What PR value? The only people who have even heard of Intrade are extreme politics wonks who already know who they’re voting for. Not the average joe whose vote the campaigns are looking for
Yeah, this is the sticking point for me. I suppose, maybe people betting on InTrade think it has some sort of influential value and may want to run up the prices, but I simply don’t see that as the most likely scenario. Does anyone really care what the fuck InTrade thinks? What %age of Americans even know about betting markets?
Among those politics wonks are pundits, bloggers, etc. who have influence over other people.
Check out post #33. I said the work involved a Fox News reporter in my example. Now why do you think I used a Fox News reporter being called?
Before he became President, Lyndon Johnson played a rather rough game of politics: believe it or not, Democrats tended to play dirtier than Republicans during the 1900 - 1950s.
Anyway, Johnson was meeting with his advisers when he was running for the Senate: they were brainstorming about how to defeat his opponent.
Lyndon B Johnson: “Start a rumor. Say he fucks pigs.”
Advisers: " **. . . **"
LBJ: “He’ll never be able to deny it.”
Now only a tiny percent of the voting public would hear that slander. But a high percentage of opinion makers would get the message, and that would color their reporting. In today’s context, there are plenty of horse race reporters who aren’t especially numerate: they are in the narrative business. Distorting the intrade market, either temporarily or merely shifting the level zone, is cheap marketing of the Romney product.
Since last evening, intrade prices have reverted to their long-run 538 + ~10 percentage points level. The small scale betfair arbitrageurs, assuming they exist, appear to maintain a certain spread. Good for business, provided they don’t run into additional competition.
I’m having trouble understanding just how Intrade works. Are people betting with other betters on Intrade, so that for every long contract there is a short contract as well (In other words, there are an equal number of Romney and Obama contracts out there)? Can you buy shares from Intrade, and sell them on the secondary market, or is it strictly a secondary market that Intrade facilitates? Is it exactly analogous to a stock market exchange?
Let’s say I buy 100 Obama contracts at (for example) $60 a pop. I pay $6,000 in order to have $10,000 if Obama is elected, and $0 if not. Where does this $4,000 come from? Does Intrade hold onto all money until the election is decided? If so, they would likely want to pay out as close to an equal amount of money under each scenario, as they seem to charge both monthly and transaction fees. This would be analogous to the vig on sports betting.
If this is the case, then the implied probabilities would not necessarily reflect reality, but instead are the numbers that ensure an equal amount of money is paid out by Intrade under each scenario. For sports betting, if Team A is favored by 7.5 points, it doesn’t mean the bookies think Team A will beat Team B by an average of 7.5 points if the game were play ad infinitum number of times. Rather, the amount of money on each side of the spread is close to equal, so that the bookies are not at risk depending on the results of the game, and instead profit by charging a 10% vig. They would make about the same amount of money no matter if Team A covers the spread or not.
Am I way off base, or does Intrade operate under the same principle? It would help explain why they do not agree with other betting markets or projections.
Maybe we’re missing something here. Maybe there’s manipulation but it’s not to benefit Romney.
Is there some other way a manipulator might benefit? Are there perhaps betting pools that are based on Intrade in some way? Something else?
Could somebody be running some sort of test to see what is and isn’t affected by Intrade?
Your initial suggestion is correct - people are betting against other parties and Intrade is ambivalent as to the outcome. Intrade is like the stock market which facilitates trading in stock but doesn’t make or lose money based the outcome.
Again, the question here is if the odds for Obama/Romney are 60/40 on Intrade and 70/30 on another exchange, then someone could bet 3,000 on Romney on the other exchange and 6,000 on Obama on Intrade. This way they’ve paid out 9,000 and are guarenteed a 10,000 return. (Aka arbitrage.) This would cause bets on Romney to migrate to the other exchange and bets on Obama to migrate to Intrade until the odds are roughly comparable.
Then why doesn’t this happen? Why is Intrade out of line with other betting markets? Are transaction costs that high? Or has it self-corrected already?
Also, this doesn’t preclude the possibility that the typical bettor is biased towards one candidate despite evidence to the contrary, so that every market will reflect the same biased probabilities on average. Not even necessarily for Romney; Obama’s odds could be artificially inflated. That happens all the time in sports betting, as I understand it.
Completely unsourced speculation: Romney’s supporters tend to be richer, thus have more money to throw around, thus are more likely to bet money on the election outcome, thus Romney’s odds are biased upwards on all political markets. Is there any way to prove or disprove this assertion, other than somehow running this election thousands of time to derive an empirical probability?
Edit: I want to emphasize my first set of questions. Seriously, if the odds are 60% Obama on one market and 55% on another, what’s stopping me from spending $4,000 on the first for Romney, $5,500 on the second for Obama, and coming out ahead $500 no matter what happens? What prevents me from doing that right this second?
In a sense no one is actually selling anything, what you need is someone willing to make an analogous opposite bet. So in your 60/40 example, i buy 100 shares of Obama at $60 ($6,000) only if someone is willing to buy (or sell in their parlance) 100 shares of Romney at $40 ($4,000). When the election is held if Obama wins, then I get the $4,000 that the Romney person bet. Intrade requires you to have enough cash in an account with them to cover any bet so it is just a transfer from one account to the other. They charge $4.99 a month to trade with them. Not sure if there are other charges.
That’s the key question that a lot of people have been discussing. On another thread, a British guy claimed to have invested over $200K doing precisely this.
Potential obstacles are transaction costs, currency exchange risk, and the fact that US investors can’t access the foreign exchanges.
Perhaps the markets aren’t large enough to make this worthwhile? In my scenario, there is an spread of 5%, but even spending $10,000 will only net you $500. To make serious money, you’d have to bet a few orders of magnitude more, which would seriously distort the markets and probably not be possible. I don’t think anybody could really bet $1M on Intrade for one candidate without driving the prices up ridiculously high. That $18K bet mentioned earlier itself swung the probabilities by 8%. I guess the arbitrage opportunity would disappear quite quickly.
So my personal explanation is that there isn’t enough money in these markets to either avoid arbitrage, give an accurate barometer of the election, or even remain consistent with each other. Is there any way to know what the ‘market caps’ of the different markets and candidates are? I see the number $15M below for the total political market on Intrade. If true, this is nowhere near large enough to give much useful information.
I think one of the best bets is this one which is a 52.5% chance that Obama wins, Dems keep the Senate and Republicans keep the House.
Was only a matter of time.
IMHO it doesn’t make sense to compare 538 (or the other models) to Intrade. Intrade is not a prediction model. Intrade is an assessment of the “conventional wisdom” at any given time. To the extent that you value conventional wisdom, you value Intrade. But it’s not comparable to a properly done statistical analysis.
I like intrade because it’s a way of boiling down all the on-the-one-hand-this-and-OTOH-that commentary into a number that reflects the aggregate consensus of what people with money on the line really think. But it’s not a true prediction.
Obama’s back up above 60 on Intrade, at least for the moment. Who knows what’s driving all the movement.
FYI, Kerry was somewhere around 80 - 90% on Intrade the day of the election. Not sure why you’re putting stock (hah!) in it.
Nobody should be putting stock in Intrade. As a measuring instrument for these kinds of things, it’s about as accurate as the “Choose your candidate!” 7-Eleven cup poll.
Cite?
Countercitation:
Intrade predicted Bush in early Sep 2004: http://www.msnbc.msn.com/id/5943565/ns/politics/t/bullish-bush-accumulating-kerry
Cite3: Internet Market Predictions - The New York Times