Modelling Financial Time Series

Rating:
Summary: Rigorous but practical survey of time-series for traders.
Review: The first edition was published in 1986. It is EXCELLENT. Taylor rigorously studies the use of nonlinear time-series (Box-Jenkins) methods to trade a variety of financial markets, including individual stocks, stock indices, currencies, metals, and agricultural commodities, finding that there is a small trend component in most markets that can be profitably traded. Taylor performed testing of time series back in the early 1980s, when computer power and financial data was much scarcer and more expensive. I am excited to see what he has come up with, now that computers and data are a zillion times cheaper.

Rating:
Summary: Modelling Financial Time series
Review: This is (still) an excellent book, ahead of its time when published in 1986. One of the issues he dealt with was the possible modifications to option pricing models (black-scholes type) that could be made to account for trending markets. This was before the crash of '87 and the subsequent wide-spread adoption of skewed volatility smiles and risk-reversals into option pricing.

This book is probably out of print permanently, but the author is working on a new book, the provisional title of which is "Asset Price Dynamics and Prediction" Target Date of March 2004. The chapters are loosely based on subjects covered in Modelling Financial Time series.

Chapter headings for Modelling Financial Time series:

1. Introduction
2. Features of financial returns
3. Modelling price volatility
4. Forecasting standard deviations
5. The accuracy of autocorrelation estimates
6. Testing the random walk hypothesis
7. Forecasting trends in prices
8. Evidence against the efficiency of futures markets
9. Valuing options
10. Concluding remarks
Appendix : a computer program for modelling financial time series