Regensburg 2004 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Downloads | Hilfe

AKSOE: Physik sozio-ökonomischer Systeme

AKSOE 3: Postersitzung

AKSOE 3.3: Poster

Montag, 8. März 2004, 16:00–18:00, Poster D

No profit in spite of a perfect prediction? How interaction destroy arbitrage. — •Roland Rothenstein and Klaus Pawelzik — University Bremen, Institute for Theoretical Physics,
Otto-Hahn-Allee 1, D-28359 Bremen, Germany

The efficient market hypothesis stated, that -because markets anticipate future returns of stock markets- nobody is able to make a profit using past (public) informations. However not all markets are all the time efficient. Under the assumption that markets are not efficient, many traders try to find inefficiencies in the market and use them to build forecast systems. But every order has an influence to the market, which changes the price and contradicts the original prediction. Can a trader neglect its own influence to the market or is this influence so important that it has to be considered in the prediction of the future? Can interaction destoy arbitrage opportunities? We study a evolutionary stock market model with a trader (super agent), that is able to make a perfect prediction. In this market, we determine how the transactions of this superagent affects his ability to use his perfect knowledge to make a profit in this market. We measure this influence under different conditions in our market model like different number of agents, different memory of the agents or different evolutionary strategies.