----- Original Message ----- From: "Robin Hanson" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, April 01, 2002 7:47 PM Subject: Re: Securities exchanges shutdowns
> Alex Tabarrok wrote: > > > Yes, in 1968 the exchange closed on Wednesday's in order to deal with > > backlog. French and Roll (1986) find that variance of stock returns on > > days when the market is closed is much lower than on days when the > > market is open which suggests that trading itself, rather than say > > information transmission, generates variance. > > Couldn't we interpret this as trading *creating* information, which is then > transmitted? > How can they measure stock returns when the market is closed? >From my understanding, trading expresses the preferences of two sides agreeing to it. When a trade is executed, this preference information becomes known to everybody else. Unless the trade happens "truly randomly", I don't see how it could be "creating information", except for the interpretation that knowledge about a particular trade only gives you part of the preference information (you don't know how low A was willing to sell or how much B was going to offer), and nobody knows what part that is going to be until the trade occurs, for if the parties involved knew that beforehand, they would have traded already. So, in a sense, each trade is random. (I am probably making several assumptions here. Would anyone care to point them out?) Given a piece of information, does it make a difference whether something is the original source of this information (if we can define such a thing), i.e. the creator of this information; or whether it is a transmitter of this information? If not, then transmitting information is the same as creating it. If it does, could you illustrate this difference? Can somebody resolve this philosophical muddle? Maybe then we can understand how Robin Hanson's theory would explain the phenomenon that inspired this thread. Gustavo
