Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/22412
Authors: 
Bartels, Patrick
Breitner, Michael H.
Year of Publication: 
2004
Series/Report no.: 
Diskussionspapiere des Fachbereichs Wirtschaftswissenschaften, Universität Hannover 300
Abstract: 
Today´s derivative pricing base on stochastic models developed in the 70´s. These models base on some unrealistic assumptions. The system Warrant Pro 1 presented here combines the software agent PISA (Partially Intelligent Software Agent) and the neurosimulator FAUN (Fast Approximation with Universal Neural Networks) to synthesize market price functions instead of theoretical price functions. PISA automatically extracts data from the internet or other (semi-)structured text sources. High quality neural networks training and validation data with predefinable denseness are provided cost ree. FAUN learns true market price functions and computes real market prices. An example for German DAX options is outlined.
Subjects: 
Market prices
software agent
artificial neural networks
neurosimulator
derivatives
JEL: 
G13
F31
C80
M19
Document Type: 
Working Paper

Files in This Item:
File
Size
430.27 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.