On Sun, 3 Nov 1996 ichudov@algebra.com wrote:
What I do is the following: I go to the Ladbroke's and offer to pay the gamblers not $6, but $6.01 if Dole wins. Being somewhat rational, these gamblers see a better deal than Ladbroke's offers, and give me their $1 bills. This is very simple.
In theory very simple indeed. But then there's the matter of trust (they know Ladbroke will probably be there after the election, but will you?) and market infringement (will Ladbroke's security allow you to hang around?) and such practical things. But I understand that you are more interested in the theoretic basis for arbitrage. I was talking more about the real world.
I take their $1 bills and go to "William Hill". I buy, however, LESS bets than dollar bills that I received. In particular, I buy $6.01 / $10.00 bets for each dollar that I receive.
Gambling institutions do these kind of insurance transactions all the time, of course. But many of them don't work only with small safe margins (changing the odds according to incoming bets pro/con so that exactly some percentage will always stay in their pockets after taxes) because they are themselves gamblers.
Of course, if gamblers could compare prices and choose gambling houses easily, no one would ever buy these bets from Ladbroke (unless they are crazy).
Some will anyway, out of convenience, if a Ladbroke office happens to be just around the corner. But those daring to give the highest odds, and in this case without insuring themselves with counter-odds, with take most of the customers and most of the profits if Clinton wins (and the losses if Dole wins).
This situation means that there is some market imperfection that does not allow arbitrage. It is not clear, though, what this imperfection is.
In part for practical reasons, as stated above. That will change when this kind of betting moves online, with digital cash (if allowed) or digital traceable money (betters will accept some degree of taxation). Then all the opportunities hitherto reserved for gamblers on the stock, commodity and monetary markets will become available to the more profane betters on sports, horce racing and elections: derivates, futures etc. And more. Some of the more esoteric cryptographic protocols will become of practical value in the gambling business. Like you could bet $n that Dole will win, prospective takers of the bet could make secret offers and the highest bidder would get your bet at the next to highest offered odds, without anybody's offer being revealed. You might have committed to take that offer, or you might not - different gambling styles. An all against all situation, serviced by a trusted entity with committed bits in escrow, living off a very small margin on all transactions. Asgaard