>The trouble is, what happens if the "right" bid/ask contains a wide
>spread? I.e., what if the right bid is 20 and the right ask is 80? You get
>no trades, except those made by eternal optimists... which don't help the
Then the traders are a bunch of wimps (sorry, had to say that). Seriously,
with a 'large enough' trader population, bids should converge to a price
(as the marginal trader is willing to take on the risk, is willing to
profit off of the spread, etc). So if the 'right' bids are 20/80/0, I'd say
it is a limit of the present market. If you can find a way to increase the
ratio of trading to the number of issues traded, then the spread should
>...One further possible change
>that could be made would be to make the prices finer by adding decimal
>places. It's too easy for a bid/ask to hit 5/6 and then never trade,
>whereas if it were 5.49/5.51 it might be trading around in that range
>quite a bit.
You've said that before, and I don't understand your reasoning.
Is it simply because the resolution is too course? The difference in your
example is a payoff of 20:1 vs 17:1. How much more accurately can you
Or is it too expensive to cover the spread? On several occasions (more
often than I care to admit), I have mistakenly sold shares when I meant to
register a buy. Every time I have been able to back out and replace the
order at the same price. The spread cost me nothing in each case.
I would counter and say that the problem at the fringes of the price range
are due to the effects of MY pet theory :-). You need a high ratio of
funding in order to get trading at extreme price ranges. Given equal
investment by each trader, the accuracy of the price is affected by the
resolution of that ratio of Yes to No traders. And..., given that we only
have a total population of 300 traders to start with, it is going to be
hard to maintain high/accurate ratios on all extreme claims.
So if the Yes/No ratio is insufficiently populated, you can get one of two
results. Either a small population can bid a small amount and
overcompensate the price toward 0.5, or there won't be any high-return bids
at all. I would argue that is where we stand today.
Have you tried testing your theory by simulating the higher resolution with
a scaled claim linked to a low trading claim?
On another note:
>Every trade indicates a disagreement on price. If everybody agrees, no
>trades will occur. Only when A thinks a contract is worth more than P,
>and B thinks it worth less than P, does B sell it to A at price P.
> Randy Hudson <email@example.com>