|
 Business
 Trading the OEX or TSE
Temporal Rules for
Trading
TSE
Market Timing Applications
Intermarket Model for
S&P 500
A Hybrid Trading System
Tapping Financials
Databases
Science
Industry
 |
|
|
These examples discuss application
of our tools or the underlying mathematical methods in scientific areas such as medical
diagnosis and cost containment. Because of confidentiality considerations, we
discuss here only applications that were made available to the public.
Trading the OEX or S&P 100
Temporal Rules for Trading TSE
Market-timing Application
International Model for S&P 500
A Hybrid Trading System
Tapping Financial Databases
Business
| Title: |
Trading the OEX or S&P 100 |
Developer:
|
Skalkos and Associates
Salem, NH, USA |
| Description: |
The developer derived rules for "bearish" and
"bullish" sentiments (Hines index) with a subsequent up or down market. The
rules were used to achieve over 68 percent profitability on long trades and over 62
percent profitability on short trades. |
| Reported Benefits: |
- rules with good generalization qualities of the OEX
- identification of key indices for profitable short-term trading
|
| Reference: |
Chris Skalkos, "Rough Sets Help Time the OEX", NeuroVe$t
Journal. November/December 1996, pp. 21-28. |
Business
| Title: |
Temporal Rules for Trading TSE |
Developer:
|
First Marathon Securities
Regina, SK, Canada |
| Description: |
The developer derived relationships between 32 stock and
economic indicators including TSE, DOW, S&P, P/E and seven major industry indexes over
a six month period. The rules describe stock behavior of major TSE trading companies and
their sensitivity to market or economic indices. |
| Reported Benefits: |
- description of relationships and stock’s sensitivities to various indices
|
| Reference: |
Robert Golan and Donald Edwards, "Temporal Rules
Discovered Using DataLogic/R with Stock Market Data", RSKD-93, July 1993, pp. 61-72. |
Business
| Title: |
Market-timing Applications |
Developer:
|
Ruggiero and Associates
East Haven, Conn., USA |
| Description: |
A set of rules was developed to predict short and long
positions in the S&P 500 while recognizing different market price cycles. The
developer claims excellent performance of strong rules while discarding weak rules in
trading. |
| Reported Benefits: |
- correct calling of over 70 percent of the positive moves in the next 5 weeks
- average transaction represented over $ 25,000 profit per S&P 500 contract
|
| Reference: |
Murray Ruggiero, "Rules are made to be traded", AI
in Finance, Fall 1994, pp.35-40. |
Business
| Title: |
Intermarket Model for S&P 500 |
Developer:
|
Ruggiero and Associates
East Haven, Conn., USA |
| Description: |
Rules to predict strong rallies in the S&P 500 of 2
percent or more were developed. The rules were also used to override short signals from
S&P 500 neural network models and to add additional long entries. |
| Reported Benefits: |
- strong generalization rules for predicting S&P 500 "buy" and
"sell" signals
- improvement on neural network models used for trading
|
| Reference: |
Murray Ruggiero, "How to build a system framework",
Futures. November 1994, pp. 50-56. |
Business
| Title: |
A Hybrid Trading System |
Developer:
|
Ruggiero and Associates
East Haven, Conn., USA |
| Description: |
A hybrid trading system was developed incorporating neural
networks, rough sets and a spreadsheet. Neural network models were supervised by rules
generated by rough sets to correct for possible errors in predictions. The system allowed
the user to address and exploit inefficiencies that exist in the market. |
| Reported Benefits: |
- reduction of drawdown by 25 to 50 percent and increase of the average winner/loser ratio
by 50 to 100 percent
- average trade length was cut by 50 to 80 percent
|
| Reference: |
Murray Ruggiero, "Turning the Key", Futures.
December 1994, pp. 38-40. |
Business
| Title: |
Tapping Financial Databases |
Developer:
|
REDUCT & Lobbe Technologies Inc.
Regina, SK, Canada |
| Description: |
A comparative study of rough sets (RS) versus multivariable
discriminant analysis (MDA) for prediction of corporate bankruptcy from five financial
ratios: working capital, retained earnings, earnings before interest and taxes, market
value of equities and sales to total assets volumes. |
| Reported Benefits: |
- correct predictions for bankrupt firms were increased from 96.0 for MDA to 97.0 percent
for rough sets
|
| Reference: |
A. Szladow and D. Mills, "Tapping Financial
Databases", Business Credit, July/August 1993. |
 Top
|
