Details
| Original language | English |
|---|---|
| Pages (from-to) | 319-333 |
| Number of pages | 15 |
| Journal | OPTIMIZATION |
| Volume | 47 |
| Issue number | 3-4 |
| Publication status | Published - 2000 |
Abstract
Today's option and warrant pricing is based on models developed by Black, Scholes and Merton in 1973 and Cox, Ross and Rubinstein in 1979. The price movement of the underlying asset is modeled by continuous-time or discrete-time stochastic processes. Unfortunately these models are based on severely unrealistic assumptions. Permanently an unsatisfactory and quite artificial adaption to the true market conditions is necessary (future volatility of the underlying price). Here, an alternative heuristic approach with a highly accurate neural network approximation is presented. Market prices of options and warrants and the values of the influence variables form the usually very large output/input data set. Thousands of multi-layer perceptrons with various topologies and with different weight initializations are trained with a fast sequential quadratic programming (SQP) method. The best networks are combined to an expert council network to synthesize market prices accurately. All options and warrants can be compared to single out overpriced and underpriced ones for each trading day. For each option and warrant overpriced and underpriced trading days can be used to ascertain a better buy and sell timing. Furthermore the neural model gains deep insight into the market price sensitivities (option Greeks), e.g., Δ, Γ, Θ and Ω. As an illustrative example we investigate BASF stock call warrants. Time series from the beginning of 1996 to mid 1997 of 74 BASF call warrant prices at the Frankfurter Wertpapierbörse (Frankfurt Stock Exchange) form the data basis. Finally a possible speed up of the training with the neuro-computer SYNAPSE 3 is briefly discussed.
Keywords
- Expert council networks, Financial modeling, Inexpensive neuro-computers, Neural networks, SQP-training methods, True market option and warrant pricing
ASJC Scopus subject areas
- Mathematics(all)
- Control and Optimization
- Decision Sciences(all)
- Management Science and Operations Research
- Mathematics(all)
- Applied Mathematics
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: OPTIMIZATION, Vol. 47, No. 3-4, 2000, p. 319-333.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Heuristic option pricing with neural networks and the neuro-computer SYNAPSE 3
AU - Breitner, Michael H.
N1 - Funding Information: The author gratefully appreciates the support by the Market Maker Software GmbH, Kaiserslautern (Data-Pool financial database), P. E. Gill, University of California San Diego (SQP methods), and the MediaInterface GmbH, Dresde:: (SYI\JIAXPS3E cmperatior?). Last but not least, the author thanks both reviewers for their valuable comments and constructive criticism which helped to improve the paper significantly. Copyright: Copyright 2005 Elsevier B.V., All rights reserved.
PY - 2000
Y1 - 2000
N2 - Today's option and warrant pricing is based on models developed by Black, Scholes and Merton in 1973 and Cox, Ross and Rubinstein in 1979. The price movement of the underlying asset is modeled by continuous-time or discrete-time stochastic processes. Unfortunately these models are based on severely unrealistic assumptions. Permanently an unsatisfactory and quite artificial adaption to the true market conditions is necessary (future volatility of the underlying price). Here, an alternative heuristic approach with a highly accurate neural network approximation is presented. Market prices of options and warrants and the values of the influence variables form the usually very large output/input data set. Thousands of multi-layer perceptrons with various topologies and with different weight initializations are trained with a fast sequential quadratic programming (SQP) method. The best networks are combined to an expert council network to synthesize market prices accurately. All options and warrants can be compared to single out overpriced and underpriced ones for each trading day. For each option and warrant overpriced and underpriced trading days can be used to ascertain a better buy and sell timing. Furthermore the neural model gains deep insight into the market price sensitivities (option Greeks), e.g., Δ, Γ, Θ and Ω. As an illustrative example we investigate BASF stock call warrants. Time series from the beginning of 1996 to mid 1997 of 74 BASF call warrant prices at the Frankfurter Wertpapierbörse (Frankfurt Stock Exchange) form the data basis. Finally a possible speed up of the training with the neuro-computer SYNAPSE 3 is briefly discussed.
AB - Today's option and warrant pricing is based on models developed by Black, Scholes and Merton in 1973 and Cox, Ross and Rubinstein in 1979. The price movement of the underlying asset is modeled by continuous-time or discrete-time stochastic processes. Unfortunately these models are based on severely unrealistic assumptions. Permanently an unsatisfactory and quite artificial adaption to the true market conditions is necessary (future volatility of the underlying price). Here, an alternative heuristic approach with a highly accurate neural network approximation is presented. Market prices of options and warrants and the values of the influence variables form the usually very large output/input data set. Thousands of multi-layer perceptrons with various topologies and with different weight initializations are trained with a fast sequential quadratic programming (SQP) method. The best networks are combined to an expert council network to synthesize market prices accurately. All options and warrants can be compared to single out overpriced and underpriced ones for each trading day. For each option and warrant overpriced and underpriced trading days can be used to ascertain a better buy and sell timing. Furthermore the neural model gains deep insight into the market price sensitivities (option Greeks), e.g., Δ, Γ, Θ and Ω. As an illustrative example we investigate BASF stock call warrants. Time series from the beginning of 1996 to mid 1997 of 74 BASF call warrant prices at the Frankfurter Wertpapierbörse (Frankfurt Stock Exchange) form the data basis. Finally a possible speed up of the training with the neuro-computer SYNAPSE 3 is briefly discussed.
KW - Expert council networks
KW - Financial modeling
KW - Inexpensive neuro-computers
KW - Neural networks
KW - SQP-training methods
KW - True market option and warrant pricing
UR - http://www.scopus.com/inward/record.url?scp=0040351384&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:0040351384
VL - 47
SP - 319
EP - 333
JO - OPTIMIZATION
JF - OPTIMIZATION
SN - 0233-1934
IS - 3-4
ER -