The prototyping of an automatic Trading System
碩士 === 國立中正大學 === 資訊管理學系 === 87 === Financial researches or technical analysis traditionally employed the closing prices of equities as their basis for analysis. It is arguable that if the singular closing price can sufficiently represent all the trading prices of an equity in a trading...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
1999
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/56049813964950716973 |
| Summary: | 碩士 === 國立中正大學 === 資訊管理學系 === 87 === Financial researches or technical analysis traditionally employed the closing prices of equities as their basis for analysis. It is arguable that if the singular closing price can sufficiently represent all the trading prices of an equity in a trading day. This paper made use of the intraday trading data to check if this high frequency data can facilitate us to screen out profitable trades more effectively.
The tick-by-tick trade data from NYSE was converted to trade data in three minutes interval to facilitate this study. The data was employed to check if the technical analysis using this intraday data could produce high return. The annual return was booked between 25.5% to 47.1% without transaction cost. The annual return fell in between 5.4% to 36.6% when proper transaction cost was introduced. The return was calculated based on a hypothetical account with $100,000 seeds fund. A trade was carried out when a buy signal was generated and sufficient fund was available in the account. A sell transaction will be carried out when a sell signal is generated. All holdings were liquidated at the end of simulated trading period.
An automatic trading system was developed to test this trading strategy using the real data from market. This trading system will issue trading orders to selected Web-based investment simulation sites, such as Final Bell. Buy/Sell trade(s) will be carried out by the trading systems whenever a trading signal is generated.
|
|---|