and you've just hung out your shingle in Manhattan,
then this guy is your natural first
prospect.) ... In February 2006, Jason Ruspini, intrepid prediction market reporter, has
exhaustively covered the recent NYC prediction market conference (part 1 - part 2). ...
Besides the names that I have
just dropped, generally speaking, you may sometimes find interesting
entries about prediction markets in blogs maintained by economics professors worth their tenure. The quirkiest economics blogger is Tyler Cowen. He
is an economics professor at George Mason University, Virginia, U.S.A. (Of course, don't expect Tyler Cowen to rat on his colleague and friend Robin Hanson in his blog.) — As for the law professors who sometimes take on prediction markets, the best is Tom Bell. (Cass Sunstein
is a fine scholar whom I highly respect, but he has still to impress me
as a prediction market commentator.) Speaking of law blogs (generally
speaking, not specifically those that I've just mentioned), and to end
on a sour note, I want to report that today's worst crap published in the blogosphere is produced by
professors venturing outside their sphere of competency and invading the economists' turf.
Let's now revisit my 2005 dispute with contrarian Wall Streeter Barry Ritholtz (who has since
incorporated himself into Ritholtz Research and Ritholtz Capital
Partners, if you can believe it). Barry Ritholtz (a prediction market-skeptic) was furious that I exposed his inaccurate 2004 U.S. presidential prediction —the great seer had forecasted that, "All four data points suggest the incumbent [i.e., George W. Bush] gets defeated in November." Barry Ritholtz labeled my piece a "less than honorable criticism" and an "inelegantly critique". Oh, mon Dieu! And what was Chris Masse's big crime, really? According to his wicked system of beliefs, MonsieurBarry Ritholtz couldn't possibly be held responsible for his predictive failure
because, get that, he had told his readers that his political
prediction was "just a
crapshoot". (Please note that, in my 2005 piece, I did quote him extensively and that the crapshoot line was cited.) What Barry Ritholtz did was to attach a low probability (say, 55%) to his prediction —just like the InTrade/TradeSports'
2004 U.S. presidential prediction market did with its own predictive
output. Indeed, if you remember, the probability of the Bush
re-election was valued at 55% at InTrade in the days prior to Election Day.
So, in both cases, we had a prediction associated with a low
probability. The outcome: the prediction market won and the
financial expert lost. I'm sorry, Barry Ritholtz, your session is over. Please make another appointment with the receptionist on the way out.
El Presidente David Perry (of newly incorporated Consensus Point) seems to be determined to pulverize the market leader into fine particles. Let's wish him good luck. Good to see GE as their first client. (David Perry bagged the elephant!)
Will the year 2006 see some of the big consulting firms entering
the field of prediction markets? If yes, why? If no, why not?
What would they bring to the table? What would
it entail for the smaller consulting firms? Any pajamas blogger with a
take on this?
4. Exchange of the Year 2005:BetFair — I give the award to this British-based prediction exchange for two good reasons. Number one, BetFair
managed to obtain a legal status in Australia (thru Tasmania,
which is the little island just South of the Australian mainland) —Tom Bell was not involved in the deal, if that's your next question. Number two, BetFair managed to prove that a betting exchange can help fighting the corruption that it suscitates.
As said one of its co-founders, Mark Davies, BetFair makes it
easier to use insider knowledge to bet on a losing runner "but harder
to do it undetected". This is done thanks to BetFair's top notch IT system (Big Brother is tracing you), and thanks to memorandums of understanding that the betting exchange signed with some
of the (European and Australian) sporting bodies. You have it all in
the five news articles excerpted just below —beware: British English used all over the place.
[Australia] - All bets are on. - Critics say Tasmania is taking a gamble with online betting, but the
Betfair agency says it is making racing cleaner. James Button and
Malcolm Brown report. - [BetFair in Australia] --- On February 7,
Betfair will open an Australian branch after winning a licence to
operate in Tasmania in a joint venture with the Packer's family PBL
company. The idea was born in the late 1990s. Betfair's managing
director in Britain, Mark Davies, says Andrew Black, a software
programmer and professional gambler, was tired of bookmakers telling
him to take or leave the odds they offered. He
had worked on the New York Stock Exchange and wanted to apply to
punting what he saw as the exchange's model of perfect competition: a
customer wants to sell stock, the trader posts up the stock price and
waits for takers. Black took a job cracking codes for the Ministry of Defence. The high-security site insisted he leave work at 5pm, and since he lived alone in a country cottage, that gave him a lot of thinking time. In a Eureka
moment one night, he worked out how he might devise the software that
eventually led to Betfair. Betfair was launched in 2000 with a
death-of-a-bookie advertising campaign, featuring an actor in a coffin.
--- Professor Jan McMillen, director of the Centre for Gambling
Research at the Australian National University, said surveys in NSW and
Queensland from 2001 to 2003 showed up to 13.1 per cent of TAB clients
had used betting exchanges and some had up to 15 offshore betting
accounts. --- "Betfair are pirates and parasites. They came along and
offered us 27 cents in every $100 bet. If we accepted that, we would go
broke. The TAB gives us $5 in every $100." --- But Matthew Engel of The Financial Times saw Betfair
as a "liberating force in betting, which increases consumer choice and
value. It offers something close to a classical perfect market."
--- Davies acknowledges Betfair makes it easier to use insider
knowledge to bet on a losing runner "but harder to do it undetected". He says customers must provide watertight identification and "every
mouse click" can be monitored. Davies thinks bookmakers, at a time of
record profits, are exaggerating the financial threat from Betfair. "We're only 5 per cent of the market. Most
punters still like to bet bricks and mortar and receive their winnings
in their hot, sweaty paws. We don't believe that will change." ---
[Australia] - Back a loser to win a buck with Betfair - [a BetFair explainer] --- While the British exchange can be used in
Australia, all wagers were placed in pounds, and the profits returned
to Britain, where Betfair has secured 90 per cent of the betting
exchange market since it started in 2000. The granting of an Australian
licence means bets can now be placed in Australian dollars, and the
company can advertise widely here. Betfair says licensing the exchange
in Australia means it can be regulated, pay taxes in Australia,
generate employment, provide much-needed competition in the wagering
market and bring new punters to the sport. --- Betfair Australia is
hoping to generate up to $700 million a year by 2009. Mr Lennon says
the deal will be worth $10 million to the Tasmanian Government and
between $35 million and $40 million to the Tasmanian racing industry by
2009-2010. --- 2005-01-11
The McCue Interview: Betfair COO, David Yu
- (alternate URL) - [BetFair] --- Now nearly 100,000 people per month use the exchange
and Betfair processes some three million transactions per day -
matching more than 12,000 bets per minute at peak times. Annual
revenues have passed the £100m mark and there is continual
speculation about a possible stock market flotation, which would make
some of its 500 employees very rich. --- While the rapid growth of
Betfair obviously has its benefits for the company and staff, it also
poses some major technology challenges. Current peak volumes can be
several hundreds of bets per second and there could be up to 3,000 different betting markets on any given day.
But Yu is already looking way ahead to see where Betfair's IT systems
need to be in a few years' time to cope with the predicted growth of
the company. Internally this is
called the "100x" project and will include planning how to rearchitect
certain components of Betfair's systems to support 100 times the volume
of transactions handled today. Yu says: "That sounds really
ambitious and like a big number but at the current growth rates we're
going to hit that in a few years' time so we better start thinking
about it now. We are strategically going in and looking at different
bottlenecks in the system and we either look to rearchitect them or
replace them with different technology or what not." --- 2005-11-14
[United Kingdom, E.U.] - Career of Carter at sad end - [BetFair and horseracing's integrity]
--- Carter was fined £2,000 and excluded for five years from the
sport he once graced after a Jockey Club investigation found him guilty
of, among other things, the corrupt practice of providing information
to a gambler in return for an undisclosed reward. --- The differences
this time were the involvement of the Betfair betting exchange with its
meticulous records, a court order seizing Carter's mobile phone
accounts, and the determination of the Jockey Club's head of security
Paul Scotney to prosecute the case to its fullest. --- Carter's
offences related to eight races in August and September 2003, when the
jockey provided information to high-rolling gambler Chris Coleman, a
tailor from Moorgate in London, that horses such as Aljazeera at
Doncaster and Lili Marlene at Ascot would not be winning. Whether
he 'stopped' the horses - gave them a bad ride so they couldn't win -
is a moot point, but Carter told Coleman they would not win and they
did not, despite most of the eight being highly fancied by their owners
and trainers. --- As soon as Coleman heard from Carter he spread
the news to his associates, Dean Coleman, Neil Yorke, Stephen Hobbs and
James Nash. They all 'laid' large sums of money on Betfair - they were
effectively betting other punters that the Carter horses would lose. They
duly cleaned up: the conspirators invested a total of £483,790
and the total return was £83,121, approximately one pound won for
every six they laid. That does not seem like good odds, but any
punter will tell you that even a 1-6 winner is a goldmine when you know
for certain that you are going to succeed. Thanks
to their technology, which tracks every bet and lay, Betfair spotted
the 'unusual' betting patterns and, as part of their agreement with the
Jockey Club, informed the security department. Scotney and crew
moved in and soon found the evidence of cheating - the key was gaining
the mobile phone records which proved the frequency of contact between
Coleman and Carter. --- 2005-10-09
[United Kingdom, E.U.] - Online Gambling raises the Ante
- Betfair is spicing up the $8 billion Internet gaming market with its
EBay-like strategy and talk of a possible IPO. Regulatory issues and a
backlash from traditional gaming houses may sway the odds against it. -
(PDF - PDF mirror)
- [BetFair | Internet gambling and betting | U.S. political prediction
markets] --- Oelbermann says Betfair stands out from competitors. He
compares it to EBay Inc., the San Jose, California–based online
auction house that matches buyers and sellers in a global marketplace.
Betfair charges nothing to a person who offers a wager. It takes a
maximum cut of 5 percent of a winning bet. The strategy eliminates the bookmaker, the agent who offers odds and
accepts bets for a fee. In traditional casinos or gambling sites,
customers play against the house, which builds a profit margin into the
odds, making it hard to win, Black says. In contrast, Betfair is a
betting exchange—an inspiration Black says he took from his stint
in 1995 trading U.S. equities and learning about the New York Stock
Exchange. A person who accesses Betfair’s Web site can offer a
bet and set the odds. If the odds are attractive, other bettors will
take the counter position. This happens with split-second timing, and
it’s anonymous. Everybody has funds in the system to back their
wagers. Transactions are reconciled immediately after the event is
completed. --- Black says Betfair avoids these legal headaches
by taking no bets from U.S. residents. Customers must give their
address to register and submit bank details to deposit money into their
account before they can wager. “We don’t accept bets from customers with a U.S. address or U.S. credit card,”
says CEO Stephen Hill, 44, sitting in a corner room in the
company’s glass and steel offices overlooking the Thames in West
London. -- “This market is
about liquidity and size,” he says. --- (first publication
maybe: 2005-09-01) - 2005-10-00
President Bush isn't the only one who gambles on on Supreme Court nominees. - (subscription) - (via Barry Ritholtz)
- [SCOTUS | Alito nomination prediction market at TradeSports/InTrade]
--- On Oct. 27, after
White House General Counsel Harriet Miers asked President Bush to
withdraw her nomination to the high court, Intrade began offering
contracts with a nominal value of $10 on each of 28 candidates thought
to be in the running to replace her. The Alito contract traded as low
as 60 cents that first day, meaning holders would win $9.40 per
contract if he was nominated. He was considered a long-shot for a seat
expected to go to a female jurist. If Bush had not picked Alito to
replace Miers, his Intrade contract would have expired worthless. In
fact, bettors who bought contracts on the other 27 candidates lost all
their bets. Hourly data provided by Intrade show an enormous spike in
interest in Alito on Oct. 28 after Erick Erickson, a blogger for
RedState.org, reported at 8:29 EST that "very credible people outside
the White House and lower-level staff people inside the White House
have Alito on their lips." RedState.org, a Website designed to appeal
to Republicans, attracts about 110,000 visitors each day. Erickson, who
tells us he does not gamble, wrote a follow-up blog later that morning.
He quoted a source who claimed, "The White House General Counsel's
Office is not doing too good at keeping this a secret." Bettors had
purchased just 180 Alito contracts on Oct. 27. Purchases of Alito
contracts shot up to 1,495 on Oct. 28. We counted just 128 Alito trades
prior to Erickson's first report. The betting continued throughout the
weekend, averaging 442 per day. Bush formally named Alito at 8 a.m. EST
on Oct. 31. --- 2005-11-05
Some other expired prediction markets: dig my cache.
6. Media of the Year 2005: Which media has most inflated the prediction market bubble in 2005?... Humm... I think it's Fortune magazine, with this story, Making a Market in (Almost) Anything. Andy Serwer wrote: "A fledgling website with some big-name backers, Intrade allows regular people to invest—not bet—on the current events they know best." See the problem with that line??? The trouble is that Andy Serwer equates speculating and risk-free investing. This is a big no-no for the CFTC, which devotes not one, but two didactic webpages explaining the economic purpose and risky business of
speculating on futures markets, for individual investors. I also take
strong exception to Andy Serwer's bold papal statement: "It also correctly signaled that Cardinal Ratzinger would be chosen as the new Pope, arguably the most secret vote on the planet." Surely you're joking, Mister Serwer! (Read my 2005 story: How the BBC, Fortune and the New York Times went
overboard claiming that the prediction markets had foreseen the name of
the new pope!) Gee, Fortune has truly become the business magazine of record —for the Elvis sightings, that is.
The market as crystal ball - [prediction markets | the accuracy of the InTrade/TradeSports futures markets] - (subscription - PDF) - 2005-07-25
Making a Market in (Almost) Anything - A fledgling website with some big-name backers, Intrade allows
regular people to invest—not bet—on the current events they
know best. - (subscription - republished - republished - republished - PDF) - [prediction markets | InTrade/TradeSports] --- Because it
involves putting money on everyday events (like sports), trading
contracts on Intrade may seem a little like gambling. But Intrade
executives say it’s not: For one thing, Intrade takes only a
commission and does not act as "the house" like a Las Vegas casino.
Also, investors on Intrade can, at least in theory, use skill to hedge
their risk, as they do in the equity market. Intrade
says it is in the process of seeking regulatory oversight by the
Commodities Futures Trading Commission, the federal agency that
regulates futures and options. (Citing the lack of a formal
application, the CFTC had no comment.) --- LeBaron
points out that as Intrade expands and liquidity increases, the value
of its predictive information and the accuracy of the pricing will grow. --- 2005-07-25
An Executive Risk Handbook - Five ways managers can use scenario planning to prepare for disasters. - (PDF - scroll down) - [risk management | Intrade/TradeSports] --- #4. Use the power of markets.
You've heard of the remarkable success of predictive markets, where
real people bet real money on the likelihood of specific events. Around
for years, these markets attracted lots of acclaim after calling the
squeaker 2004 presidential election. You can check
them—InTrade.com is the best known—to see if they're
offering contracts on events that might affect your business. For
example, InTrade recently showed chances of private Social Security
accounts being enacted by December 2006 at about 17%. But
you can also create such markets, focused on exactly the questions you
want answered, among your own employees (generally with small amounts
of money furnished by the company). Hewlett-Packard has used
internal markets to forecast sales more accurately than the marketing
manager could; Eli Lilly has used them to predict the success of drug
research with uncanny accuracy. Well-designed predictive markets can give managers new insight into specific risks and how they might affect the company. --- by Geoffrey Colvin - 2005-10-03
Could be that I'm speculating too much further into the near future, but I believe that the winner for the year 2006 will be CNBC. — Watch this: CNBC Video - InTrade - MP4 file - [U.S. political prediction markets @ TradeSports/Intrade] - 2006-02-09
Trade Exchange Network Cooperates with Federal Investigation and Agrees to Pay Civil Penalty of $150,000. - [CFTC
fines TradeSports/InTrade] --- The CFTC consent order finds that between
January 2003 and May 2005, TEN, through its websites, offered the
following commodity option contracts, among others, to U.S. residents: Gold
Futures Year End; Daily Crude Oil; Light Sweet Crude Oil Futures Year
End; Intraday Euro versus U.S. Dollar Rate; and U.S. Dollar versus Yen
Cash Rate. According to the findings in the order, TEN actively
solicited U.S. residents to trade such contracts by retaining an
individual in the U.S. to market TEN’s products throughout the
country, and TEN ultimately developed a U.S. customer base that was
roughly 33 to 40% of its total customer base. The order further finds
that the commodity option contracts offered on TEN’s websites
were not excepted or exempted from the Commission’s regulation
banning options trading, and therefore violated the CEA. --- (+ CFTC consent order - PDF) --- 2005-10-04
Using the 'Wisdom of Crowds' to Tell the Future - (archive - republished) -
[InTrade/TradeSports] --- The Commodity Futures Trading Commission
found that the company had offered some kinds of contracts banned by
the agency. The company did not admit or deny it, but agreed to pay a
fine of one hundred fifty thousand dollars. --- 2005-11-11
Wanna know more about the ongoing saga between TEN and the CFTC? The digest is here.
The best prediction market stories of the past year: NEWS 2005. — Indexing the best prediction market stories of this year: NEWS 2006.
8. Opinion of the Year 2005: (Are you sitting? The last thing I want is to be responsible
for you, my dear readers, falling to the ground out of an emotional shock and hence busting your ass.) Well, here's how Chapman university law professor Tom Bell blogged his reaction to the above story: "Good News: CFTC Busts Intrade."
(Excerpts in the paragraph just below.) — Was it an attempt at humor??? Does Tom Bell need a timeout??? — Part of it is the
Californian attitude (a mix of denial and audacity,
maybe, coolness run amok, more surely), and the other part of it
is Tom Bell's clever (but economically unrealistic) attempt at finding a loophole or shaping a bill on which to base the legality of U.S.-based prediction exchanges.
Good News: CFTC Busts Intrade. - [CFTC fines TradeSports/InTrade]
--- Surprisingly, even to me, I find it this turn of events chock-full
with good news. --- The implied good news comes out of that last
sentence, as well as the whole of the settlement: The CFTC expressed no
interest in regulating any claims not already traded on CFTC-regulated
U.S.-based exchanges. It went after TEN only for offering contracts on,
inter alia, Gold Futures Year End, Daily Crude Oil, and U.S. Dollar versus Yen Cash Rate. That sends the signal that the CFTC won't mess with an exchange offering trades on purely political or scientific claims. Note that TEN does not claim that it will
trade political or scientific claims on the CFTC-regulated, U.S.-based
market that it has announced. It mentions only pretty traditional sorts
of contracts. But it doesn't say it won't
offer trading on political or scientific claims. And, even if it does
not, its wrangle with the CFTC hints that somebody else might do so
without suffering similar regulatory action. That should come as good
tidings to The Washington Stock Exchange and the (soon to be relabeled)
Simon Exchange. --- by Tom Bell - 2005-10-10
As for Tom Bell's proposal aiming at legalizingprediction exchanges in the U.S., my take is that there's no business model for a commercial organization running real-money "science,
technology, or public policy" prediction markets, only.
As a matter of fact, I can't even remember any real-money prediction
markets dealing with these topics —wait, yes, there were
those lousy social security reform prediction markets on which my ideological friend Donald Luskinlost his shirt (alas). Tom Bell has a fertile imagination, but it won't help our cause, until our good doctor takes an Air Lingus
flight from JFK to Dublin, Ireland, to receive a crash course on the
economics of betting exchanges. As reported by Wolfers and Zitzewitz in Five Open Questions (PDF - PDF), sports event-driven futures markets are the bread earner for the betting exchanges, as of today.Tom Bell's proposal is ethereal —and entertaining enough to pique the curiosity of a small bunch of sofa conspirators— but it has no business groundings.
2006-03-13 update: The Prediction Exchange: Progress in Promoting the Sciences and Useful Arts - (PDF) - by Tom W. Bell - 2006-03-11
The best prediction market takes of the past year: OPINION 2005. — Indexing the best prediction market takes of this year: OPINION 2006.
9. Paper of the Year 2005:Justin Wolfers and Eric
Zitzewitz's 2005 answer (Interpreting
Market Prices as Probabilities)
to Charles Manski's 2004 question (PDF), "What is the logical
basis for interpreting the price of an all-or-nothing futures contract
as a market probability that the event will occur?". — Here's a
short list of papers, with excerpts:
Market Prices as Probabilities -
- by Justin Wolfers and Eric
While most empirical analysis of prediction markets treats prices of binary options as predictions of the probability of future events,
Manski (2004) has recently argued that there is little existing theory
supporting this practice. We provide relevant analytic foundations,
describing sufficient conditions under which prediction markets prices
correspond with mean beliefs. Beyond these specific sufficient
conditions, we show that for a broad class of models prediction market
prices are usually close to the mean beliefs of traders. The key
parameters driving trading behavior in prediction markets are the
degree of risk aversion and the distribution on beliefs, and we provide
some novel data on the distribution of beliefs in a couple of
interesting contexts. We find that prediction markets prices typically provide useful (albeit sometimes biased) estimates of average beliefs about the probability an event occurs.
Aggregation of Information and Beliefs in Prediction Markets - (PDF) - by Marco Ottaviani and Peter Norman Sørensen - 2005-12-19
model trading in prediction markets by a population of individuals with
heterogeneous beliefs and private information. The resulting market
price is interpreted as an average of the information-updated private
beliefs. This price can effectively aggregate all available information, but may be a biased predictor of the event due to the heterogeneity of beliefs.
key theoretical questions are: Will trade lead to a price that
aggregates the diverse expectations of the market participants? Can the
price be interpreted as the probability of the event?
We contribute three observations to the prediction market literature. First, we allow for asymmetric information as well as heterogeneous beliefs. Second, we analyze a tractable example with constant absolute risk aversion. Third, we argue that the arithmetic average of the traders beliefs is not necessarily the most natural benchmark for the price even in the absence of asymmetric information .
Prediction Markets in Theory and Practice - (PDF - PDF) - by Justin Wolfers and Eric Zitzewitz - 2005-11-23
The claim that prediction markets can efficiently aggregate information is based on the Efficient Market Hypothesis.
Explaining why there is any trade in prediction markets
remains an important open theoretical question. Wolfers and Zitzewitz
(2006) provide a simple adaptation of the Kyle (1985) model in which
trade is driven by uninformed outsiders with either hedging- or
entertainment-driven demand for the prediction security, or by
manipulators attempting to influence market prices. Another
important role of prediction markets is that potential trading profits
provide an incentive for information discovery. --- Another key
advantage of prediction markets over alternative approaches to
information aggregation is that they provide incentives for truthful
revelation of beliefs.
First, market prices tend to respond rapidly to new
information. --- Second, in most cases, the time series of prices in
these markets appears to follow a random walk, and simple betting
strategies based on publicly available information appear to yield no
profit opportunities. That is, these markets appear to meet the
standard definition of weak-form efficiency. Third, the law of one
price appears to (roughly) hold, and the few arbitrage opportunities
that arise in these markets are fleeting, and involve only small
potential profits. Fourth, attempts at manipulating these markets
typically fail. --- Finally, prediction markets usually provide quite
accurate forecasts and have typically outperformed alternative
The potential to apply these markets to determine the
consequences of a range of contingencies has led Hanson (1999) to term
these “Decision Markets”. Indeed, Hanson (2000) has
suggested that such markets could be used to remove technocratic policy
implementation issues from the bureaucracy, a suggestion endorsed in
Hahn and Tetlock (2006). Moreover, while the previous example involves
only one contingency, Hanson (2003) suggests that market scoring rules
can allow traders to simultaneously predict many combinations of
outcomes. The basic intuition of his proposal is that rather than
betting on each contingency, traders bet that the sum of their errors
over all predictions will be lower.
Five open questions about prediction markets - (PDF - PDF) - by Justin Wolfers and Eric Zitzewitz - (2005-01-21) - 2006-02-00
Interest in new applications of prediction markets is focused in three domains: forecasting, decision making, and risk management.
1. How to attract uninformed traders? Counterintuitively, the
problem for most prediction markets is attracting sufficient uninformed
An important implication of the model sketched above
is that the success of the prediction market in generating trade
depends critically on attracting uninformed traders. ---
Risk love or the “thrill of a gamble” provide obvious
motives for uninformed traders, and both Tradesports and Betfair have
successfully attracted many sports bettors to their markets. --- As the wonkishness of the contract rises, however, volume and liquidity falls rapidly. --- To summarize, three routes to attracting order flow have been successful thus far: offering sports betting, subsidization, and, possibly, exploiting career concerns. Each has their drawbacks, however.
2. How to tradeoff interest and contractability? A
fundamental problem in mechanism design is that the outcomes of
interest are often impossible to write into contracts.
Observable indicators of contractability have thus far
had little explanatory power for a prediction market’s success.
--- Tradesports traders have thus far been willing to trust the exchange to make a good faith determination,
and this trust is helped by the fact that Tradesports does not itself
hold positions in any of its contracts and that it has a clear need to
maintain a reputation for fairness among its traders. But as the sums
traded become more substantial, traders may begin to worry more about
3. How to limit manipulation?
Just as thinking about the contractability of the
outcomes of interest raises questions about outcome manipulation, the
prospect of basing decisions on prediction markets raises questions
about the potential for price manipulation.
4. Are markets well calibrated on small probabilities?
An exception to the generally good predictive track record of betting and prediction markets is their performance on low-probability events.
The best-documented example is the favorite-longshot bias in horse
racing (Thaler and Ziemba, 1988). --- Wolfers and Zitzewitz (2004a)
discuss a similar mispricing from S&P 500 index contracts on
Tradesports. In both examples, market prices overestimate the probability of unlikely events.
5. How to separate correlation from causation?
Decision market securities are designed to allow one to
estimate how expectations of policy outcomes vary with the policy
Conclusion: The first question arguably falls more in the field of marketing than economics: how to attract uninformed order flow to markets? This is important because these traders provide the potential profit motivating informed groups to trade.
Lower transaction costs, in both the monetary and the convenience
sense, are important, but inherent interest or buzz is clearly an
The Iowa Electronic Market: Lessons Learned and Answers Yearned - (PDF) - by Joyce E. Berg and Thomas A. Rietz - 2005-01-00
Chris Masse's comments:
I am not able to excerpt this paper. They locked the
content, within Adobe Reader. Very smart. Congrats. Some scholars live
with their time, I see.
In page 2, the authors treat TradeSports as "an online betting operation that has incorporated IEM-like contracts on political events, sometimes even duplicating IEM contracts." Number one, just like IEM, TradeSports/InTrade is a real-money prediction exchange (a.k.a. a betting exchange, in the U.K.), where trading is going on, not an "online betting operation" (which is a moniker that would better qualify any
Web-based bookmaker or sportbook). If you want people to have respect
for IEM (a non-profit organization that monopolistically prospers on an anti-democratic CFTCgraciosity),
please show respect for innovative commercial ventures that create
wealth and jobs and that form the basis of the capitalism that you're
supposed to study, as tenured professors of economics. Number two, nobody contests
that IEMpioneered (political) prediction markets in 1988. Number three, thank God,
financial innovation can't be locked into tiny university laboratories
and will always find a way to spread over in the real world of
business, so as to benefit us all.
Presidential Reality Check: An
Assessment of NewsFutures' 2004 Presidential Election Prediction
- by Emile Servan-Schreiber - 2005-01-27
Manipulating Political Stock Markets: A Field Experiment and a Century of Observational Data - (PDF - PDF trading strategy) - by Paul W. Rhode and Koleman S. Strumpf - (2005) - 2006-02-00
Political stock markets have a long history in the United States. Organized
prediction markets for Presidential elections have operated on Wall
Street (1880-1944), the Iowa Electronic Market (1988-present), and
TradeSports (2001-present). Proponents claim such markets
efficiently aggregate information and provide forecasts superior to
polls. An important counterclaim is that such markets may be subject to
manipulation by interested parties. We analyze this argument by studying alleged and actual speculative attacks -- large trades, uninformed by fundamentals, intended to change prices-- in these three markets. We
first investigate the speculative attacks on TradeSports market in 2004
when a single trader made a series of large investments in an apparent
attempt to make one candidate appear stronger. Next we examine the historical Wall Street markets
where political operatives from the contending parties actively and
openly bet on city, state and national races; the record is rife with
accusations that parties tried to boost their candidates through
investments and wash or phantom bets. Finally we report the results of
a field experiment involving a series of planned, random investments--
accounting for two percent of total market volume-- in the Iowa Electronic Market in 2000.
In every speculative attack that we study there were measurable initial
changes in prices. However, these were quickly undone and prices
returned close to their previous levels. We find little evidence that political stock markets can be systematically manipulated beyond short time periods.
Prediction markets trade contracts with payoffs explicitly linked to future events.
An example is a binary option which pays a dollar on the outcome of a
specific event, such as a candidate's victory or an on-time product
launch. An efficient prediction market aggregates available information, yielding prices that are the best forecast of the event's probability.
For the purpose of this study, fundamentals are any information that influences the underlying value of the contract. A speculative attack is defined any trade, uninformed by fundamentals, intended to change prices. A (successful) manipulation
is a speculative attack that achieves its objective of changing prices.
(A successful manipulation is usually not possible unless the trades
influence the beliefs of other market participants. An investor’s
beliefs are defined with respect to the fundamentals, as well as the
future actions and beliefs of other investors.)
TradeSports operates several online prediction markets. It ran the most influential market on the 2004 US Presidential election, which attracted more than $15M in trade volume.
Shares in the main election market paid a fixed amount if Bush won, and
the prices were scaled between zero and a hundred to give the usual
Conclusion: In almost every speculative attack that
we study there were measurable initial changes in prices. However,
these were quickly undone and prices returned close to their previous
levels. Our investigation of evidence from field experiments and
contemporary as well as historical observational data suggests it is difficult and expensive to manipulate political stock markets beyond short periods. And the period appears to become shorter over time --from days (New York Markets) to hours (IEM) to minutes (TradeSports).
Designing Real Terrorism Futures - (PDF) - by Robin Hanson - 2005-09-00
Policy Analysis Market was accused of being a terrorism futures market,
and was quickly canceled due to the resulting outrage. Many expressed
simple incredulity and repugnance, but some attempted to identify more
specific concerns. While some of these concerns seem to be based on
misunderstandings, others are more reasonable and would deserve
attention if one were actually going to create a terrorism futures
market. In this paper I have discussed five big concerns. Two of these concerns, manipulation and moral hazard, were mentioned often in the publicity surrounding PAM. Manipulation does not seem to be a real concern, nor do noisy traders more generally, at least when informed traders are not too severely limited in how much they can trade. Moral hazard is potentially a problem for high-volume markets,
and several methods were identified for dealing with this problem,
especially slightly reduced anonymity. But in fact terrorism futures is
likely to be low-volume, making moral hazard a non-issue. The three
issues discussed here that were not mentioned much in the publicity
surrounding PAM are: combinatorics, hiding prices, and decision
selection bias. The combinatorics and decision selection bias problems are relatively technical, and once understood have relatively technical solutions. The problem of hiding prices, so as to not unduly alarm the public or inform the terrorists, also turns out to have a reasonable technical solution. Thus in the end none of these problems seem insurmountable.
The Informed Press Favored the Policy Analysis Market - (PDF) - by Robin Hanson - (2005-03-18) - 2005-09-08
This paper describes such an analysis of over five hundred
media articles mentioning the Policy Analysis market. --- This result
suggests that while uninformed opinion disliked PAM, informed opinion
favored it. Unfortunately, public policy regarding PAM continues to
reflect the uninformed opinion, a situation that seems unlikely to
change anytime soon. This should give pause to those who think that our
democratic processes typically produce informed policy, even when the
electorate is highly uninformed. In the PAM case, as in many others
(Dixit, 1997; Stiglitz, 1998), policy reflects the uninformed position.
Searching for Google's Value: Using Prediction Markets to Forecast Market Capitalization Prior to an Initial Public Offering - (PDF) - by Joyce E. Berg, George R. Neumann and Thomas A. Rietz - 2005-12-00
underpricing is endemic. Many theories have been developed to explain
it. To inform theory and to investigate the practical application of
prediction markets in an IPO setting, we conducted markets designed to
forecast post-IPO valuations before a particularly unique IPO: Google.
The combination of results from these markets and the unique features
of the IPO help us distinguish between underpricing theories. The
evidence leans against theories which require large payments to buyers
to overcome problems of asymmetric information between issuers and
buyers. It is most consistent with theories where underpricing is in
exchange for future benefits. The prediction market results also show that it is possible to forecast post-IPO market values and, therefore, avoid losses associated with underpricing when a firm wishes to do so.
The effectiveness of pre-release advertising for motion pictures - (PDF) - by Anita Elberse and Bharat Anand - 2005-03-05
source for data on market-wide expectations, the Hollywood Stock
Exchange (HSX, www.hsx.com), is a popular Internet stock market
simulation that revolves around movies and movie stars. HSX has over 520,000 active users, a 'core' trader group of about 80,000 accounts, and approximately 19,500 daily unique logins.
New HSX traders receive 2 million 'Hollywood dollars' (denoted as "H$2
million") and can increase the value of their portfolio by, among other
things, strategically trading 'movie stocks'. The trading population is
fairly heterogeneous, but the most active traders tend to be heavy
consumers and early adopters of entertainment products, especially
films. They can use a wide range of information sources to help them in
their decision-making. HSX stock price fluctuations reflect information
that traders privately hold (which is only likely for the small group
of players who work in the motion picture industry) or information that
is in the public domain – including advertising messages. Despite
the fact that the simulation does not offer any real monetary
incentives, collectively, HSX traders generally produce relatively good
forecasts of actual box office returns (e.g. Elberse and Eliashberg
2003, Spann and Skiera 2003). According to Pennock, Lawrence, Giles and
Nielsen (2001a; 2001b), who analyzed HSX's efficiency and forecast
accuracy, arbitrage closure on HSX is quantitatively weaker, but
qualitatively similar, relative to a real-money market. Moreover, in
direct comparisons with expert judges, HSX forecasts perform
10. Presentation of the Year 2005:John Ledyard
lecturing at Northwestern University, last spring. — What I liked: one can
read his slides like a paper. Very didactic. He takes you from the
basics and brings you to the hotest research topics. Smart guy. —
What I disliked: his choice of the phrase "information markets"
(as opposed to "prediction markets", "idea futures markets", or
"event-driven futures markets"), which I labeled a quasi-fraud (in my trademarked, outrageous, take-no-prisoner rhetorical style). — Here's a short list of presentations, with excerpts:
There is theoretical, experimental, and field evidence that traditional IM’s may work.
But there is also evidence that there are impediments to complete
information aggregation especially in environments with informationally
large agents. Should we worry about small numbers? Can we find better information mechanisms?
To be useful in many potential applications, Information
Markets need to perform well with small numbers and informationally
large traders. Traditional markets and pari-mutuels are not up to this.
Two possible approaches: – Subsidize the action in the traditional designs. – Design new mechanisms.
Do traditional IM’s work? -- YES: Large numbers
of informationally small agents and reasonably simple environments seem
to overcome some of the no-trade incentives in traditional markets and
pari-mutuels. – I don’t think we yet know exactly why.
(Open Puzzle!) -- NO: With informationally large agents, much less
incomplete markets or complex decision problems, it does not look so
good. – N = 3 creates serious problems for both markets and
pari-mutuels, even in simple environments. – More research is
Can we design IM’s that work? -- YES: For informationally small agents in simple environments, – The Market Scoring Rule gets it early and often.
– Polling with incentive payments also works pretty well but is
slower to get the job done. – Markets are mirage prone, even with
noise traders. – Pari-mutuels are slow, even with subsidization
-- MAYBE: For informationally large agents, – The four best (ms,
msr, p, pms) are all off by.10 to .15. – All mechanisms we looked
at can be improved upon. – Or is there an impossibility theorem?
Old tech (~1950+): Proper Scoring Rules -- N/Q á
1: works well, N/Q à 1: hard to combine --- New tech (~1990+):
Info/Predict Markets -- N/Q à 1: works well, N/Q á 1:
thin markets --- The best of both:Market Scoring Rules –
modular, lab tests, compute issues, …
MSR Usage Concept • User browses current probs,
expectations – Can set assumptions, browse other values given
that – Can see market value of portfolio given assumptions •
Upon finding an “odd” current value E(x|A) – Can see
value’s price/trade history – Can see how far long/short
are from past trades • User proposes new value to replace old
– Told exact bet required to implement change – Can accept,
and/or make book orders & trading agents
Using Combinatorial Market Scoring Rules: Common Knowledge Markets,
YooNew, US Air Force Research, etc.
Summary • How elicit informed estimates? –
Scoring rules if N/Q á 1, info/predict markets if à1
• Market scoring rules do both: – Key insight …
reused scoring rules are market makers – Browse billions of
estimates, change ones want via bets – Lab tests confirm ability;
are growing # groups using • Log rule has many advantages –
If bet on p(A|B), keeps p(B), I( Α, Β, Χ ), I( Β,
Α, Χ ) … – Noisy choices easier to interpret
– Costs no more for all var/value combos – Computation
simplified, but still issues to explore
Interpreting Prediction Market Prices as Probabilities - (PDF)
- [Answer to Charles Manski | Exploring the relationship between
prediction market prices and the mean of the distribution of beliefs. |
Are Prediction Market Prices Close to Probs?] - by Eric Zitzewitz and Justin Wolfers - @ Mini-Conference on Information and Prediction Markets - 2005-12-19
Calibration results: Prices
are “close” to mean beliefs for plausible utility functions
and distributions of beliefs (especially for π ≈ ½).
Empirical results: »
Market prices are generally close to mean beliefs. » Market
prices are generally close to actual outcomes. » Market prices
are fairly efficient estimators of outcomes.
To see a list of other interesting delicious links (a.k.a. social bookmarks), please go to my resources webpage.
To be frank with you, I am not too much happy with Wikipedia's prediction market entry. I have contributed to the external links
section (only), and will do again in the future (when I have time). Speaking of the textual content, I
think that the prediction market field is too youngish
to be able to aggregate knowledge coherently and pertinently. Besides, I have
noticed vendors over-contributing, but overall, the wisdom-of-crowds
policing process would end up clearing up the slate. Thus, my reasonable hope is that the content quality will improve over
the next years.
By giving this award to David Pennock, Chris Masse does not make any value judgment on the dynamic pari-mutual market or on the Yahoo! Tech Buzz Game.
(Gee, I sound like a blogging law professor drafting his/her website policy, now.) My purpose is here is to salute David Pennock's attempt at creating a better mousetrap.
Taking a brand-new product/service idea from concept to market is the
toughest job on Earth. Success cannot be bought. For a third of their
growth, modern economies rely on inventors and innovators, but God
knows how hard it is to persuade companies to gamble on new ventures. Thank you YAHOOOOOOOOOOOOOOOOOOOOO!!!!!!!!!...
2006-03-06 evening update: I gave the reason for granting this award: David Pennockinvented something, and put it to try. That's that approach that I did salute. I wrote one full paragraph to explain just that. Most inventors/innovators fail on the marketplace, but there should be a system of societal values that reward people for trying, and trying again, and again. Innovation is something that everybody should embody, including scholars. That is why I don't have a specific innovators category.
On a technical note, I was aware of the shortcomings of
the DPMM/YTBG (Version 1). And I know that our good doctor
is a bit
secretive about the current version of DPMM/YTBG (Version 2). OK, but that's peripheral to the core reason for my granting this 2005 award.
The share-ratio price function DPM (the mechanism in current use on the TBG) is provably arbitrage free, meaning that it does not suffer from the fatal flaw of the original money-ratio price function.
I haven't been purposefully secretive about the share-ratio price function DPM. I simply haven't found the time to write it up for publication.
13. Software of the Year 2005:Consensus Point's Idea Futures — David Perry has teamed up with Canadian Ken Kittlitz (historically speaking, the first Robin Hanson-follower, circa 1994) in order to offer two versions of the software that powers the ForeSight Exchange (the most prestigious play-money prediction exchange): one open-source version that they hand out for free (Idea Futures), and one proprietary version that they license or webhost for a fee. See, Consensus Point open sources a major portion
of its software so the user base grows larger. Humm... David
Perry, clever guy!... His marketing strategy reminds me of the
James Bond villain in Live And Let Die.
His plan was to inundate the American market with cheap cocaine, and
then, when just everybody in the nation is hooked, hop, he would engineer a giant price hike
so as to pluck clean the hordes of addicts and secure his seat in the
Chris Hibbert (the Zocalo project manager) has also his reves de grandeur. His long-term goal: "Deploying public prediction markets with partners." (Source: his PPT file). TradeSports/InTrade, BetFair, HedgeStreet, and even the Chicago Mercantile Exchange,
are trembling in their pants. The terror is welling inside
them. Somebody should give Chris Hibbert a crash course on the
liquidity problem that any new prediction exchange (a.k.a.
betting exchange) faces when entering the field.
Chris Hibbert is the
proud 2005 (re)inventor of the smartest idea of 2004 (which was originally hatched by NewsFutures for the now-defunct MIT Tech Review's play-money prediction exchange, baptized Innovation Futures, and which, I was told, is currently featured in some version of the NewsFutures prediction market software). Chris Hibbert described a way for prediction exchanges to increase the liquidity of N-way claim markets by posting orders that can trigger the simultaneous buying or selling of the full set of possibilities. For more info on Chris Hibbert's technique, read his slides (PPT file). For a critical review, please go reading the last part of this Jason Ruspini's blog post (which subtly makes the point that the method just creates the appearance of liquidity). And for the original idea, namely the N-ary
outcome markets with automatic arbitrage (so as to enable logical
bid/ask matches across alternative outcomes) used by NewsFutures, thank God (namely, Brewster Kahle, the ultimate librarian), the Way Back Machine has archived the webpage featuring the Innovation Futures FAQ.
EXCERPT OF THE (ELLIPTIC) FAQ: "Are the outcomes' order books
independent from each other? No, there are several rules that link the
order books of the outcomes in a single market: ..." ... (Read the rest
at the bottom of the FAQ webpage.)
Speaking of innovation, I was looking at my software webpage the other day, and something struck me like an unexpected Windows blue screen in the middle of a crucial data processing session. Of all the prediction market software that I list, the NewsFutures family of products has the wider range of designs (CDA, scoring rule, dynamic pari-mutuel market). Very few software makers venture beyond CDA, it's a pity.
I publish my 2005 awards in early March 2006 (as opposed to the end of December 2005), for two reasons. Number one, during the month of January, I want to forget everything about the year 2005, so as to start up with a fresh mind in mid-February, when I begin typing this webpage. Number two, my strategic objective is to upstage the Oscars —it's time that clarity replaces futility.
Person of the Year 2005: Bill and Melinda Gates + Bono - (subscription) - 2005-12-18
"Zocalo and my name came up a few times, sometimes positively, sometimes not."
Chris Masse's comments:
fantasy about creating a brand-new public (real-money) prediction
exchange while downplaying (by omission) the difficulty of solving the
new entrant's liquidity problem is a very common theme among the prediction market crowd.
As for Chris Hibbert's
(re)invention, it's very often in science and technology that the
same discovery/hypothesis/theory/invention/innovation pops up in
two different minds, at around the
same time. The Nobel foundation deals with this kind of problem on a daily basis. To take an example close to us, Robin Hanson created his concept of idea futures in the fall of 1988 not knowing that the IEM gang had already invented the political prediction markets, earlier that year. (IEM was not on the Web in 1988. It was an experiment internal to the University of Iowa, involving only their faculty.)
Synonyms of "real-money prediction markets":
derivatives, European binary options, European call options, event-driven futures
markets, event futures markets, idea futures markets, information
futures markets, predictive market-based technology.
Inappropriate terms: event markets, idea markets, information markets, opinion
markets, predictive markets.
Sub-category: "decision markets" are
prediction markets used as decision tools.