Bryan Caplan  

Obamacare Market

My Book List... The Case for Competitive Gover...
Can anyone explain what happened in the Obamacare betting market over the past 24 hours?  Why did it fall from near 70, close at 35.1, then rebound to 65.1?

Comments and Sharing

COMMENTS (19 to date)
razib writes:

i'm curious too. but this isn't the only time it's been volatile, during the health care summit it jumped up to 90 for an hour or so before drifting back down.

Ted writes:

Probably just like the stock market - sometimes things have nothing to do with fundamentals.

NathanM writes:

The fees InTrade charges (a minimum of 10% and maximum of 15% per trade, plus fees for withdrawing money) are so high that someone interested in trading only to maximize profits will avoid InTrade unless the InTrade price is extremely out of whack with the fundamentals of the contract.

So, as one would expect, InTrade has extremely low trade volumes (for example, 608 trades on that contract today, which is considerably higher than the mean). So it is cheap for someone who isn't interested in making money to manipulate InTrade, and I am left wondering why anyone cares about InTrade. Price fluctuations in a market with such high transaction fees are unlikely to mean very much.

razib writes:

nathan, do you think intrade offers any value in relation to pundits? for example, megan mcardle, who isn't a partisan hack, recently implied that there's almost zero chance that the democrats would keep the house this fall. that seems too low of a probability and way too certain. intrade has been around 50/50, which seems to mean "no idea."

Eri Culver writes:

The reason that the Obamacare betting market rebounded to 65.1 is because of Rep. Dennis Kucinich, D-Ohio.

Although he was against the health care bill, he had four meetings with President Obama to discuss it. He then switched his position and announced his intention to support it. The key defection from the opposition greatly improves the chances for the bill's passage.

Peter Twieg writes:

Yep, from what I gather a single trader managed to manipulate the market so it closed yesterday at 35.

NathanM writes:

@razib I wouldn't say InTrade is completely valueless, as McArdle's extreme prediction seems to be. I just keep in mind that so long as the InTrade price is within plus or minus 10-15 cents of the "correct" value, then even someone who knew the true odds and had a way to risklessly arbitrage wouldn't be interested in InTrade.

So, basically, the long term InTrade price is set within a very broad range by people who (a) don't understand how the market they are trading in works or (b) don't care if they lose money.

As for the short term price, the volume is so thin, and the reason people would game the market is so obvious, I don't think drastic shifts require any explanation.

Matthew Gunn writes:

This is actually an instance of a classic question... do asset prices have downward sloping demand? (Most theory says they should be flat.)

Empirically though, the general answer appears to be that if there are limits to arbitrage, crazy things can happen.

In this case, people that normally act as market makers probably didn't have enough money or time to keep the price up. One guy probably came in at the end of the day and bought up the whole ask order book.

Completely crazy things can happen in ACTUAL financial markets too when arbitrage arguments break down!
is a classic and troubling paper on the topic. Worth a read to anyone who hasn't seen it already.

geckonomist writes:

my first thought was that Goldman Sachs must have cornered the market.

Blackadder writes:

The fees InTrade charges (a minimum of 10% and maximum of 15% per trade, plus fees for withdrawing money) are so high that someone interested in trading only to maximize profits will avoid InTrade unless the InTrade price is extremely out of whack with the fundamentals of the contract.

If this is true, why did the Intrade price almost immediately spring back to just below where it was before?

Joe Teicher writes:

I'm with the people who say that this doesn't require an explanation. A total of 49 contracts traded below 60 yesterday, and only 18 contracts below 50. If someone was trying to manipulate the price, assuming they can get back out at around 67 they spent about $100 to accomplish this feat. Its just a meaningless blip.

Joe Cushing writes:

Maybe it has something to do with the Obama deadline to get the bill passed.

caveat bettor writes:

Intrade does suffer from a lack of liquidity (thanks, Congress, for the Internet Gambling Regulation and Enforcement Act).

Single, atomic decisions made by committee without the benefit of survey data (e.g. Nobel prizes) such as healthcare legislation that may test the limits of Coase (since the 2 houses of Congress might not vote constitutionally for it) are, I concede, difficult for prediction markets to provide stable price signals.

But Intrade is still quite predictive, as much as Congress has constrained the space. For instance:

jb writes:

Tinfoil Hat Theory (offered in jest)

1. If the InTrade numbers are high, that reassures potential technically sophisticated on-the-fence Democrat legislators that the bill is likely to pass, and they will have political cover.
2. So it's in Obama's best interest to make the InTrade numbers as high as possible
3. And if anyone tries to capitalize on this, Obama's Federal InTrade Manipulator Agency (FIMA) is responsible for re-bidding up the price

Ari Indik writes:

Regime uncertainty.

Megan McArdle writes:

If I implied that, I inadvertently sounded way too certain. I think the chances are worse than inTrade--somewhere in the 80-90% range. But not 100%.

Paul Hewitt writes:

Nathanm: The InTrade fees are not 10% to 15%. They are a maximum of 5 cents trading in and 10 cents on expiry. Given the $10 expiry payout, this is 1.5% (though it would be higher when calculated on cost).

So, the market prediction *could* still be fairly accurate. I'm not saying it is, but it won't be inaccurate, because of transaction costs.

At any rate, I'm not a fan of these types of prediction markets (discrete options vs. continuous variables, such as quarterly sales). In these discrete markets, if you're almost right, you're completely wrong.

The manipulation in this market was done in the middle of the night, just prior to newsrooms starting their morning news stories. Coincidence?

mulp writes:

It is predicting the outcome just as well as the market predicted the ultimate value of all those CDOs.

Maxim writes:

The claims above about Intrade fees are false.

The fee is actually 1% to 1.5% per trade for winners of a bet; and 0% to .5% for losers (different fees for price takers and makers.)

The above poster likely assumed that each Intrade contract is for $1... in fact, they are for $10. That is why they got the fees off by an order of magnitude.

I would also note that the Intrade irregularity was corrected almost immediately after the irregularity.

Comments for this entry have been closed
Return to top