260 D. Pelusi
fore, there are some technical filters that assure the highest profit. In this way, an
important issue is the choice of the training and trading sets.
In this paper, we describe a pattern recognition algorithm to optimally match
a training period and a trading period in the DMAC filter rule. The trading rule
described is a long-short strategy investing in foreign exchange rates. We illustrate
a practical example choosing the semester as the testing period and obtaining stable
results. This stability is verified also for different periods, such as monthly, yearly
and two-yearly periods. Moreover, for these temporal ranges, we realise a statistic on
the short and long operations separately. In particular, we compute the mean and the
standard error of the operations number, obtaining some interesting information. It
might be convenient to also report standard indicators such as performance, volatility
and Sharpe ratio, typical of the finance industry.
The aim of this work is to obtain positive profits in accordance with similarity
degrees between training and trading periods. Our method gives a similarity index
that can be useful to establish how a training set has valuable information for a future
trading set. The results show that the similarity index is a good starting point for this
kind of study. Therefore, we will need to analyse how differences in shape have an
impact on profits for global similarity indexes of comparable magnitude.
References
- Allen, F., Karjalainen, R.: Using genetic algorithms to find technical trading rules. J. Finan.
Econ. 51, 245–271 (1999) - Arnold, C., Rahfeldt, D.: Timing the Market: How to Profit in Bull and Bear Markets with
Technical Analysis. Probus Publishing, Chicago (1986) - Berkowitz, J., Kilian, L.: Recent developments in bootstrapping time series. Discussion
Paper, no. 96–45, (1996) - Bollersev, T.: Generalized autoregressive conditional heteroskedasticity. J. Econ. 31, 307–
327 (1986) - Brock, W., Lakonishok, J., LeBaron, B.: Simple technical trading rules and the stochastic
properties of stock returns. J. Finan. 47, 1731–1764 (1992) - Brooks, C., Clare, A.D., Persand, G.: A word of caution on calculating market-based
minimum capital risk requirements. J. Bank. Finan. 14, 1557–1574 (2000) - Chang, P., Osler, C.: Methodical madness: technical analysis and the irrationality of
exchange-rate forecast. Econ. J. 109, 636–661 (1999) - Davison, A.C.: Hinkley, Bootstrap Methods and Their Applications. Cambridge University
Press (1997) - Dempster, M., Jones, C.: Can technical pattern trading be profitably automated? 2. The
head & shoulders. Centre for Financial Research, Judge Institute of Management Studies.
University of Cambridge, working paper (1998) - Dempster, M.A.H., Jones, C.M.: A real time adaptive trading system using genetic pro-
gramming. Judge Institute of Management. University of Cambridge, WP n. 36 (2000) - Edwards, M.: Technical Analysis of Stock Trends, 5th edn. John Magee, Boston (1966)
- Efron, B.: Bootstrap methods: another look at the jackknife. Ann. Stat. 7, 1–26 (1979)
- Efron, B.: The Jackknife, the Bootstrap, and other Resampling Plans. Society for Industrial
and Applied Mathematics, Philadelphia (1982)