Prediction Market Movie

[Emile Servan-Schreiber]

For your viewing pleasure, NewsFutures invents the "prediction market movie": a new way to look back at recent history, or just relive the excitement of past trading. For instance, the link below takes you to our fist production: a review of the epic trading battle between Clinton and Obama from March 2007 to March 2008 (after Texas and Ohio, it was over for HIllary).

http://us.newsfutures.com/hcboWidgetCode.html

Enjoy!

--Emile
NewsFutures

[David Pennock]

Very cool. If only I had bought when Obama's head was so tiny. :-)

Emile Servan-Schreiber wrote:
> For your viewing pleasure, NewsFutures invents the "prediction market
> movie": a new way to look back at recent history, or just relive the
> excitement of past trading. For instance, the link below takes you to
> our fist production: a review of the epic trading battle between Clinton
> and Obama from March 2007 to March 2008 (after Texas and Ohio, it was
> over for HIllary).
>

> *http://us.newsfutures.com/hcboWidgetCode.html*
>
> Enjoy!
>
> --Emile
> NewsFutures <http://www.newsfutures.com>
>
> >

[Bob Holley]

Did any prediction markets predict the stock market meltdown, or was it one
of those unpredictable events that Taleb calls black swans?

Dr. Robert P. Holley
Professor, Library & Information Science Program
Wayne State University
Detroit, MI 48202
313-577-4021 (phone)
313-577-7563 (fax)
aa3...@wayne.edu (email)

[Nigel Eccles]

Hi Robert,

I believe the best forecast indicator would be derived from out of the money
put options on the DJ index. I'm not sure what they were trading at however
I believe the volatility measure was trending down prior to the turmoil.

I've still to read Taleb's recent book but I think the recent turmoil would
come under his definition of a black swan.

Nigel

[Michael Strong]

I'd be happy for someone else to provide a more rigorous response, but after seeing Taleb ridicule prediction markets, it is worth providing a bit of a response.

First, Taleb himself acknowledges that some Black Swan problems can be shifted into more manageable territory by means of improved information.

It strikes me that, as a tool, prediction markets are in an early stage of innovation in which we may need dozens, or hundreds, more prediction markets (esp. real money markets) to be implemented in various realms to reduce uncertainty.

To take one example, anonymous corporate prediction markets have increased the accuracy of sales predictions and product release dates.  Abramowicz and Henderson have a paper on using prediction markets to improve corporate governance,

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=928896

Given that Freddie Mac's Chief Risk Officer expressed concerns in mid-2004 leading to his being fired in 2005,

In an interview, Freddie Mac's former chief risk officer, David A. Andrukonis, recalled telling Mr. Syron in mid-2004 that the company was buying bad loans that "would likely pose an enormous financial and reputational risk to the company and the country."
http://econlog.econlib.org/archives/2008/08/freddie_macs_wh_1.html

Suppose Andrukonis and others within Freddie Mac and Fannie Mae had had the opportunity to use internal prediction markets through which they their concerns about various aspects of their portfolios could have been expressed?  Again, Taleb acknowledges that moral hazard increases the probability of Black Swans, and for very large quasi-governmental organizations as large as Freddie Mac and Fannie Mae to be firing whistle-blowers, especially chief risk officer whistle blowers, is a non-trivial event when they are not subject to normal market pressures and incentives.  If, as government-backed financial institutions such "internal" prediction markets had been transparent to the public, who knows what kind of additional information would have been available how early?

A more developed HPI futures market might also have changed outcomes.  In Nov. 05 the Housing Price Index (HPI) peaked and began a long decline.  In May 16, 2006, the Chicago Mercantile introduced a futures market based on the Housing Price Indices, but it grew in volume slowly (as new futures products often do).  By mid-July of 2006 there were only a few dozen open contracts, worth a few million dollars, in the major metro areas for which the market existed.  One blog commenter complained as of Nov. 21, 2007,

There two problems wih using CME Housing futures data to analyze the housing market (1) Lack of free historical data (if anyone can resolve this, let me know) and (2) the relative illiquidity of the marketplace due to its size. The Fed Board of Gov's published a nice report on using housing data see: http://www.federalreserve.gov/pubs/feds/...
It's a bit technical (and now a little dated) but they make good points about using CME data. Note that the open interest in SD housing futures was only 58 contracts or $3.6M which obviously wanes in comparison to the total SD real estate market or any other large price setting market mechanism (e.g., stock markets or currency/commodity futures). Note that open interest today (11/21) is only 72 contracts.

http://piggington.com/august_case_shiller_hpi

The fact that historical data was not freely available and that the market was relatively illiquid as recently as last November does not imply that the futures in HPI does not provide valuable information, it only implies that it was a relatively new market which is still in a growth phase.  At the time, by the way the HPI futures at CME were predicting a decline on into 2011.

To take yet another new information source, Realius is a real estate prediction market launched last year in San Francisco and now available for five regional markets,

http://www.realius.com/

Unlike the HPI, which calculates housing data through a formula based on historical data, which may introduce various inaccuracies vis-a-vis actual prices, Realius is based on the predictions of on-the-ground local real estate experts estimating the sales price of individual homes, a far richer source of local knowledge than is available through the HPI.  The founder of Realius suggested that had his prediction market been working five years ago we might have avoided this mess.

I am not claiming that any of these three specific markets,

1.  A hypothetical internal prediction market within Freddie Mac and Fannie Mae
2.  The CME HPI futures
3.  Realius

necessarily would have, on their own, prevented the meltdown.  Nor am I claiming that other factors were not involved; they most certainly were.

Instead, I'm proposing that in the coming decades, as hundreds and hundreds of new prediction markets are developed for very diverse niches, our ecosystem of knowledge will change.  There will still be Black Swans, for certain.  But it is premature to claim that prediction market technologies have failed in this particular instance.  We don't know how much information could have been made available had all three of the proposed prediction markets mentioned above, and dozens more not imagined here, all been functioning for twenty years in advance.

It is important to see prediction markets as a versatile new technology about where the desktop computer was in 1980.  "Gee, look at the cool stuff it does."  And yet compared to today, a 1980 Apple III looks pretty primitive.

Again, this is by no means rigorous; perhaps there are elements to this particular meltdown that are for some reason completely impervious to the endless possibilities of ever more granular knowledge discovery that I'm imagining.  If so, I'd love to hear why.

--

Michael Strong
CEO and Chief Visionary Officer
FLOW, Inc.
www.flowidealism.org

Liberating the Entrepreneurial Spirit for Good

Links to my articles:

http://www.flowproject.org/michael.html

When once you have tasted flight, you will forever walk the earth with your eyes turned skyward, for there you have been, and there you will always long to return.?

Leonardo Da Vinci

[Joel Elconin]

Why is everyone so interested in criticizing other prediction market outcomes, instead of creating

their own measurements?

 

Why analyze things after the fact, you all are starting to sound like historians, not prediction mar-

ket proponents?

 

                                                         Alvin

[Tom D'Eletto]

My 2 cents.
 
It seems to me that asking "Did any prediction markets predict the stock market meltdown?" is the wrong question since prediction markets aren't fortune tellers.
 
There would have had to have been a market that was asking the appropriate question. Perhaps if there had been a question on a market that was something like: Will the DJIA will experience a 400+ point one day drop before the end of Q3 2008? Or Will Lehman Brother declare bankruptcy by the end of 2008?
 
then we could have had predictions for these events or similar events.
 
Now, I don't believe that many people would have been buyers of either of those contracts...but I do think there would have been noticably more buyers of those contracts then there would have been a year or two ago. And that might have indicated something. The extraordinary events of the past few weeks were always going to be long shots at best...but sometimes the long shorts come through.
 
-Tom

[Joel Elconin]

You would be suprised.  Sophiciated investors will wager on any thing, just have a few market-makers

that will provide some liquidity.  All this talk, is after the fact, what good is that for predicction markets?

 

              

                                                  Jerry

 

.

----- Original Message -----

From: Tom D'Eletto

To: Predictio...@googlegroups.com

[Richard Jaycobs]

Re: Did any prediction markets predict the stock market meltdown?

Absolutely and of course, there were markets that predicted the
increasing likelihood of the events of last week.

We all read the news and know that both (the now infamous) short sellers
in the stock market and those trading Credit Default Swaps on the
banking sector predicted an increasing probability that the "events" of
last week would happen.

However, given that there are no meaningful real money prediction
markets, it's a completely self absorbed question to ask, "did any

prediction markets predict the stock market meltdown?"

Mapping the real money market activity back to a "prediction money
binary contract" on, for example, "will Lehman enter bankruptcy in the
next X days" may (or may not) be an arbitrage linked proposition.

So, any prediction market that WAS NOT "predicting" an increasing
probability of the meltdown was, by arbitrage, a completely disconnected
and uninformed market.

The website, electoral-vote.com made an obvious but noteworthy
observation today with regard to Intrade's political markets:

"In general, the betting sites are lagging indicators with respect to
the polls. When new polls come in, the bettors see that and adjust their
bets accordingly. It doesn't work the other way (people responding to
pollsters don't do so on the basis of what they saw on Intrade)."

http://www.electoral-vote.com/evp2008/Pres/Maps/Sep23.html

When I read some of the comments and questions on this googlegroup, it
seems some folks are hoping to demonstrate that prediction markets are
the dog somehow able to lead "event tails".

Personally, I think prediction markets are the tail that follows the
event dog wherever it's going. And that's probably the way it should be.

Rj

[Byrne Hobart]

There's a serious problem with this discussion. You're asking if one market predicted the movements of another market. But since the Dow predicts its own movements (if everyone decides it's going to 5000, that's exactly where it'll go). So what you're really asking is "Were there any arbitrage opportunities between prediction markets and financial indices?" And that's another way to claim that prediction markets are flawed.

Asking a market to predict events that will move the market is a bad idea, and asking another market to predict the returns of the first market is only going to work as long as there isn't arbitrage between them. Prediction markets serve a more useful purpose when they are aggregating information and expediting bets on something other than a giant information aggregator on which people place bets.

[Richard Jaycobs]

Thank you Byrne. You made my intended point more succinctly than I did.

Rj

[Bala Pillai]

Byrne,

>You're asking if one
> market predicted the movements of another market.

Isn't there room for a Prediction Market instrument that addresses "wisdom of the crowd" versus groupthink contentions?

I agree this may not be what happens in the early days of Prediction Markets.

I have been following moral hazard pundit, Henry Liu -- see "Too Big To Fail Vs Moral Hazard" at http://www.atimes.com/atimes/Global_Economy/JI23Dj12.html  -- for years and resonate with his views. And I was not surprised that his well argued predictions of investment banking collapses from years ago, would not get traction in the financial press that matters. I run into that all the time. Why? Groupthink & Myopia.

I disagree with Nigel Eccles that this is a black swan. It is lots less black swan and lots more predictable. The solution is to map what gets traction in media that matters with what gets traction in Prediction Markets, so that there is a greater coupling between the two. I do agree that there is a danger of groupthink that spreads across both the financial markets and Prediction Markets then -- don't know a solution to that yet.

Bottom line on question of this thread: it is very very very early days for Prediction Markets

cheers../bala
Bala Pillai
Sydney & Kuala Lumpur
+61 2 9807 8545
Linked In: http://www.linkedin.com/in/balapillai
Facebook:  http://profile.to/balapillai
.

[Joe Seither]

Dr. Holley,

my two cents:

"Stock market meltdown" is probably not the right way to frame the problem, the damage to the overall market is still pretty well contained with the Dow only 24% off its all-time high and real carnage mostly limited to the finance and insurance sector of firms. The Wilshire 5000 is off just 23% from its all-time high reached in Oct 2007. By official definition and historical standards, were in a bear market, but nothing extraordinary - yet.

The inherent complexity of derivatives and lack of transparency in derivatives markets would make it pretty tough for external prediction markets to operate efficiently and predict the toxicity of the investments and current resulting condition of the financial markets. 

It might be more incisive to ask whether any prediction markets predicted a well-defined point event like the price, in basis points, of credit default swaps or residential. This is a real "commercial prediction market" in action, with prices reflecting the state of relative confidence between the Investment Banks themselves, who are in the best position to assess the relative risk of the underlying investments.

What were CDS prices doing over the last 2 years?

Liquidity will return - the market will redefine reasonable leverage based on a credible balance sheet.

The really troubling long term issue for US markets and firms is the fact that the ultimate guarantor of a lot of US investments - the US Government - is $55T overcommitted versus projected revenues.

www.pgpf.org

[Kristoffer Thomas Hartwig]

Interesting subject indeed.

I wonder if the prediction of "Black Swans" in the event driven
derivatives-department might be linked to the usage of Idea Markets/
Futures in the search for Radical Innovations?

Seems that it's basically the same challenge? Painting Black Swans
grey, if you will.



/Kristoffer Thomas Hartwig

[Jason Ruspini]

"Black Swan" is a rough piece of jargon that's of little application
once one understands the basic concept.

As many have suggested, better transparency and simply having assets
trade on public exchanges would have helped to prevent the crisis.
(which was only partly a failure to predict insofar as payoffs were
skewed to the upside for executives.)

Conditional contracts such as the chances of AIG defaulting given a
Lehman default or on relations between broader macro variables might
also be interesting as they can lay bare potential non-linearities as
a crisis sets in. This is similar to how implied volatility curves
show that a further X% decline is more likely given a X% decline,
etc. The 1987 crash was difficult to predict in this sense of non-
linearity and magnitude, not that the market was likely to go down.
In particular cases such as those of individual firms, such feedback
loops can lead to irreversible outcomes.

Jason Ruspini

[Bala Pillai]

I wrote earlier:-

>Isn't there room for a Prediction Market instrument that addresses
> "wisdom of the crowd" versus groupthink contentions?

In view of below, perhaps we can call them King Charles instruments.

cheers../bala
http://profile.to/balapillai

http://www.atimes.com/atimes/Letters.html

During the high Middle Ages, members of the nobility fought each other usually for land and gold but often just for the sake of fighting. Usually the armored knights would not be killed in battle but were captured and held for huge ransoms, extorted from their subjects by other knights. Famous examples are King Richard of England ("the Lionheart", ransomed twice), and John II, King of France, each captured after needless battles, lost due to their own stupidity. Often the ransom included giving huge chunks of their territory to the enemy. The unarmored infantry and attendant troops, not worth ransom, were usually massacred on the spot. While prisoners, the nobles were treated with every courtesy and consideration, being called "brother" by their battlefield enemies and made fully at home. The epoch was sung and praised as a wondrous time for high deeds of valor by their troubadours.

Nobody considered the multiple disasters caused to all others by their incompetence: war, ruin, famine, disease, ignorance and the black death that destroyed half the population. "Ransom us" cried the kings, "or the kingdom is ruined!" But they had already ruined it, had they not? Very exceptionally, King John stayed a guest at the English court for 25 years because his son Charles refused to pay ransom (which was to surrender half the kingdom to the English). France recovered and eventually won the Hundred Years' War.

In modern times the new nobility of Wall Street and corporation barons engage in needless battles (mergers and acquisitions that for the most part have gone bad, the globalization and outsourcing frenzy), wreck their own companies (some recent examples are Enron, GM, Ford, and now Wall Street giants), and then demand bailouts from their subjects (taxpayers) extorted by other barons (lobbyists, Congress) who have expectations of similar rewards when it's their turn. They receive wonderful treatment ("golden parachutes") while their infantry (all those below VP ranks) are massacred (unemployed, lose pension plans). The epoch is sung and praised as a wondrous time for high deeds of finance and "freedom" by their troubadours in the media. Nobody considers the multiple disasters caused to all others by their incompetence ... "Ransom us" cry today's barons, "or the economy is ruined!" But they have already ruined it, have they not? Is there a Charles today to ensure that John does not return to the kingdom and complete its ruin? Nobility and especially royalty "learn nothing and forget nothing". History's corrective action has been revolutions, usually violent ones that often make things worse in the short term.
Kali Kadzaraki
Houston, Texas (Sep 25, '08)