Summary: appears in the IEEE Trans. on Neural Networks, vol. 12, July 2001
FINANCIAL MODEL CALIBRATION
USING CONSISTENCY HINTS
Yaser S. Abu­Mostafa
California Institute of Technology
Abstract: We introduce a technique for forcing the calibration of a financial model to
produce valid parameters. The technique is based on learning from hints. It converts
simple curve fitting into genuine calibration, where broad conclusions can be inferred
from parameter values. The technique augments the error function of curve fitting
with consistency hint error functions based on the Kullback­Leibler distance. We
introduce an efficient EM­type optimization algorithm tailored to this technique. We
also introduce other consistency hints, and balance their weights using canonical errors.
We calibrate the correlated multi­factor Vasicek model of interest rates, and apply it
successfully to Japanese Yen swaps market and US Dollar yield market.
Keywords: computational finance, canonical error, consistency hint, cross entropy,
EM algorithm, financial engineering, interest rates, Kullback­Leibler distance, model
calibration, multi­factor models, overfitting, optimization, Vasicek model, volatility
term structure.
The author is with the Learning Systems Group in the Departments of Electrical Engineering,
Computer Science, and Computation and Neural Systems, at the California Institute of Technology,

Source: Abu-Mostafa, Yaser S. - Department of Mechanical Engineering & Computer Science Department, California Institute of Technology

Collections: Computer Technologies and Information Sciences