Inefficient Markets
I just took a look at the bid-side markets on Intrade for the Democratic nominee for Vice President. Here are the levels (for those names with bids above zero):
Tim Kaine – 30.1
Evan Bayh – 26
Kathleen Sebelius – 16
Joe Biden – 10
Brian Schweitzer – 5.5
Hillary Clinton – 5
Sam Nunn – 4
Jack Reed – 4
Anthony Zinni – 4
Al Gore – 3.5
Ed Rendell – 3.5
Wes Clark – 3.2
Chuck Hagel – 3
Janet Napolitano – 3
Bill Richardson – 3
John Edwards – 2.8
Chris Dodd – 2.5
Dick Gephardt – 2.2
Claire McCaskill – 2.2
Michael Bloomberg – 2
Bob Graham – 2
Mark Warner – 1.7
Tom Daschle – 1.2
Bob Casey – 1
Barack Obama – 1
Ted Strickland – 1
Tom Vilsack – 1
Bob Kerrey – 1
John Kerry – 0.6
Caroline Kennedy – 0.5
Jim Webb – 0.5
That adds up to 147. Say you take that bet for $1 per unit – you’re out $100 if you short a name and that person is picked. I don’t know the margin rules, but let’s assume that if you short the whole list you have to put up the maximum you could have to pay out – which is $100, since only one name can possibly pay out. That means you can short the entire list and immediately take home $47 risk-free – with the potential to make an additional $100 if a really dark horse like, say, Bill Bradley or James Jones turns out to be the nominee.
Or suppose you think Evan Bayh’s going to be the nominee. You short everybody but him, and take home $21, no money down (the $100 margin comes out of receipts on the other trades). If either Bayh or a dark horse gets picked, you get an additional $100, for a total profit of $121. By contrast, if you bought Bayh, you’d put down $29.9 for a chance to receive $100 if you’re right.
Or suppose you think one of the top four contenders is going to get the nod – Kaine, Bayh, Sebelius or Biden. To buy all four, you’d need to pay a total of $93.80. But if you just shorted everybody else, you’d receive $64.90. Assuming you have to put up $100 in margin, you’re putting down a net $35.10 with the potential to receive $100 – or, better, $35.10 with the potentially to earn $64.90, nearly a 2-for-1 payoff – versus putting down $93.80 to potentially earn $6.20. Why would anyone play the long side when you can put the same bets on for such better odds on the short side?
There must either be some very screwy margin rules, or there must be absolutely no liquidity. Either way, the existence of a substantial apparent arbitrage like this suggests that these markets should not be taken too seriously as a guide to what “smart money” thinks about the odds of anything.
Intrade has two markets for the Democratic VP nomination. The one you listed will add any new candidate upon request so you can bet on him. The second one has a short list of candidates and a contact for the field. That second market has 100x the volume of the one used for this post, which makes it much more useful as a guide.
— David · Aug 1, 05:05 PM · #
The fees on the tails sells are a killer on intrade, and selling them is the strategy above. Contracts are in $10 numeraire units. Putting the fees in above, if nobody above wins you make $10.96 (instead of $14.70). If someone does, you make 96 cents.
If I understand the margin rules, you have to put up the maximum you can lose across all bets, which would be putting up $295.3 to make 96 cents. Suddenly the expected value doesn’t look so good.
— Mike · Aug 1, 05:26 PM · #
Mike’s right, which is why Intrade isn’t very accurate for low percentage trades. The amount of capital you have to immobilize to meet your margin makes the deals worthless even if you stand to make some money, theoretically.
— Adam Greenwood · Aug 1, 05:51 PM · #
3 cent commission on tails right away. The killer is that you pay 10 cents (equal to 1% of $10) if your trade wins, so if you sell “Jim Webb – 0.5%” (5 cents) and your trade wins, you pay a 10cent fee on that 5 cents. (I’m not sure if they just give you zero in that case or not).
In theory, since the inefficiencies bias the selling of the tails (there are, as the kids say, “limits to arbitrage” there), buying mass tails should have a statistical arbitrage. Of course you would need some serious capital to survive it. I wonder if they have historical data…
— Mike · Aug 1, 08:05 PM · #
“buying mass tails should have a statistical arbitrage.” Wait, no it wouldn’t. (exact opposite.) That’s why I don’t work for a hedge fund.
— Mike · Aug 1, 08:12 PM · #