1 / 57

NEURAL NETWORKS FOR TECHNICAL ANALYSIS: A STUDY ON KLCI

NEURAL NETWORKS FOR TECHNICAL ANALYSIS: A STUDY ON KLCI. 授課教師:楊婉秀 報告人:李宗霖. Outline. Introduction Forecasting the Stock Market Neural Network and its Usage in the Stock Market A Case Study on the Forecasting of the KLCI Discussion Conclusion and Future Research. Introduction.

lilian
Download Presentation

NEURAL NETWORKS FOR TECHNICAL ANALYSIS: A STUDY ON KLCI

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. NEURAL NETWORKS FOR TECHNICAL ANALYSIS: A STUDY ON KLCI 授課教師:楊婉秀 報告人:李宗霖

  2. Outline • Introduction • Forecasting the Stock Market • Neural Network and its Usage in the Stock Market • A Case Study on the Forecasting of the KLCI • Discussion • Conclusion and Future Research

  3. Introduction

  4. Introduction • It is generally very difficult to forecast the movements of stock markets. • Artificial neural networkis a well-tested method forfinancial analysis on the stock market.

  5. Introduction • Using neural networks in equitymarket applications include: • Forecasting the value of a stock index • Recognition of patterns in trading charts • Rating of corporate bonds • Estimationof the market price of options • Indication of trading signalsof selling and buying

  6. Introduction • Feed-forward back propagation • Without the use of extensive market data or knowledge useful prediction can be made and significant paper profit can be achieved.

  7. Forecasting the Stock Market

  8. Forecasting the Stock Market • There are three schools of thought in terms of the ability to profit from the equity market. • Random Walk Hypothesis&Efficient Market Hypothesis • fundamental analysis • Technical analysis

  9. Forecasting the Stock Market • Use a variety of techniques to obtain multiple signals. • Neural networks are often trained by both technical and fundamental indicators to produce trading signals.

  10. Forecasting the Stock Market • For fundamental methods: • retail sales • gold prices • industrial production indices • foreign currency exchange rates • For technical methods: • delayed time series data • technical indicators

  11. Neural Network and its Usage in the Stock Market

  12. 3.1 Neural networks • Neural networks

  13. 3.2 Time series forecasting with neuralnetworks

  14. 3.2 Time series forecasting with neural networks • Three major steps in the neural network based forecasting proposedin this research: • Preprocessing • Architecture • Postprocessing

  15. 3.3 Measurements of neural network training • Normalized Mean Squared Error (NMSE) • Signs • Gradients

  16. 3.3 Measurements of neural network training • We argue that NMSE may not be the case for tradingin the context of time series analysis.

  17. 3.3 Measurements of neural network training

  18. A Case Study on the Forecasting of the KLCI

  19. A Case Study on the Forecasting of the KLCI • Kuala Lumpur Composite Index (KLCI) • Neural networks aretrained to approximate the market values. • To find the hidden relationship between technical indicators and future KLCI.

  20. Data choice and pre-processing

  21. Data choice and pre-processing • The major types of indicators • moving average (MA) • momentum (M) • Relative Strength Index (RSI) • stochastics (%K) • moving average of stochastics (%D)

  22. Data choice and pre-processing • inputs • It−1 • It • MA5 • MA10 • MA50 • RSI • M • %K • %D • output • It+1

  23. Data choice and pre-processing • Normalization

  24. Nonlinear analysis of the KLCI data

  25. Nonlinear analysis of the KLCI data

  26. Nonlinear analysis of the KLCI data

  27. Neural network model building • Historical data are divided into three parts • Training sets (2/3) • Validation sets (2/15) • Testing sets (3/15)

  28. Neural network model building • A trade-offbetweenconvergence and generalization. • Numberof hidden nodes

  29. Neural network model building

  30. Neural network model building • Primary sensitive analysis is conducted for input variables. • Low influencefactors: M20, M50, %K, %D • High influencefactors: It, RSI, M, MA5 • Chosen for input: It, MA5, MA10, RSI, M, It−1

  31. Neural network model building

  32. Neural network model building

  33. Paper profits using neural network predictions

  34. Paper profits using neural network predictions

  35. Paper profits using neural network predictions • In a real situation, this might not be possible as some indexed stocks may not be traded at all on some days. • The transaction cost of a big fund trading, which will affect the market prices was not taken into consideration in the calculation of the “paper profit”.

  36. Benchmark return comparison • Benchmark 1: • Benchmark 2:

  37. Benchmark return comparison • Benchmark 3:

  38. Comparison with ARIMA • Autoregressive Integrated Moving Average (ARIMA) Model • Introduced byGeorge Box and Gwilym Jenkins in 1976. • Provideda systematic procedure for the analysis of time series that was sufficientlygeneral to handle virtually all empirically observed time series data patterns. • ARIMA(p, d, q)

  39. Discussion

  40. Discussion • A very small NMSE does not necessarily imply good generalization. • Bettertesting results are demonstrated in the period near the end ofthe training sets. • We have no data to test the “best” model. • There are two approaches for using the forecastingresult. • best-so-far approach • committee approach

  41. Discussion • Measures • NMSE • Sign • Gradient • Fourchallenges • inputs andoutputs • types of neural networks and the activation functions • neural network architecture • evaluate thequality of trained neural networks for forecasting

  42. Conclusion and Future Research

  43. Conclusion and Future Research • It shows that useful prediction could be made for KLCI without the use of extensivemarket data or knowledge. • It shows how a 26% annual return could be achieved by using the proposed model. • It highlights the following problems associated with neural network based timeseries forecasting: • the hit rate is a function of the time frame chosen for the testing sets; • generalizability of the model over time to other period is weak; • there should be some recency tradeoffs.

  44. Conclusion and Future Research • A mixture of technical and fundamental factors as inputs over different time periods should be considered. • Sensitivity analysis should be conducted which can provide pointers to the refinement of neural network models. • The characteristics of emerging markets such as KLCI should be further researched to facilitate better modeling of the market using neural networks.

  45. Thanks for Your Listening

  46. Forecasting the Stock Market • The research done here would be considered a violation of the above twohypotheses above for short-term trading advantages in, Kuala Lumpur StockExchange (KLSE for short hereafter)

  47. Forecasting the Stock Market • There are three schools of thought in terms of the ability to profit from the equity market. • The first school believes that no investor can achieve above averagetrading advantages based on the historical and present information. • Random Walk Hypothesis • Efficient Market Hypothesis • Taylor provides compelling evidence to reject the randomwalk hypothesis and thus offers encouragement for research into better marketprediction.

  48. Forecasting the Stock Market • The second school's view is the so-called fundamental analysis. • It looks indepth at the financial conditions and operating results of a specific companyand the underlying behavior of its common stock. • In 1995 US$1.2 trillion of foreign exchangeswapped hands on a typical day [10]. The number is roughly 50 times the value ofthe world trade in goods and services which should be the real fundamental factor.

  49. Forecasting the Stock Market • Technical analysisbelongs to the third school of thought. • It attempts to use past stock price and volumeinformation to predict future price movements. • These five series are open price, close price, highestprice, lowest price and trading volume. • Most of the techniques used by technical analysts have not been shown to be statistically valid and many lack a rational explanation for their use

More Related