>


What if the asset was some illiquid exotic/hybrid derivative? What I mean is that wouldn't you need a model of sorts. Just wondering .....

>

If I was a trader at some investment bank, and you called me up asking me "what I was carrying", what would be the motivating factor for me telling you? What would motivate any other trader to tell you what their book was?

MobPsycho is just trying to wind you up, don't fall for it!

P

MobPsycho has proposed what economists call the "direct revelation solution." Many problems that are complex if you analyze each individual's actions are simple if you assume everyone tells everyone else what they have and what they want. Sometimes people assume that the market somehow finds the direct revelation solution, but this seems like a heroic assumption. Most of the literature is devoted to designing systems that give everyone the incentive to go to the direct revelation solution.

The simplest example is an auction. Everyone has a value, Vi, for the object (the owner's reserve price is just one more bid). Everyone could reveal their V's, the object could be awarded to the person with the highest V, and this person would pay the owner some price between the highest and second highest V. All such solutions are Pareto-optimal.

In a traditional auction the price paid is the highest Vi. But this gives an incentive to lie. If I understate my true value, I might get the item cheaper (the risk is that I will not get it at all, although I was willing to pay the price it was eventually sold at). Each person has a complex decision, the object might not go to the person willing to pay the most, the owner might not collect as much as possible.

If the winning bidder instead has to pay the second highest bid, all incentive to lie disappears. Rational people will bid their true V, the object will go to the person willing to pay the most, and the owner will always receive the second-highest bid. Total utility increases, although depending on the circumstance some individual's utility could go down (but, if necessary, these individuals could be compensated in a manner so that everyone is better off).

I agree with MobPsycho that the financial markets are structured to encourage direct revelation. The most important protection against crashes is voluntary sharing of position information and needs. This is done in a manner that scrubs price direction information.

However crashes still occur. One important reason is that some people are doing illegal things, or things forbidden by their employers. These people will not share position information, despite their theoretic economic incentive to do so. This is why crime and crashes are so closely associated.

Simply saying "everybody must reveal" is not unlike saying "everyone must rent apartments in Harlem for $100 a month." It does nothing to convey the actual utility of what is supplied, to those who could supply more or less, in the form of an incentive. >>

That's not quite true. The direct revelation solution does not depend on everybody revealing, if it did it wouldn't be much good.

A closer example goes back to your original post. You wrote "What I can't understand is, how is somebody's first instinct not to start calling around and see what everybody else is carrying?" I agree with you that it would seem to be in everybody's interest to get together and share information, and I also believe (you may or may not) that a lot of financial machinery is designed to facilitate this. However the system does not work perfectly.

A direct revelation analysis would be one way to either (a) figure out how to make money from the imperfections or (b) design and impose a regulatory solution. I suspect you are more inclined toward the former.

The basic approach is to figure out how things would work if everyone did reveal, and who would be better or worse off. Assuming there is a net gain, you should be able to figure out a way to get to the same solution, using some of the tools of the analysis.

So while it is actually fairly straightforward to transition from routing and publishing orders, to buying, collecting, and publishing information packets, nobody gets it. >>

Maybe I'm not catching your meaning, it won't be the first time I've been way off. Anywhoo, I trade in an esoteric space so information is not that readily availible. That being said someone that just routes and publishes prices is not really adding a whole lot, the money made on picking price descrepancies is smaller than the amount made on the trade in general. Basically you have wide bid asks and everyone is reasonably close on both levels. That being said I think that on a daily basis I buy, collect and publish information packets. I'll look at where my position is what I want and what other people have sent me. I then agregate this info and send out my axes to people that I think can help me out and people that I do business with regularly. I don't think that this is too different from what most traders do every day. The people that are there for me on trades stay on the list, the people that take the information and never give back, still stay on the list because who knows, maybe one day they'll be there when I need them, and if they're not trading with me, I don't think they're using my info against me. The people that take the information, lean on me for more and are never there get removed from the list. I'm not publishing a list of what was just done or where I think things should be at, I publish a list of trades that I'm willing to get done and the levels that I care at. So I think that at least in my arena a lot of people buy (by trading with people), collect (by keeping track of where they're getting done at), and publish (sending out an axe) everyday. Just my .02