This Blog Has Moved!

My blog has moved. Check out my new blog at realfreemarket.org.



Your Ad Here

Friday, February 22, 2008

The Defect of Intrade

Intrade's trading system is widely described as a great invention. However, there are fundamental flaws in its system that cause its markets to be inaccurate.

Intrade has a defect in its trading system that causes longshots to be overpriced. For example, "Ron Paul wins the Republican nomination" obviously has a value of zero, but its actual price is far greater than zero.

The problem is that, when making a long-term prediction, you are required to post margin equal to your worst-case loss, and you don't get credited with interest. For a longshot, short selling it would involve tying up a lot of capital without earning interest. This means that people can't profitably short-sell longshots.



I liked this post on Marginal Revolution about Intrade.

Mr. Ravitch has made a nice profit betting against Ron Paul, the libertarian who late last year was, amazingly, given almost a 10 percent chance of becoming the Republican nominee. “If you asked anyone in politics whether there was ever, at any point, a 10 percent chance of Ron Paul being the nominee,” Mr. Ravitch said, without finishing the sentence. “That sort of makes my case for me.”

There's a problem with Intrade's margin rules. Suppose you sold $1000 that Ron Paul would not win the nomination at 10% odds. In other words, you collected $1000 when you sold the "Ron Paul not nominated" contract. If Ron Paul would have won the nomination, you would owe another $9000. If Ron Paul is not nominated, you can keep the $1000 as profit.

The problem is that you are required to post the $9000 as margin collateral. To sell $1000 of "Ron Paul not nominated" futures, you need $9000 in your account. In other words, your profit (if you're right) is 11%. The "Ron Paul not nominated" contract does not settle until the nomination is officially announced. At the time, it was still a year until the convention. This means your profit rate would be 11%, which isn't even comparable to a stock market investment.

The real problem is that you don't earn interest on the $9000 collateral you were required to post as margin. If Intrade credited everyone with interest, then their market would be far more accurate.

If you wanted to create a fair online betting market, you should credit all accounts with interest. The market would automatically correct prices for long-term contracts. Currently, Intrade pockets the interest on all accounts.

Notice that short-term contracts aren't as distorted by the lack of interest credit. If you're betting on the Super Bowl outcome in late January, the effect of interest is negligible. If you're betting now on "Who will be elected President", then the lack of interest credits distorts the market.



This post on Techdirt is completely wrong, regarding inefficiencies in online prediction markets. The article cites "Ron Paul is still priced at 1.2% to win the Republican nomination." At this point, it's obvious that Ron Paul's true odds of winning the Republican nomination are essentially zero.

Can you profitably sell "Ron Paul does not win the nomination"? No. There's a defect with Intrade's margin rules. When you take into account Intrade's margin rules, this trade is not profitable.

Suppose you sold $120 of "Ron Paul does not win Republican nomination". The worst-case loss is that Ron Paul actually does win the nomination, in which case you owe $9880. You are required to post $9880 in your account to meet the margin requirement, which is the worst case loss. This capital cannot be used for other trades until the "Ron Paul not nominated" contract settles. The contract does not settle until the nominee is officially announced at the Republican convention late this summer, approximately 6 months from now. Your return is $120/$9880 = 1.2% over 6 months. You'd be better off investing your money elsewhere.

The problem is that you don't earn interest on the money in your account. You should earn interest on the $9880 plus the $120 you received when you sold the contract. If Intrade credited accounts with interest at the Fed Funds Rate minus a small fee, then the Ron Paul contract would correctly trade at nearly zero.



If Intrade credited all accounts with interest, then Intrade's betting markets would be far more accurate. Under the current system, the longshots for a long-term contract tend to be overpriced. They can't be profitably short sold, due to the margin requirements and lack of interest credits.

4 comments:

JC said...

You also missed the part where shorting Ron Paul costs 0.3% in trading fees and they take out another 1.0% on expiry. So if you short Ron Paul at 1.2 for $9800 to win $120, Intrade will charge you $30 in trading fees and $100 for your "winning" trade on expiry.

Anonymous said...

you also missed that the fee is 3% not .3 because the contracts are .1 dollars not 1 dollar contracts and its .03 per... also the expiry fee.

so even if you win you lose a lot in fees... this site is a total scam

Anonymous said...

And they fix the results so insidertraders can take your money. Daily DOW was higher at 12:00PM in NY? They just add or subtract however many seconds they need so their insidertraders win as much as possible. An insidertrader is long 100 for DOW higher at 12:00:00? But the DOW was lower in NY? Then just wait a minute or two. Or take the DOW at 11:59:00. "But it was 12:00:00 when you thought it was only 11:59:00 there". Sure it was. Whatever makes insidertraders the most money.

And only their insidertraders can be market makers on daily dow.

Anonymous said...

Two comments:

Since this post in 2008, Intrade has revised its fee schedule. Now there is simply a monthly fee of $4.99 to maintain an active account. There are no other trading costs. This alone has made the market much more liquid.

Next, in linked markets, there can only be one winner. For example Ron Paul *and* Mitt Romney can't EACH win the Republican nomination. Only one can win.

So Intrade calculates the total margin requirement for each contest as the worst case if you lose ALL your bets. The margin necessary for one bet can be partially offset by the potential to win in a mutually exclusive bet.

For example, lets say you're short Romney at $5 a share. If Paul wins the nomination then your Romney short will also be a win.


Consequently you can *also* short Paul shares if you like, with a lower margin requirement because if you're wrong about Paul and he wins, any loss you'd have with the Paul short shares will be partially (or totally) offset by your corresponding gain in short Romney shares.

The point is, if you're willing to take a position in one of the "major" candidates, shorting one of the "also ran" candidates like Paul doesn't require much (or any) extra margin.

This Blog Has Moved!

My blog has moved. Check out my new blog at realfreemarket.org.