There's something funny going on over at Intrade with respect to the pricing of the Obama and McCain contracts.
Right now, Obama is trading at 52.3 points. That is, Intrade implies that he has a 52.3 percent chance to become the next President.
Now, I happen to think that is a patently absurd price. But you don't have to take my word for it. Over at BetFair, another large UK-based gambling and futures site, you can also buy an Obama contract. But the price there is 1.62, which implies a 61.7 percent chance that Obama will become the next President.
That is a huge spread, 51.5 points versus 61.7 points. This is the equivalent of the Giants being 3-point favorites at the Bellagio Sportsbook, and 7-point favorites at the Mirage down the block. Those things just don't happen in efficient, sufficiently liquid markets, because they create arbitrage opportunities: you'd lay $10,000 on the Giants at the Bellagio and $10,000 on their opponents at the Mirage. Any time the Giants win by fewer than 3 points or more than 7 points, you lose nothing, since your two bets cancel out. But any time they win by fewer than 7 points but more than 3, you win both bets, and take home $20,000 (less the casino's vigorish) for absolutely no risk. Pretty good deal, right? That's exactly what's happening with these futures contracts.
It does seem to be Intrade specifically that's out of line, rather than Betfair. At Iowa Electronic Markets, yet another political futures exchange, the probability of the Democrats winning the popular vote is about 61 percent. They don't have an electoral vote contract, but if they did, presumably that number would be a little higher because of the structural advantages Obama has in the electoral college this year that we've discussed here at length.
In fact, the Intrade pricing doesn't even seem to be internally consistent. If you look at their pricing in individual states, they have Obama at no lower than 60 percent or so in each of the Kerry states, as well as in Iowa, Colorado and New Mexico. And Virginia, Nevada and Ohio are all at about 50:50. The relationships between the odds of winning any given state and the odds of winning the electoral college are difficult to determine, but I'm pretty sure that Obama should be higher than 51.5 percent given those parameters.
Finally, take a look at the actual pattern of the trading within the last 36 hours:
It's pretty obvious that this is not some sort of random walk. Rather, every so often, some individual trader or some small group of traders are shorting all the Obama contacts in bulk and resetting the entire market. The markets then organically climb back upward until the rogue trader strikes again six or eight hours later. The volumes on these contracts have been very high for the past week as a result.
Most likely, this is just some idiot degenerate gambler who is trying to have some fun. Between the Obama and McCain contracts, there appears to be about $400,000 in contracts changing hands every day, which is a lot by Main Street standards, but minuscule by either Vegas or Wall Street Standards.
What's a little weird, however, is that this rouge trader is not only selling Obama contracts and buying McCain contracts .... they also seem to be buying Hillary Clinton contracts:
This is the exact mirror image of the Obama trading -- in fact, the trades seem to be occurring at exactly the same times. So someone is betting on some sort of disqualifying event happening to Obama.
I don't think this is any cause for alarm... if Joe Biden contracts were being bought up as part of this scheme, that might be more concerning, but they aren't. Still, if I were the Secret Service FBI (**), I would probably want to know the identity of this trader. And if I were one of those Beltway pundits who thinks I can be hip by referring to Intrade, I would take a look at the gross inefficiencies in these markets and think better of it.
(**) It seems to me that too much attention is being paid to this particular point. No, I do not think it is at all probable that this is connected to some sort of physical threat to Obama. But I do think it is the job of the FBI to investigate improbable threats. This is the whole reason the Pentagon was arguing for terrorism futures markets a couple of years ago.
9.24.2008
Intrade Betting is Suspicious
@ 4:54 AM
Subscribe to:
Post Comments (Atom)
542 comments
What's the probability it's just a load of PUMA nutcases? Frankly, I'm surprised that Hillary stock is still that high.
Intrade is not an oracle - it is far more flawed than the worst polling, possibly even worse than Zogby Interactive.
It's interesting that your first reaction was about the Secret Service, i.e. a physical threat. My first impression was something entirely different: a scandal that Unknown Trader believes will break and kill Obama's chances, leading the Democratic Party to hold an emergency replace-Obama-with-Hillary-to-salvage-the-election meeting. After all, as you said, the physical-threat idea would be more worrying if they were Joe Biden contracts being bought, and there were quite a few news articles discussing the purchase of "McCain drops Palin from the ticket" contracts after the convention, providing precedent...
Daniel,
I don't think any sort of physical threat to Obama is at all likely. I do think that it's the Secret Service's job to investigate unlikely threats, however.
Hmm.
Several things seem interesting. First, are you suggesting that we place the election betting market to win either way? Second, getting the Secret Service involved this seems a little odd. The FBI, sure, maybe even the NSA. The SS doesn't really cover this sort of thing.
On the other hand, this stuff is a little odd.
Nate Silver, you are one sharp dude!
The Free Marketer in me says the problem is too much stock in the hands of a few (or one) traders and that the solution would be to reduce barriers to entry... of course given that it's intrade, the barriers to entry in the market is what, like a dollar? Given this, I think I agree with your nutso rogue trader theory. Still, I think it'd be a mistake to discount the predictive power of these markets the closer we get to the election, although then again, you can say that about the polls you guys track, too.
What about the possibility that it is someone who can afford to do this, and who wants to keep this indicator of race dynamics look as close as possible? Not too much money for someone super rich who wants to make McCain look like he is doing better than he is.
How could there be funny business involved? If there's actually someone betting on HRC becoming president, they're incredibly stupid, or:
1. There's a huge scandal about to hit Obama AND Mccain
2. Obama and Biden are both going to be assassinated
There is no other way she would have a chance. A scandal knocking Obama out of the running would taint the democratic party and no one from it would win this year.
betting markets are dangerous, it's real money and people don't throw it away like they were trolling on a blog. I think Nate has great knowledge of sports betting markets and understands these must be regulated and controlled just to avoid frauds. So his call to the ss must be taken seriously.
Thanks Nate, I've been observing the Intrade patterns lately and did get suspicious as well. The same is going on with the Rasmussen market.
Some may have insider's informations about polls to be released, I can tell the outcome from incoming poll result by watching the directions of the Intrade and Rasmussen market.
I believe though, someone from the McCain camp is doing these trades even though is loosing money, just to gave the appearance that the race is close.
The betting markets, have been the most accurate way to predict the winner, thus the need to manipulate them.
LOL! I love how "idiot degenerate gambler" links to Mike Matusow's wikipedia. Classic Nate.
I suspect it could be someone over on the Republican Side, just trying to keep McCain's numbers from sliding too low, in an effort to prop up his credibility as a viable candidate... that is, to sort of help define the narrative of a competitive race.
As we speak the usual early morning Obama's short selling is just happening at Intrade, Obama just went down 3.3 points
Keep deleting 'em, Nate.
I'll keep posting the chart of the Clinton for President contract since she dropped out of the race.
Who should the Secret Service be investigating for the Clinton spike in early July, you or Marty Peretz?
Or is the more serious question whether or not Nate has a positon on Intrade he's trying to influence?
RCP just listed the ANC/WO poll giving Obama the 9 points lead, strangely this coincides with the 3.3 loss for Obama at Intrade.
Cold any one working at RCP be the Intrade manipulator? I would not be surprised.
Clearly the SEC's new anti-shorting rules have forced short sellers below ground into Intrade! XD
I am a bit confused.
I have asked before if it was legaly possible for Palin to be replaced (legaly not politicaly), and the general opinion seemed to be no. Is is not too late to substitute Hillary? Are there any precidents?
I was alienated from feminism at Vassar, precisely because it became clear that the business there was less about eviscerating notions of gender than it was about white women claiming an oppressed identity.
I didn't think that exclusivity affected me at first...not until years later, when I was working in a prison, and I saw how devastating conceptions of gender were for these men, who would die on a dime just to avoid being emasculated. And later when I went to teach in after school programs, I'd see young boys internalizing the same destructive lessons about what a man is supposed to be.
It made me angry, because I suddenly realized how important feminism was for men.
You can open an account at Rasmussen market for free you get 10,000 dollars to play with.
Rasmusen markets tend to follow Intrade
I wonder if there is any possibility that this is someone from the McCain campaign trying to make the election look closer than it really is? I don't know how possible it is, or how legal it is, and maybe backing Hillary rather than McCain gives them some form of cover? Or maybe its just some poor sad lonely nut having some fun!
As I understand the legal position, if something were to happen to Barack Obama, then Joe Biden would become the Democratic candidate? Though I am not entirely sure how the rules of the DNC cover this?
YES. Intrade is generally rubbish for futures. Betfair has much higher volume and seems to be fluid with events. You can look at the graph of prices and pinpoint the actual events that triggered each correction. Not to mention the interface isn't god awful...
Second thought. Just how well controlled is this market? Is it possible that this is somebody selling to himself?
He has 2 accounts. One sells Obama short in big batches hoping to cause a panic. The other sells back quietly in small batches.
If it is done carefully it should be posible to make a steady profit. He just buys a bit more at the peak of the panic & sells after it settles back.
"I wonder if there is any possibility that this is someone from the McCain campaign trying to make the election look closer than it really is? I don't know how possible it is, or how legal it is, and maybe backing Hillary rather than McCain gives them some form of cover? Or maybe its just some poor sad lonely nut having some fun!"
The Obama, Clinton, Giuliani, and Paul contracts have all traded far too high at various points during this campaign season, because all of those campaigns have had a disproportionate number of supporters with more disposable income than rationality.
Similarly, Dean contracts traded too high in '04 for the same reasons.
All trades have a winner and a loser. If a candidate is good at fundraising, they're likely to have true believer supporters with enough disposable income to make losing trades on Intrade.
Cue the Secret Service!
For those interested in more insight on the structural oddities of futures markets, check out www.worthwhilemoney.com
---Or is the more serious question whether or not Nate has a positon on Intrade he's trying to influence?---
One, this is an inane thing to postulate given the nature of this site and the future value to Nate of being an accurate predictor of political outcomes. Two, the existence of a wide disparity in spreads provides opportunities for arbitrage Nate is more than qualified to notice and take advantage of if he chooses.
reminds me of Hanlon's Razor - "Never attribute to malice that which can be adequately explained by stupidity."
this could simply be a true believer with some cash, who thinks they know better than the polls, could it not?
Most likely, the Obama contracts are being dumped by one or several McCain supporters who then buy up Clinton contracts to stoke the (now-discredited) theory that white women are going to vote for McCain because they "feel slighted."
A hell of a lot is at stake in this election, and if a market can be manipulated (as Intrade can) it's gonna be.
Thanks, Nate, for your eagle eye on this.
striac
A stupid person with that much money?
I vote for malice.
PPB, I'm confused. What are you actually trying to argue?
LOL @ Matusow reference. Definitely betting my house imo.
"One, this is an inane thing to postulate given the nature of this site and the future value to Nate of being an accurate predictor of political outcomes."
Meh. One can walk away with $100k+ in a week if one correctly plays the heavy volume markets.
Do that for the length of the campaign, and it blows Nate's future value away. Plus, he can always change his tune in the election's final week and have the correct results for posterity.
"Two, the existence of a wide disparity in spreads provides opportunities for arbitrage Nate is more than qualified to notice and take advantage of if he chooses."
But, of course, those markets are much easier to profit from if one is able to influence them via some website that gets pickup in the wider media.
------
FWIW, if I had to bet, I'd bet that Nate probably isn't playing Intrade and manipulating his site to profit.
But the idea of Nate pumping and dumping is about seventy times less wacky than Nate's idea that it's the Secret Service's bailiwick to be worrying about the Clinton for President market.
---Most likely, the Obama contracts are being dumped by one or several McCain supporters who then buy up Clinton contracts to stoke the (now-discredited) theory that white women are going to vote for McCain because they "feel slighted."---
Donating the money to the RNC would be a great deal easier than churning through transaction fees for no particular reason.
This is most likely just the relative illiquidity of the market combined with the demographics of the Intrade customer as Nate postulates. Actual McCain manipulators would be far more likely to drive the contracts into showing an advantage for McCain generating news fodder of "the smart money" seeing McCain winning regardless of polls.
"PPB, I'm confused. What are you actually trying to argue?"
That this particular post of Nate's is loony in several ways.
---But, of course, those markets are much easier to profit from if one is able to influence them via some website that gets pickup in the wider media.---
No, arbitrage is by definition easier than trying to manipulate an individual market for personal gain. Arbitrage involves no risk.
---But the idea of Nate pumping and dumping is about seventy times less wacky than Nate's idea that it's the Secret Service's bailiwick to be worrying about the Clinton for President market.---
I'd say they're about equally wacky. Although I'd agree that the fact that FBI lacks any jurisdiction or authority to investigate the records of market in Dublin certainly adds to the wackiness.
LOL @ the insinuations that Nate's trying to manipulate the market for personal gain. If Nate was at all interested in making money on intrade, this post would never have been made. Do you see why?
Anyone freaking out about potential threats to the safety of Obama should check out Intrade's contract-specific rules:
"In the event of the death of one of the candidates the following steps will be taken:
1. All contracts in the group will be paused and expired at the last closing price set for each contract
2. A new market will be listed for the remaining candidates"
Although I'd agree that the fact that FBI lacks any jurisdiction or authority to investigate the records of market in Dublin certainly adds to the wackiness.
Actually, the fact you think the FBI has no jurisdiction or authority to investigate the records of a market in Dublin is much more wacky.
International manipulation of gaming over the Internet would usually cross into multiple countries jurisdiction. If anyone involved in a market manipulation of any sort lives in the US, the FBI has jurisdiction. If there is anything resembling a threat to any Presidential candidate, the FBI or SS will have jurisdiction.
You concede jurisdiction far below where you should.
And finally, it's worth noting that if someone wanted to try to pin Obama's value at its current level contrary to market demand, they'd probably have to throw a minimum of $100k per day down a well.
If the McCain campaign thinks that throwing $3m - $10m a month away on this project is the best investment of their money, folks should say godspeed to them.
Of course, one who follows Occam's razor would suggest that the market simply doesn't agree with Nate's assessment of the election at the moment, rather than proposing bizarre assassination theories that require Secret Service, FBI, NSA, and MLB investigations...
Nate,
I have noticed the same thing over the past couple of weeks. Someone is playing with the Intrade presidential market in order to favor McCain's odds. I'm an equity (stock) trader who is all too familiar with manipulation of markets and it wouldn't take much $ for an experienced trader to push these numbers all over the place. There's simply too much money involved here to expect it not to happen (Think many billions of $ are at stake over energy policy?). Of course it doesn't make a whole lot of sense/cents as the polls should be driving the betting and not vice-versa, but given the number of pundits tracking Intrade, it makes sense to the manipulators to make every effort to influence the direction of this race. Even a losing effort like this.
---Actually, the fact you think the FBI has no jurisdiction or authority to investigate the records of a market in Dublin is much more wacky.---
False.
---International manipulation of gaming over the Internet would usually cross into multiple countries jurisdiction. If anyone involved in a market manipulation of any sort lives in the US, the FBI has jurisdiction.---
Country's. Also: False. What law is it you think they would be violating? What mechanism is you think would be available to DoJ to access records at Intrade?
---If there is anything resembling a threat to any Presidential candidate, the FBI or SS will have jurisdiction.---
False.
---You concede jurisdiction far below where you should.---
False.
Think this through for just a moment. For DoJ (or whomever) to establish that the imaginary trader in question resides in the US, they'd have to access the records of an Irish corporation. Intrade could of course *voluntarily* cooperate, but there is simply no mechanism to *compel* them to. Nor is it particularly likely that another law enforcement or regulatory agency would force Intrade to comply.
---Of course, one who follows Occam's razor would suggest that the market simply doesn't agree with Nate's assessment of the election at the moment---
You're missing the point. It's not the Intrade market disagreeing with Nate, it's the Intrade market disagreeing with other markets.
Looking forward to the deletion of troll/hack "peteypowderblue"s inane straw men.
Bottom line: The money manipulating Intrade is well spent (in the rogue traders' eyes) if it influences the pundits to present a narrative that the race is closer than it actually is.
That's what going on with Intrade.
One only has to look at the myriad of polls on the right hand side of this website. There are different views of the same event. I am sure Betfair has a European bias and Intrade has a US bias, just like each poll has a bias for thier weighting
And finally, finally, it's also worth noting that the Obama and McCain contracts for President have been trading for a while now as if the Clinton for President contract were trading at zero.
If you add up the bids on Obama and McCain at a randomly chosen point in the day, you'll generally find that they add up to 99. Given that Intrade takes one percent off of winning long contracts in commission, this means that they Clinton for President contract is being completely ignored.
A similar phenomenon occurred during the GOP primary race when the Ron Paul contract eventually got ignored by the rest of the market.
Wow. This site is heading off a credibility cliff to explain Intrade via a lone "rouge trader" causing Obama to not look as likely as you all think he is.
Hysterical.
Here's something relevant to actually talk about, Obama springs to 9 point lead nationally ABC/WaPo (one caveat - 17 point Dem advantage ID with leaners, likely off, see Gallup yesterday)
More voters trust Obama to deal with the economy, and he currently has a big edge as the candidate who is more in tune with the economic problems Americans now face. He also has a double-digit advantage on handling the current problems on Wall Street, and as a result, there has been a rise in his overall support. The poll found that, among likely voters, Obama now leads McCain by 52 percent to 43 percent. Two weeks ago, in the days immediately following the Republican National Convention, the race was essentially even, with McCain at 49 percent and Obama at 47 percent.
http://www.washingtonpost.com/wp-dyn/content/article/2008/09/23/AR2008092303667.html?hpid=topnews
I stopped paying attention to the intrade numbers awhile ago when I realized that too many people look at is a gambling. The contracts are simply a game of chance and people started buying up McCain when he was really low on the "chance" that he'd win therefore giving them a bigger return. When someone starts thinking from angle of trying to maximize gains rather than buying the position that is motivated the current state of the race, then reality gets edged out of the equation.
It's like doubling down your twelve against a dealer showing a face card. That's not a reality based bet, your just looking to score some gains.
---Wow. This site is heading off a credibility cliff to explain Intrade via a lone "rouge trader" causing Obama to not look as likely as you all think he is.---
Wrong. That wasn't the point of the post, *at all*. Give reading it a shot and see how that goes.
"The money manipulating Intrade is well spent (in the rogue traders' eyes) if it influences the pundits to present a narrative that the race is closer than it actually is."
You need to add up the actual dollars required to execute such a strategy.
It's simply not cost-effective.
And that's why Intrade is an interesting market to watch. The campaigns can't afford to really move values around in the high volume markets like President.
"When someone starts thinking from angle of trying to maximize gains ... then reality gets edged out of the equation."
You may not be fully clear on how markets work in general...
---I stopped paying attention to the intrade numbers awhile ago when I realized that too many people look at is a gambling. The contracts are simply a game of chance and people started buying up McCain when he was really low on the "chance" that he'd win therefore giving them a bigger return. When someone starts thinking from angle of trying to maximize gains rather than buying the position that is motivated the current state of the race, then reality gets edged out of the equation.---
All markets are an exercise in maximizing return. This doesn't make them unable to provide price discovery.
I did read it Poker, carefully. Total hysterical inanity. :)
For example:
It's pretty obvious that this is not some sort of random walk. Rather, every so often, some individual trader or some small group of traders are shorting all the Obama contacts in bulk and resetting the entire market. The markets then organically climb back upward until the rogue trader strikes again six or eight hours later. The volumes on these contracts have been very high for the past week as a result.
Its "pretty obvious" that there is a rouge trader holding Obama down? Be happy Obama is up in the polls and on Intrade, quit being insanely paranoid.
---Its "pretty obvious" that there is a rouge trader holding Obama down? Be happy Obama is up in the polls and on Intrade, quit being insanely paranoid.---
It has nothing to do with Obama. It has to do with one market not reflecting other markets. If Intrade had Obama at 90% it would be as strange. If one were able to buy GM stock for $2 per share in Tokyo but at $4 per share in NY, it would be as strange.
Understand? It's the disparity of the relative value of the Intrade options when compared to other prediction markets, not the fact that the Obama options are priced lower. Nate makes this abundantly clear when he briefly offers the sports betting arbitrage metaphor.
When there is a disparity in markets, something is occurring. The most likely things to be occurring are that the out of synch market is much more or less efficient than the other markets, is being manipulated for whatever reason, or is subject to a large amount of irrational trading.
One of these is very likely happening at Intrade. The most likely of these would seem to be irrational trading driven by the demographic political leanings of traders there.
I found the analysis of what's happening on Intrade brilliant, quite apart from the questions of jurisdiction and motivation that most of you are debating. I might add that it shows incredible forbearance for Nate not to throw out the asses of the posters who dare to slander him on his own site. What's YOUR motivation for accusing HIM of being the rogue trader?
Its a 10 point spread, according to nate, 50-60%, between the markets.
Markets disagree. That's where arbitrage comes in. Money if fungible. If McCain was trying to fix Intrade, a laughable possibility in and of itself, traders would crush the plan by selling obama at 60 on the other two sites and buying obama at 50 on intrade. Money is fungible.
PeteyPowderBlue said...
You may not be fully clear on how markets work in general...
Sure, I'm not avid follower of markets. But intrade isn't a pure market at this point. It's a game of chance now. You have (at this point) two positions you can take and one of those positions will be correct. Investing takes some thought. Seeing McCain at 40 and thinking "well hot damn, I can make more off of that than Obama" isn't a strategy based on reality. It's a strategy based on maximized returns. That mindset his infiltrated intrade with it's new-found popularity over the course of this election. It might not be the majority mindset of the traders there, but it's enough to throw things out of whack. The numbers are no longer sound because that "market" is tainted. Therefore, I ignore them.
I dont think Nate is the rouge trader, there is no rouge trader.
hehehheh
---I might add that it shows incredible forbearance for Nate not to throw out the asses of the posters who dare to slander him on his own site. What's YOUR motivation for accusing HIM of being the rogue trader?---
I think the actual implication was that he was trying to drive the price of Obama options higher by projecting Obama as more likely to win than the data would suggest.
I'm reasonably certain no one was suggesting that Nate would be intentionally driving the price of the Obama options down.
I don't understand. Am I missing something, or is there easy money to be made by shorting Obama on Betfair and backing him on Intrade?
Exactly my point Ian.
---Sure, I'm not avid follower of markets. But intrade isn't a pure market at this point. It's a game of chance now. You have (at this point) two positions you can take and one of those positions will be correct.---
This is true of options on anything. Stocks, indexes, commodities, etc. etc.
---Investing takes some thought. Seeing McCain at 40 and thinking "well hot damn, I can make more off of that than Obama" isn't a strategy based on reality. It's a strategy based on maximized returns. That mindset his infiltrated intrade with it's new-found popularity over the course of this election. It might not be the majority mindset of the traders there, but it's enough to throw things out of whack. The numbers are no longer sound because that "market" is tainted. Therefore, I ignore them.---
Due respect, the argument you make is silly.
Can some explain exactly how the prediction markets are pulling this off? I imagine that there is someone with a pool of Obama-to-win shares, lets say 1000. Every time they think the price is too high, they dump some shares and the price drops. Then other people start buying back shares and the price slowly rises. Repeat.
At some point, won't the manipulator run out of shares to sell? Is there someone with a HUGE pool of Obama shares that they are slowly draining to keep the price constant? Or is my understanding of how Intrade works incorrect?
Geoff posted:
"I dont think Nate is the rouge trader, there is no rouge trader."
Geoff, have you been to any cosmetics stores lately?
For instance right now the high bid on obama is 49.3. Anyone who wants to print some free cash can buy up obama at 49.3 and sell obama at 60 on the other two sites. Print money all day and all night, from here to eternity.
Go for it folks, see if you can pull it off. :)
heehe yea i can't spell in the morning
I tried registering at Betfair, but then discovered they do not accept bets from people in the US. Does that affect their market for US politics? Perhaps people outside the US have, on average, a different view of the US presidential election?
---I don't understand. Am I missing something, or is there easy money to be made by shorting Obama on Betfair and backing him on Intrade?---
Nope. Buy McCain on Betfair at 40, buy Obama at Intrade at 52, profit.
The problem, of course is how liquid both markets are.
The most likely cause I see is someone who ALREADY shorted Hillary is now transferring his money (at a slight discount compared to waiting for the contract to expire) by covering his Clinton position. As his money is freed up, he puts it back in the market by shorting Obama.
It might sound silly to you because you're more versed in it. But to ignore that you've got people who have that same silly idea in the markets taking positions on a blind "i wanna make some easy money" mindset is purely ignorant on your part.
GP - volume today in a few hours is 5044 contracts. Too much volume to control, but you wouldnt need "extra contracts". You can sell as many Obama contracts as you want at any price....i suppose you'd just keep on lowering your ask price until there is no demand.
Remote at best.
[b]
I tried registering at Betfair, but then discovered they do not accept bets from people in the US. Does that affect their market for US politics? Perhaps people outside the US have, on average, a different view of the US presidential election?[/b]
Perhaps.
@gprimos1, you can short a position withot previously having bought that position, up to the limits of your capital, which is needed to pay the contract if you lose.
---It might sound silly to you because you're more versed in it. But to ignore that you've got people who have that same silly idea in the markets taking positions on a blind "i wanna make some easy money" mindset is purely ignorant on your part.---
No, it's not. What's ignorant is that you don't seem to realize that this is how ALL markets work. People buy things they think are priced below value and sell those they think are overvalued, be they wheels of cheese, stocks, bonds, barrels of oil, or option contracts.
Betfair's "No Americans" stipulation does not prevent arbitrage. It just means Americans can't do it. The market should still adjust properly as European traders short Obama on Betfair and buy on Intrade.
Maybe there just aren't any European arbitrage agents with enough money to counter this?
The thought once crossed my mind that the email hacking was a setup by Republicans to frame Obama. (I highly doubt it, but it's just one example of a way Obama could be pressured to step down before the election.)
---Maybe there just aren't any European arbitrage agents with enough money to counter this?---
More likely there are real liquidity questions. The entire market here is really extremely small change to arbitrageurs.
Thanks DVD...
So if my understanding is correct, for us to see this kind of trading pattern, there would have to be some VERY big investor who had earlier bought a huge number of Obama shares when they were low, lets say at when they were $10 in '07. This investor has recently decided to cash out of the market. They want to keep the price they get high so they stop selling once the price drops below $50. For this to produce this kind price stabilization, they would have to be selling at the same rate that the "normal" traders are buying to raise Obama's price to the "true" price of $60. If my theory is correct, at some point this investor will have run out of shares to burn, and the market traders will buy Obama up to $60 to match the other markets.
"Value" doesn't enter the mind of the people who are blindly diving in with no idea other than "X is lower than Y, so I'll take X". That is what I'm trying to say here, but you aren't getting it for some reason. There are people out there who are not looking at the real value of a position, but a perceived chance at making money because they are only looking at the relative values of the two positions. If you have people who are playing off of a perceived value rather than a real value, then the market odds can't wholly be considered a good indicator of a prediction for the outcome at that point.
It's like day-trading and people who used to drive up the price of tech stocks. They didn't really know what they were investing in, they just thought "oh, well I can make money there." Qwest, 3Com or Google siting at $400 a share no longer indicated the real value of that company. If enough people jump on McCain at 40 because it's a hot buy they hope will have good returns and it drives McCain up to 50, it doesn't reflect that McCain has a 50% chance of winning. Again, this is why I ignore the intrade numbers.
GP-
That explanation fails to account for McCain at 48.
"The entire market here is really extremely small change to arbitrageurs."
Well, it's small change to the high-end Wall Street folks. But it's pretty real scratch in normal terms.
The Presidential winner market alone is trading over half a million dollars a day...
Intrade is garbage. Americans aren't able to buy options on the Intrade market, so I fail to see how it can accurately reflect the American candidates' political chances. I have an interesting project for you nate: try to correlate breaking US events with changes in Intrade stock. My guess is you'll likely see a 6+ hour delay in Intrade price shifts. I'll leave the conclusions to you, but my takeaway is that Intrade has too isolated of a potential market, so it doesn't accurately reflect the outlook of a open market.
Weird. I always have an Intrade tab open and had noticed this, to an extent. I'm not so savvy on i-trade,, but saw Obama dip to below 50 last night for no reason and was wondering how much volume it would take for one person to affect movement to a significant degree.
---"Value" doesn't enter the mind of the people who are blindly diving in with no idea other than "X is lower than Y, so I'll take X". That is what I'm trying to say here, but you aren't getting it for some reason.---
Oh, sorry. I see now. Just kidding, I understood your point immediately. It's not that there's confusion, it's that you're unequivocally wrong.
---There are people out there who are not looking at the real value of a position, but a perceived chance at making money because they are only looking at the relative values of the two positions. If you have people who are playing off of a perceived value rather than a real value, then the market odds can't wholly be considered a good indicator of a prediction for the outcome at that point.---
There is no "real value". Markets are price discovery mechanisms.
---It's like day-trading and people who used to drive up the price of tech stocks. They didn't really know what they were investing in, they just thought "oh, well I can make money there." Qwest, 3Com or Google siting at $400 a share no longer indicated the real value of that company.---
The price people pay for something *IS* the "real value". There is no imaginary "true value" of anything.
---If enough people jump on McCain at 40 because it's a hot buy they hope will have good returns and it drives McCain up to 50, it doesn't reflect that McCain has a 50% chance of winning.---
Of course it doesn't. That's not the point of the market.
---Again, this is why I ignore the intrade numbers.---
Ok, ignore away.
GREAT post!!!! It will be the Secret Service who investigates if it is percieved as a threat to Obama.
Looks like a repug operative play to me. The repug is trying to get a better number on RCP and other repug sites, and breed disension in the dems by inflating Hillary. Might be Karl Rove or Schmidt themsleves, seems like their kind of manipulate the world play.
Asholes, luckily us Dems have big brains like Nate's to catch them.
Nate - did you post a bet on BO? You sure should have! I thought BO was low on intrade for days.
Note that Intrade also has generic party contracts. The Democratic Party tends to score 3 points higher than Obama for the Presidency, so yeah, it's clear there are high-volume traders on Intrade who are betting as though they believe Clinton could somehow win.
Since PUMA stubborness is based partly on the perception that Clinton still has a shot, we can speculate that
A. Boosting Clinton's share price, and the generic Democrat share price, could juust maybe help McCain.
B. If this Intrade weirdness goes away, it'll be a good sign for Obama: not because Intrade is being a good predictor, but because it would show that the perception of Clinton's chances has shifted, potentially leading to PUMAs coming home.
Personally, I think a lot of the people polling as PUMAs are actually partisan Democrats whose bluster won't match their votes--yet another reason Obama may exceed his poll numbers.
geoff-
OK I see your point. For the tinfoil hat explanation to be correct, the GOP "operatives" would have had to have bought Obama at some time in the past and are now transferring their shares to McCain to stabilize the price. In any case, at some point they are going to run out of shares and these big drops are going to stop.
Nate - I always appreciate your attention to the details. I never would have stumbled on this.
In the arena of details, though, I got a giggle from the typo in the paragraph above the Clinton chart.
I'm taking it that should be 'rogue' trader and not 'rouge' trader (selling make-up on e-bay? or maybe suggesting the trader is a pinko?).
Thanks for your good work. You're allowed a typo now and then.
I REALLY hope this can be investigated and we can figure out who is doing this. Times Online - please go get them!!!!
Tito: You're missing the point. Yes, there are idiots throwing around money without thinking. That's enough to throw off the price of one market. But when two different markets have a ten percent disparity in their trading prices, it means that something weirder is going on.
Let me give a more straightforward analogy. Let's say there are two textbook stores next door to each other. Outside one, there is a sign proclaiming "WE BUY PHYSICS TEXTBOOKS! WILL PAY $60". The other has a sign that says "PHYSICS TEXTBOOKS FOR SALE, $50". You can surmise that one of the shopkeepers isn't thinking very much, but that doesn't explain the fact that none of the people who walk down this busy street every day have bothered to buy all of the books from store B and resell them to store A.
If McCain drops to realistic levels his fundraising and supporter enthusiasm will dry up. $400,000 a day is a small price to pay to maintain ambiguity and prevent this from happening. It costs a lot less than a 30 second TV ad, and pays for itself many times over by keeping funds flowing and volunteers excited. This is an affluent McCain supporter (or group) trying to garble the message that markets are sending.
Nate – as the only chap on the net that seems to be dealing with the Intrade markets in an appropriate statistical way (adjusting for state independence, accommodating the long shot bias in the probability tails and accounting for implied probabilities being only approximately correct at any given time etc etc), let me say that you’ve nailed a couple of points, but a few things need to be clarified.
Firstly – Betfair… don’t bother. It’s a bookies market and by definition it isn’t pure.
Secondly – pay absolutely no attention to the legacy Contender Markets on Intrade – the 2008.Pres.Obama and 2008.Pres.McCain markets. Even though they have (historically at least) deeper trading volumes – most of that is back weighted, and they also seem to be disconnected from the information happening around them and are polluted with legacy contracts related to third party contenders that didn’t make the final nomination cut (like Clinton for the Dems still being traded and Romney for the Reps). Their behaviour often appears to be a random walk, interrupted by an exogenous event that, which to my mind at least, is usually either a gaming play or an overcooked poll reaction. These Contender markets are shearing markets where newbie sheep full of naivety get fleeced by traders with experience, or in this case, calculators.
You are absolutely correct – the prices on the Contender Markets are absurd, and the absurdity comes from the almost reversion to a fixed mean response these two markets exhibit (think typical cobweb regression model for commodities with fixed long term price and short term fluctuations), where the further away they get from a 50/50 implied probability, the more resistance they face and the more money is pumped into them to bring them back to a near equal split equilibrium. When you see an elastic band effect in a prediction market that runs contrary to every piece of evidence on the table – move on, it’s being gamed or fleeced.
Instead, the party as Presidential winner markets; the President.Dem2008 and President.Rep2008 markets are much more consistent with not only the aggregated polling sims you do here and the raw polling results aggregated at places like Electoral-vote.com, but also the underlying State by State market results which are the smallest and most local unit of information in the vast Intrade universe.
In those Party based Presidential markets, The Democrats are currently running at around 56% as an implied probability of victory - but here, still, there is a flow on effect from the Contender Markets where the Party markets are partially anchored via arbitrage opportunities to the polluted Contender Markets. Not by a lot (and I’d imagine this mostly has to do with the costs involved in keeping the required float in your trading account for what is essentially very smallish ROI’s), but by a fair chunk none the less. A few percent.
If you want more precise information from Intrade – of which there is actually mountains – in the same way that trying to extract richer information from polling is more complicated, so too is extracting similarly valuable info from the States You need to look at the State markets collectively, but you also need to treat them appropriately.
That means adjusting for State non-independence, it means that you must, absolutely, take into account the longshot bias in the tails. Because you need to cover your bets at Intrade, you need to keep a certain amount of money in your account. For this reason, the return on investment in driving a probability from 10% down to 2% via your trades is simply not worth it because of the amount of money you have to tie up in your account for a pissy little return of a few percent – hence the probabilities in the tails get inflated. It’s why you never see probabilities of 1% and 2% until the few days before the election (when people are trying desperately to minimise their losses), and it’s why places like Hawaii which would never in a million years go Republican this election, still exhibit probabilities of Republicans taking Hawaii of nearly 4% and of Republicans winning equally ludicrous places like New York having chances of around 8%.
We can adjust for that in the specification of any simulation of the State markets we use. For instance, if we assume that States with a 10% or less probability of going to a given party are in reality approximately 0% (and adjusting for the other problems mentioned) then the Democrat probability of winning the presidency stands at around 60%
If we tighten that up and assume that States with an implied probability of 20% or less for a given party obtaining victory are effectively approximately 0%, then our probability of a Democrat victory jumps up to 64%.
If we assume 30% and 70% are the cut-off limits, it again jumps up to 69-70%.
The 30/70 rule is important in this case because I can’t find, in the history of Intrade political markets, any candidate ever winning with a probability below 30% immediately before election day. With all the elections that Intrade has participated in and with literally hundreds of candidates over the years with probabilities of less than 30%, yet not a single one ever winning – well that’s longshot bias in action.
Interestingly enough, my State market sims seem to track your Win Percentage fairly well, although with a slight bias toward 50/50 compared to your measure. From Election day minus 40 I’m planning to gradually tighten up my longshot bias specs and the standard deviation on my secondary probability distribution draw for the States in the monte carlo sim (which effectively deal with the uncertainty of a given State probability at any given time). I’d imagine the Sim results from the State Intrade markets will pretty much track your Win Percentage results.
Markets like polls have a lot of info. The headline Contender markets are like online polls – worthless as an information source. The Party markets are like a 500 sample poll done by people that wouldn’t know sampling error from a biased sampling frame; dubious but some value.
The State markets on the other hand contain very rich information that is unfortunately clouded by a bit of tom foolery and superficial bias. But tom foolery of that type and the independence bias can be statistically accommodated. Let us not forget that on the 23rd of September in 2004, the Intrade State markets had predicted every state correctly with the exception of New Mexico and Wisconsin. The polls on the other hand were remarkably different.
On electoral eve, Intrade had every State correct – the polls were still uncertain in a handful of States.
There’s rich information in the State markets that we shouldn’t dismiss simply because the headline markets are berko and full of dodgy behaviour. I’m from Australia so I don’t know if the saying “Runs on the board” has any meaning across the pond, but the Intrade State markets certainly have form by any yardstick.
For all the details of this stuff:
http://blogs.crikey.com.au/pollytics/intrade-data/
Scott Steel (aka Possum Comitatus)
No, he meant rogue.
http://dictionary.reference.com/browse/rogue
Don't question a U of C grad on words very often, you tend to lose.
How much does it cost to manipulate intrade like this? Maybe several K per day? That's alot of money over a long run! So somebody thinks it's worth the investment. Why? My guess is because intrade has high visibility on realclearpolitics...
Of course it doesn't. That's not the point of the market.
Then it's completely irrelevant for people to cite intrade numbers as though they have any bearing or hold any reflection of the state of the race. Finally, we can agree on that.
Folks, seriously, how about some talk about the huge new poll of Obama plus 9?
9/18-9/22 - ABC - Obama plus 9
9.17-9/22 - Ipsos - Obama plus 1
9.1718-9.2122 - Battleground - McCain plus 2.
Strange times....Ras and Gallup should be telling today to see if there is another stronger Obama push underway (beyond the economy bump) or if ABC is an outlier.
Abhas Gupta said...
<< Intrade is garbage. Americans aren't able to buy options on the Intrade market, so I fail to see how it can accurately reflect the American candidates' political chances. >>
Not true. People in the US can participate in Intrade, they can't participate in Betfair. Is that a significant difference that might account for the trading value difference?
The quick answer is that Intrade doesn't have enough participants (or, more specifically, liquidity). You can see this in most of their political contracts, but what Nate's picked up here is a great example: that one better can control a market price like this is a testimony to how badly the online markets work.
A merger between some of these betting sites would be a fine idea...
Possum-
Lots of words, but you do not explain the actual rogue market manipulating trades. The market may not be pure, but we have evidence of actual market manipulation and your whole "smarter than you" explanation seems to ignore this basic fact.
---Well, it's small change to the high-end Wall Street folks. But it's pretty real scratch in normal terms.
The Presidential winner market alone is trading over half a million dollars a day...---
Right. This is virtually nothing.
LOL!
You guys are trying to come up with a rational expklanantion when it sits before like an elephant in the room.
These trades are repug manipulation, why should this surprise anyone? They manipulate facts and lie constantly, it was a Cheney aid who said "we create our own reality"
More repug lies in this market, don't overthink it.
I think the Secret Service needs to look at this VERY closely. LOL!
LOL!
You guys are trying to come up with a rational expklanantion when it sits before like an elephant in the room.
These trades are repug manipulation, why should this surprise anyone? They manipulate facts and lie constantly, it was a Cheney aid who said "we create our own reality"
More repug lies in this market, don't overthink it.
I think the Secret Service needs to look at this VERY closely. LOL!
Sensible Monday -
Nah, I got the market discrepancy point of the post. Nate's sports betting analogy was pretty good on that point. My main argument is that the intrade numbers aren't the Holy Grail of the state of the race as some people around here like to treat them. I'm not trying to be some sort of market wizard (obviously), just trying to make the point that they don't really mean anything to me because there is noise in the system from too many non-experts (such as myself) getting involved.
LOL!
You guys are trying to come up with a rational expklanantion when it sits before like an elephant in the room.
These trades are repug manipulation, why should this surprise anyone? They manipulate facts and lie constantly, it was a Cheney aid who said "we create our own reality"
More repug lies in this market, don't overthink it.
I think the Secret Service needs to look at this VERY closely. LOL!
---we have evidence of actual market manipulation---
No.
---These trades are repug manipulation---
Unlikely in the extreme.
I have no real clue about betting on the internet. Yet, I noticed that there are still some contracts on a Clinton victory. My most likely theory is that there is someone who can´t get off his Clinton bets and is trying to generate some fake buzz.
That doesn´t explain the rather low number for Obama though. It´s currently at 50.2 Obama vs. 47.6 McCain which looks like a popular vote forecast. Even Rasmussen markets have a higher number and they are as pro-McCain as you can be without losing your credibility.
So I came up with a few wacko theories, some coming from Nate..
It could be some gambler (loved the Matusow reference) but he would be trying to fill a bottomless hole.
It could be the McCain campaign trying to give a better impression of where the race is at. That doesn´t change the race though... but the difference between fact and image has eluded the McCain campaign so far (how strangely similar to the Clinton campaign!).
It could be some radical underground cells of PUMAs. But they don´t have the money, I would guess...
It could be Obama himself knowing a big scandal is about to break :-P.
Or it could be, not an accurate reflection of the race, but an accurate reflection of how people perceive the race. The whole media is shouting that it´s close, a nailbiter, within margins of error etc....
Or it could be market failure... maybe McCain is getting bailed out, maybe his stock is about to collapse... THE HORRORS
Gigantic LOL at Mike Matusow link.
My GOD, Nate, My GOD:
You are right! I think there may be someone out there, maybe a white supremacist group that has a plan or contract along the lines of Huckabee's comments at the NRA.
Very likely this is an NRA/white supremacist plot and some low level person is trying to bet on that info using Intrade!!
I do hope the FBI looks into it carefully, like you said!!
@Scott
Great post, very informative!
Poker-
So you think a trader is so dumb as to think Obama will leave the scene and Clinton will appear in the 40 days to save the dems? You are kidding right.
If you follow sites like RealClearPolitics, they post the Intrade numbers in RT and it is MUCH more likely these trades are a repb operative making sure McCain at least gets some good numbers.
I hope you like it whereever you live poker, because it sure is not my reality.
FYI folks re polling:
From NBC News poll:
Virginia
McCain 47, Obama 44
Dates conducted: Sept. 17-22. Error margin: 4 points.
Fred - it's not "smarter than you" - apologies if it came across that way. Gaming happens in markets, at Intrade it seems to be more predominant in some markets than others such as the Contender Markets compared to the Party markets, and both more than the State markets.
I agree market manipulation is happening in the Contender markets - they arent worth touching and certainly aren't worth serious statistical analysis. But that doesnt mean that the other Intrade markets arent worthy of analysis.
Got to say, manipulation doesnt always equal a Republican/Democrat/Ross Perot redux (or insert absurd enemy of choice) conspiracy.
It looks to be more like traders with an abacus rather than a bunch of underpaid, oversexed and overworked young guys in poorly fitting suits playing poker with party slush funds on Intrade!
Because, honestly, the political return aint worth that effort.
VA is an OK number for BO. I still have hope he will do better in VA and NC than any recent dem.
BO needs to defend NH, send Hillary up there, now!
---So you think a trader is so dumb as to think Obama will leave the scene and Clinton will appear in the 40 days to save the dems?---
Yes. Or, more accurately, there is a trader who thinks that likelihood is greater than 30 to 1 against.
---You are kidding right.---
I am not.
---If you follow sites like RealClearPolitics, they post the Intrade numbers in RT and it is MUCH more likely these trades are a repb operative making sure McCain at least gets some good numbers.---
No, in fact it's vastly less likely by several orders of magnitude. Were it more likely, it STILL WOULD NOT BE EVIDENCE.
---I hope you like it whereever you live poker, because it sure is not my reality.---
I do like it. I live where correlation is not causation, and where an explanation being exciting doesn't make it more likely than a boring one. There is no *evidence* of anything here, other than Intrade traders valuing these contracts differently than other markets. Any speculation about why this occurs is just that, speculation. Speculation isn't evidence.
Possum-
Sorry for the slam, it seems we agree. These trades are manipulation, and these trades are in the markets the repubs place on their sites and the talking heads on Fox discuss. It is too much for me to think it is not manipulation.
I like you model BTW. Good luck with it.
Here's hoping the ABC poll is not an outlier, and BO is getting a second bounce from the economy.
The dems looked for very good yesterday as they seemed to sanely approach the economic crisis as the repubs screamed the sky is falling (and tried to get not oversight for 700 billion in funds).
Poker-
Good luck with that - you are wrong, but I respect your opinion.
---Good luck with that - you are wrong---
Exceptionally unlikely, on the order of being hit by a comet.
---I respect your opinion.---
It would seem you do not. I'm not offended by this, however.
There are some high-profile PUMAs with lots of disposable income, and they may simply be trying to manipulate the media narrative here. There were people over on the Kos freaking out over the Intrade prices, so to some extent the plan must be working.
I don't think this necessarily represents a threat to Obama as much as "teh crazy" of the few genuine PUMAs, but it would seem to be against trading rules to manipulate a market like that, so maybe this person/consortium can be caught and banned from trading.
1. All contracts in the group will be paused and expired at the last closing price set for each contract
2. A new market will be listed for the remaining candidates"
Does anybody not think there is a lurking scandal?
Geoff,
I'm happy to talk polls. The national ones you cited seem all over the place. That Battleground poll has been way off center in McCain's direction since it's inception (just a few days ago, right?). I don't know why this is, but it's noteworthy. If you take the other two polls, O+1 and O+9, I suspect that the real story is somewhere in between: O+3 or O+4 seems most likely. Of course, much of that polling was done over the weekend or earlier this week, when Obama's gains seemed to be strongest. Today's trackers should give us an idea of whether Obama was cresting a day or two ago or whether the slight movement towards McCain is just noise. My guess is the latter, since there hasn't been any event to change the positive shift towards Obama that came with the huge economic news.
Of course, national polls don't mean much. The VA O+3 result for Obama by NBC is the fourth out of the last five polls to show Obama with a 2-point or larger lead. If this represents more than just a shifting of soft support, then Obama's chances of winning are dramatically increased. A win here in VA essentially guarantees and Obama win, as long as he holds MI and PA (likely, in my view).
the +9 poll is an outlier
OK, I think I was only partially understood. It could be someone who's holding Obama contracts, but it doesn't have to be and that's not what I said.
On Intrade, you can buy (go long) or sell (go short) on any contract. You don't have to buy first, you can just sell. If you do that, you get money upfront and have topay off if the contract proposition comes true.
So long as you have *any* position, long or short, you've got capital tied up. Unless they've changed the rules in the last few years, mot traders have to have on account the full amount they would lose if they lost the bet. Large traders who agree to become market-makers have to put up some fraction of what they could theoretically stand to lose.
So if you've sold Clinton contracts, you know you're not going to have to pay out, but you've got all this money that can't go to other bets. So instead of waiting for expiry, you might buy Clinton contracts on the cheap, canceling your net position and thereby freeing up your betting money.
Someone else might sell you the contract, knowing it's a sure money-maker, but that you'll have to wait for the money. For that person, the contract is like a bond instead of a stock.
That happens on every Intrade proposition that becomes a foregone conclusion significantly before it officially expires.
Someone who had previously shorted Clinton is doing balancing his position as I described and is putting their freed up money into shorting Obama. And I see no reason to assume that's for propaganda reasons. That's possible, but it's more likely just a gambler who made money betting against Clinton and now thinks he can make more money betting against Obama.
Vanessa posted:
"Does anybody not think there is a lurking scandal?"
Nah, most scandals post.
What do you mean they post?
I was making a joke about the term "lurker," which refers to someone who reads a forum but doesn't post. :(
Yesterday - Man charged with approaching Obama home with gun
http://www.suntimes.com/news/politics/obama/1180467,sengstacke092308.article
http://ap.google.com/article/ALeqM5jU-1-ZxXVvlRUAGnSJkSlXlAZffQD93CPGP00
Nate Dogg makes some strong points but there are also some flaws in his argument :
1) In terms of liquidity, Intrade is by far the MOST liquid of the markets he mentions, especially compared to the Iowa Political Futures market. If any of these markets would be too difficult for a single trader to continually control the price it would be Intrade. And if they have that kind of money to do it on Intrade, why wouldn't they do it on Iowa political futures market as well where it would be easy as heck.
2) The price of the Obama and McCain contracts have consistently moved inversely to each other suggesting that there isn't any major market manipulation going on with one or the other. Add the value of the 2 contracts together and it's usually around 98 which is about right because there's always a chance someone else could win the election.
3) Yes it's true that it would seem that the state markets are not in line with the overall election market however it's the state markets that are highly illiquid by comparison. If one were to argue let's say that McCain's value is way too low in Pennsylvania and should be at least 40 (let's just say), that would completely change the outlook of the race.
4) Thousands of people on this site alone now "know" that Obama contracts are vastly undervalued on Intrade. OK, so doesn't anyone have any money anymore? What are you guys doing just sitting there? Are you guys going to buy Obama contracts to at least get him over 55? Anyone?
And for those who think that the McCain campaign is spending a hundred thousand dollars a day or more to keep Obama's price low and McCain's price high just to make it look like McCain is doing better than he is, well that's, laughable. In the big picture Intrade has very little to do with the public's overall perception of the race and even to the extent which it does, it doesn't really have much of anything to do with translating into votes on Election Day. That kind of money would be much better spent on television and radio ads, or given to the RNC for their GOTV efforts.
"the +9 poll is an outlier"
Agreed. But it is a telling one. The further away from the actual result, the less likely it is that a poll will produce that outcome. So, it is far, far more likely that this race is O+3 than M+3, based on this poll. Sure, Obama isn't really leading by 9 points, but it does suggest real movement in his direction. An outlier like this would have been bizarre two weeks ago. Now it's simply just beyond the edge of believable. If the election were today, it's conceivable (though not likely) that the undecideds could break heavily for Obama and he could win by 7 or 8 points.
Huh. Nate was right, again.
http://www.pollster.com/blogs/pew_research_missing_cellphone.php
The Cell Phone Effect. Conclusion: Very close to what Nate predicted.
Sedi, the VA poll is McCain plus three, not Obama. And it was done from September 17-22 by NBC, a notoriously pro Obama organization these days.
Re the national polling, i agree Ras and Gallup will be important today to see trends...as well as Hotline and even Daily Cozacks.
I find it so sad that there are so many bitter, paranoid people in the world like Nate Silver and so many of the other lurkers here.
I feel sorry for them for leading such a pathetic, unhappy life.
This articule is more in a mounting pile of evidence.
Another day, another Nate Silver post stealing one of my talking points from the comments the day before...
Ras: 49-47 Obama, big move his way yesterday apparently.
Ras NH Poll - 49-47 McCain
Nate Silver = human garbage
So nobody thinks that the nature of this intrade activity suggests a looming scandal?
That's not even the most egregious problem. Over the last few days, Intrade's percentages have sometimes added up to more than 100%. In particular, there were a few times where Viriginia was split 52% Obama to 52% McCain.
Daily Kos back to 4 point lead -- 48-44.
Ras movement partially explanable in Ras uptick in Dem portion of sample which started on Sunday so is fully in today.
Thanks geoff,
strange how the numbers differ. Given Rasmussen´s tilt to the right, New Hampshire is looking in fact almost tied at the moment, which is worse than I expected. On the other hand, I expected the national tracker to be tied... and the Research 2000 tracker lost a point. 48-44 Obama today.
Geoff, that Ras "Dem uptick" was miniscule.
Rasmussen shows Obama up 49-47, two point jump, which is big for that poll.
However, Ramussen shows McCain up 49-47 in New Hampshire; Obama better spend more time in that state, we know how popular McCain is there.
VA will come down to turnout, nothing more, nothing less. If the economy stays sour, NC and FL will also come down to turnout.
National polling is kind of mixed, but its clear Obama has at least a 2 point lead nationally, and likely more.
Guys there's actually a very easy explanation:
Somebody believed that (1)McCain will win and (2)that Obama would clinch the democratic nomination.
So he was in a huge short position with Clinton stocks. For whatever reason he waited until now to cover his position hence the raise in Clinton position.
As he's realizing his profits from his short position he's going short on Obama.
As the market doen't have a lot of volume he can't cover his short position as quick hence the slow short going on the Obama position.
You see perfectly logical and absolutely no need for secret service or any other conspiracy theorie...
I think Nate just found an anomoly, expounded, and put it out there. I just wanted to put that out there before people started accusing the guy of meeting a guy who looked like a smoking Hal Holbrook in a Washington, D.C..
I think we have the difference here between worldwide exuberance for Obama on BetFair and the hesitance an American futures folks like InTrade might take. BetFair's observers are betting more mightily with the heart than the head. InTrade's people know from 2000 and 2004 that certain political parties invest time and money into interfering with the electoral process. They are closer to the trenches and know Rove is on the McCain payroll and behind the scenes.
Don't think it can happen? It has happened. The world assumes the 'greatest democracy in the world' counts all of its' votes, let alone has manditory participation in voting. We do neither. We challenge votes, throw them out, and do everything legislatively to suppress the vote. This goes on with both parties. It is just the GOP has become masters at it over the past few decades.
This is heart versus head. Pat Buchanan was asked why Nixon never contested the results of what Republicans were calling a stolen Illinois for Kennedy in 1960. His response: "We couldn't. Nixon stole Kentucky and Louisiana".
The wonder of this website is that it deals in the rational. Barebones numbers do not lie. You have to look at real factors that back up trending numbers. BetFair is in the countryside observing while we all sit in the trenches watching the war.
Matt, the tick was about .5 points to the D's. It's not enough, but it is meaningful. One decent polling segment last night could push it, but it doesn't necessarily correlate to a monster night.
If it's above O+4 tomorrow, then yes, it will have been a monster night, but from here it could have easily been that last night was maybe O 50 M 45-46. Not good for McCain, but not a race-changer.
In short: Need more data.
Matt, the tick was about .5 points to the D's. It's not enough, but it is meaningful. One decent polling segment last night could push it, but it doesn't necessarily correlate to a monster night.
If it's above O+4 tomorrow, then yes, it will have been a monster night, but from here it could have easily been that last night was maybe O 50 M 45-46. Not good for McCain, but not a race-changer.
In short: Need more data.
Thanks, magnum. I gave what I think was a very good explanation of how that works about 10 minutes ago, but like about half of my comments here, it never showed up. Very frustrating.
Thanks for the advance post of Ras numbers Geoff. You're the kind of Republican that I'm grateful for!
Obama must have had a huge day yesterday to go from a tie to a 2 point lead. A 6 point lead would be required which, given Rasmussen's party weighting, is a really good result for Obama.
If I have calculated Gallup's single day scores correctly, Obama is due a big jump tonight as well. I have the last 3 days results as M+2 O+4 O+7 to give O+3 for the 3 days. Assuming that Obama is say +4 tonight he will jump to a 5 point lead in the Gallup tracker. We'll see.
The NH poll is also very interesting and appears to confirm that it is one of the few Kerry states (together with PA) that appears to be tightening as the race continues. The prospects of 269-269 tie get dramatically bigger if McCain can swipe NH. I still suspect that Obama will win it but this race is far from straightforward.
By the way, do you get any internals for state polls. For instance, what party weighting does Ras use for NH? If you have weightings for other battleground states I would love to see it.
I think that what we have here is an apples and pears problem. The two bets are not the same. The Betfair market is assessing the chance of Obama being President at 60% or so, whereas Intrade is doing something quite different, I believe.
I think that Intrade (I've never bet on it, so I can't be sure) is a spread betting platform. The bet that you're making at 52% relates to the percentage of the Electoral College that Obama will get. If he gets higher, you win when you buy at 52%, If he gets lower than 52% then you lose. The amount that you win or lose depends on the amount that you stake per point.
So it's theoretically entirely possible for Obama to have a 60% chance of winning and for the probable Electoral College percentage to be 52% in favour of Obama.
Mystery solved: markets work!
Re: Hemonth
That's not a flaw, it was a temporary situation. It's possible for the percentages to add up to more than 100%, temporarily, and soon somebody will sell and take the profit. Assuming there is a buyer, and I doubt there are a lot of buyers on Intrade (which is really, surprisingly, a pretty small market).
Intrade does indeed look suspicious, but I think Obama being assassinated and replaced by Clinton is a pretty unlikely scenario. A scandal is possible, but I doubt Obama has that much baggage, and he has been a pretty clever politician so far. So, as has been suggested by several users above, it's more likely someone who sold short on Clinton in the primaries and is now buying it back at the low price to trade against Obama.
Geoff,
Thanks for the correction on the VA poll. With McCain up 3, I think it really brings back into focus the state's status as a toss-up, given the other recent polls. Are you sure about the NBC house effect, or are you just assuming that it has one because it's NBC? Recall that PPP doesn't have a house effect, even though it is a Dem pollster.
Regardless of the nationals, then, I'd say that McCain has the best results of the day so far, with leads in VA and NH. Does anyone have an explanation for why NH seems to be moving in McCain direction lately?
I suspect that most of the movement in the trackers is noise. We saw over the summer how much movement there can be in the tracking polls without any real movement of voters.
Dvd Avins has it spot on - the Intrade Contender markets are legacy markets filled to the brim with loss amelioration, which leads to fleecing and gaming. They aren't representative of the true market perception of the political 'state of play' because of that. The Party markets and the State markets are the ones to watch - which for some reason, most people don't.
Most Intrade coverage starts and ends with the markets with the poorest information - go figure!
@Hemonth - that often happens with Intrade contracts, but quickly closes up over a short period of time. The pursuit of market equilibrium writ large ;-)
Not to mention a few free bucks if you can be bothered to tie some money up if the discrepancy gets too big.
The fact of the matter is that somebody out there is betting aggressively on a disqualifying event. It sure as hell isn't some violent event because Intrade has a policy that all futures expire in the event of a candidates death.
I mean this suggests to me that there is a looming scandal. I can think of no other reason.
No internals on NH yet.
Another new unreleased poll (will be tomorrow)
VP favorability:
9/9 1000 LV
Biden 53/40 (25% v. fav, 21% v unf)
Palin 56/40 (39% v. fav, 26% v unf)
9/23 1000 LV
Biden 49/41 (22% v. fav, 21% v unf)
Palin 54/42 (36% v. fav, 31% v unf)
So both have lost overall, but Palin is holding up in this poll better than Biden.
Keep in mind this is 9/23 sample, as noted, a good night for Obama on Ras.
Any comment the inkstain?
Intrade is not a legitimate market. If it was, then people would be pursuing arbitrage positions within 10 minutes of some idiot shorting Obama every morning. That's exactly what would happen on the NYSE or any real market. Because this doesn't happen I am afraid to touch the market, which might end up being illiquid (ie. major event happens and I can't sell, even at a loss).
Clinton short sellers are covering their position
A commenter posted awhile ago that the Clinton buying is clearly covering a short position. A short needs to be explained. A traditional stock purchase is a "long" where you make money if the stock goes up. A "short" position does well if the stock goes down. When you short a stock, you sell stock you don't own at the current price and are obligated to buy it back later at the current market price. Hence, if someone shorted hilary at $35/sh they have to buy it back to make up that cash. If the cover is at $5, you make $30 per share.
People still own those worthless Hilary 4 Prez contracts, but they are holding out selling it until they get the highest price. Some may sell at $1, some at $2 or some at $5.
Did anybody watch what happened to Oil on Monday? It jumped from 108 to 121, because traders had to cover their shorts as the October futuers contracts expired. Tuesday the oil price opens at 108. Knowing that short sellers had to buy oil on Monday, people who owned the commodity pushed up the price to those covering short contracts, since they knew that they had to pay whatever the market asked. The market is not the intrinsic value of the product, but a price that someone is willing to pay for the asset.
I just think people on Intrade are betting with their hearts and conventional wisdom.
It's almost impossible to convince people that any election can be anything but "anybody's game!" and "a tossup!" with more than a month to go, because the two most recent ones were very close and that's the media narrative.
@Mule Rider: if you have such a low opinion of Nate and his readers, then why are you here?
I firmly think this is an election where how you are perceived in a state, coupled with your ground game in a state, will matter a lot more than previous years.
Obama appears unlikely to win Ohio regardless of what the national trends are. Similarly, Pennsylvania will remain very tight (+2 or +3 Obama) regardless of how the national race turns out. McCain will outperform in New Hampshire, guy has been campaigning there for 8 years.
On the plus side for Obama, he will remain competitive in Virginia and Colorado even if he loses a couple of national points. Iowa is a lock for him regardless of the national vote or the state's demographics. His ground game in Florida, Indiana, and North Carolina have the potential to flip those states independent of any national trends.
As I have been saying for weeks, I think Obama should concede Ohio and move his considerable resources from that states to PA, MI, and I guess NH. It is more important for Obama to protect PA than to win OH at this point.
Looks to me like a seller simply easing out of a large position. Maybe some other Intrade contract looks more lucrative to the seller right now. Nothing to sweat about IMHO.
Can someone explain to me how the Intrade state of Oregon has 87% for Obama and 17% for McCain. Isn't that 104% total??? I can understand minor rounding issues but 4% over????
There's also a 3-point discrepancy between the Pres.Obama and Pres.Dem contracts. This suggests that it's primarily the Pres.Obama contract that is being manipulated and the other one is staying somewhat low from everyone taking the free money due to the discrepancy.
---McCain will outperform in New Hampshire---
Hahahaha, no.
Are you sure about the NBC house effect, or are you just assuming that it has one because it's NBC?
The NBC polls are being conducted by Mason-Dixon, highly rated on Nate's list, and not partisan.
Will New Hampshire like Hillary as much without the tears?
McCain can have New Hampshire. Punt it.
Only a small fraction of Obama's paths to victory (with NH) have him stopping at 269-272 EVs. There are a few, but none are particularly likely.
"NBC, a notoriously pro Obama organization these days. "
I have to disagree, NBC is not pro-Obama at all, MSNBC is pro-Democrat on their evening shows, but even their morning show is a an ex-republican Congressman.
NBC is not biased, and MSNBC is significantly less biased than Fox.
---The fact of the matter is that somebody out there is betting aggressively on a disqualifying event.---
No.
this has got to be scandal.
Richard : The number displayed is the most recent price that the contract traded at. If the last trade on one contract happened two days ago and the last trade on the other happened two weeks ago, then the "snapshot" obtained by summing them doesn't accurately correspond to a single moment in time. If you actually check out the asks and bids on the book, you'll probably get more reasonable numbers.
I'll admit it. Nate Silver has officially blown my mind. Especially the end part about the Pentagon pushing for a terrorism futures market that I had somehow managed to miss until now.
I agree with the others saying that it's more likely someone is betting on a scandal or some other event that causes Democrats to scramble, but regardless, I don't even know how Nate thought to look for this. I'd noticed the weird trending in Hillary's stock, but...really, who would think to connect that to Obama short trading. Nate's mind works in interesting and seemingly very effective ways.
Inkstain-
The Kerry states plus IA, CO, and NM, minus NH is a 269-269 ties and seems pretty likely right now. Of course, all the close big states in the East would need to go McCain (FL, VA, NC, etc.) but that is not highly unlikely.
d,
in the event that a candidate dies the contracts expire and reset. There is no need to be concerned for his safety. The concern is that there is scandal, somebody knows about it, and is making a few bucks off of it.
"The Kerry states plus IA, CO, and NM, minus NH is a 269-269 ties and seems pretty likely right now. Of course, all the close big states in the East would need to go McCain (FL, VA, NC, etc.) but that is not highly unlikely."
And Obama almost certainly wins a tie. That was my point. Almost all his likely win scenarios put him at 273 or better, few in 269-272, where losing NH would actually lose him the election.
d-
Nate is a U of Chicago grad, they eat each other alive there (I know) and it forces you to think through all contingencies before opening your trap.
Go U of C! Go Nate!
inkstain-
I agree, but lets get Hillary in NH ASAP and just take away the one tie scenario that worries this morning.
# Daily Presidential Tracking Poll
The Rasmussen Reports daily Presidential Tracking Poll for Wednesday shows Barack Obama attracting 49% of the vote while John McCain earns 47%. It’s the first time in more than two weeks that Obama has enjoyed a lead larger than a single percentage point.
"There's also a 3-point discrepancy between the Pres.Obama and Pres.Dem contracts. This suggests that it's primarily the Pres.Obama contract that is being manipulated and the other one is staying somewhat low from everyone taking the free money due to the discrepancy."
Do you ever care to think that a Democrat besides Obama could win the election? Crazy stuff does happen sometimes you know. That's the reason for the discrepancy. Not your BS.
inkstain,
Does the trading activity not concern you at all?
Thanks for the VP figures Geoff. For what it's worth I think we are now past the point where the VPs are going to make any real difference (apart from their debate). The Palin bounce has gone with the residual effect being a firming up of the GOP base but with no overall change to the national lead.
The favorability figures do seem to be kinder to Palin than other pollsters but this seems to be a pattern. For instance Rasmussen consistently has Obama tied with McCain on the economy whereas every other pollster gives Obama a lead of 5 to 15 points. Not saying that Ras is biased but that there is certainly some house effect which may or may not be proven right.
Ras has results for Colorado later today also. Will be interesting to see whether McCain's lead in the last CO poll has gone.
Richard, re Oregon prices summing to more than 100: this is common on Intrade due to the relatively thin volume. The thinner the volume the wider the buy/sell spread, usually.
---all the close big states in the East would need to go McCain (FL, VA, NC, etc.) but that is not highly unlikely.---
It is, in fact, highly unlikely.
@inkstain: Without NH, the well-known victory path of Kerry + IA + NM + CO only gets 269 electoral votes. It's probably a win, but will be problematic in terms of real governing, and is probably something Obama wants to avoid. (I think Nate spoke to this yesterday).
So he'll need to find another EV somewhere. And NH is probably the easiest place to get it, easier than going to somewhere like VA, NV or OH.
---The concern is that there is scandal, somebody knows about it, and is making a few bucks off of it.---
They'd have to be an abject drooling simpleton to have decided this was the best way to profit from knowledge of a scandal large enough that Obama would drop out of the campaign.
"Does the trading activity not concern you at all?"
Not really. As Nate said, $400k volume isn't much at all. One person with a little cash really likes McCain.
I've been trying to convince my wife now to go along with a small purchase of Obama contracts from our savings. Nothing life-changing, but you almost *never* see a deal this good in gambling.
Besides, does anyone want to deal with the crisis that a tie vote would cause. Pundits would love it but in terms of actual governing, the next 4 years would suck.
"@inkstain: Without NH, the well-known victory path of Kerry + IA + NM + CO only gets 269 electoral votes. It's probably a win, but will be problematic in terms of real governing, and is probably something Obama wants to avoid. (I think Nate spoke to this yesterday)."
I'd take the risk of getting stuck in that situation, where he probably wins, over the lost utlility of putting resources into that state. Too many other states seem like fertile grounds.
RCP average keep listing the outdated battleground tracking poll, the only one showing a lead of 2 pointsfor McCain, this poll tresches from the 17 to 23 of Sept. and a sample of 800. Smells fraudulent to me.....
"Besides, does anyone want to deal with the crisis that a tie vote would cause. Pundits would love it but in terms of actual governing, the next 4 years would suck."
I doubt it. Once February hits, nobody but the partisans on the losing side care how the President got elected. And they would find a reason to hate him anyway.
Maybe Lynn Forester de Rothschild recently discovered Intrade?
Post a Comment