Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Department of Electronic and Information Engineering | en_US |
| dc.creator | Chung, Suet-lam Felix | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/5145 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | - |
| dc.rights | All rights reserved | en_US |
| dc.title | Prediction of Hang Sang Index (HSI) blue chips : a neural network approach | en_US |
| dcterms.abstract | Neural networks (NNs) have been applied extensively to the prediction of securities and foreign currency exchange rates in local and aboard. However, none of the previous studies addresses the specific properties of Hong Kong Hang Seng Index (HSI). The objective of this work is to answer the question: Is it possible to use NN predictors to make trading decisions on HSI securities? Three utmost market focus blue-chip securities, namely Hong Kong Bank, Hang Seng Bank, and Hutchison Whampoa were selected for the present study. Two NN architectures, multi-layer perceptrons (BP) and Elman Recurrent Neural Networks (ERNNs), were investigated and two data sampling methods, namely discrete sampling (DS) and sequential sampling (SS), were used to assess the performance of the NN predictors. Three input parameters were used, they are weekly average of securities price, simple moving average (SMA) and the relative strength index (RSI). The input data duration is from 30th September 1994 to 6th September 2004. The predictors were trained to predict the average price of the next week, given the price of current week. Trade signals, BUY or SELL , were generated to calculate profit. The outcomes of these decisions were compared with the ground truth for computing profit or loss. The results show that both architectures are able to train NN predictors to make profitable decisions. BP performs better than ERNN in terms of mean squared error (MSE). DS has lower MSE than SS. Among the three predictors involved in the investigation, Hutchison Whampoa predictor has the smallest MSE and is the best predictor. The simulation results also show that the predictors of both Hong Kong Bank and Hang Seng Bank perform equally well with similar range of MSE, which further concurs the market custom that the price of these two securities are highly correlated. The results also confirm that NNs are target specific, which means that the prediction performance of some targets is better than that of others. By using the best performing configuration, i.e., BP with DS, Hutchison Whampoa produces the highest profit, a 169% per annum. The results conclude and answer the project question that NNs are able to provide profit-making decisions for HSI securities trading. | en_US |
| dcterms.extent | x, 91 leaves : ill. ; 30 cm | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2005 | en_US |
| dcterms.educationalLevel | All Master | en_US |
| dcterms.educationalLevel | M.Sc. | en_US |
| dcterms.LCSH | Hong Kong Polytechnic University -- Dissertations | en_US |
| dcterms.LCSH | Stock exchanges -- China -- Hong Kong | en_US |
| dcterms.LCSH | Blue-chip stocks -- China -- Hong Kong -- Forecasting | en_US |
| dcterms.LCSH | Financial institutions -- China -- Hong Kong -- Forecasting | en_US |
| dcterms.LCSH | Neural networks (Computer science) -- China -- Hong Kong | en_US |
| dcterms.accessRights | restricted access | en_US |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| b1809983x.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 2.8 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/5145