In this issue:
I. Enhancing the Cybernetic Analysis indicators
II. Commentary by Steve Ward
III. Jeff Parent, Director of the Canadian Society of Technical Analysts
IV. The CSTA welcomes the world Nov. 3-5, 2005
V. Don’t let you emails to us get lost!
VI. One way to stop this newsletter
*******************************************************
I. Enhancing the Cybernetic Analysis indicators
You can learn how to enhance the indicators included in the Cybernetic Analysis Add-on by reading Marge Sherald’s upcoming article entitled “A Genetic Approach to Your Profits – Turbocharge It”. The article will appear in the August 2005 issues of Technical Analysis of STOCKS & COMMODITIES magazine, which should be distributed mid-July. The article details how to examine a number of indicators included in John Ehlers’ book, “Cybernetic Analysis for Stocks and Futures” using the optimization capabilities of the NeuroShell Trader Professional.
*******************************************************
II. Commentary by Steve Ward
1. In our March 2005 newsletter I described how the NeuroShell Classifier can be an important tool in predicting the financial markets, especially when using fundamental data for long term predictions. There are maybe a half dozen companies with household names that I have assisted in building portfolio management systems in that fashion. However, the NeuroShell Classifier can also be utilized for shorter term profits, as long as you have some vehicle (like NeuroShell Trader Professional or TradeStation) from which to obtain the necessary input data streams, and from which the model can be called. In our next article, the Director of the Canadian Society of Technical Analysts (CTSA) describes his short term market prediction system which utilizes the NeuroShell Classifier.
2. And speaking of CTSA, Jeff is busy organizing the 18th Annual International Federation of Technical Analysts Conference, which could be a lot of fun in Vancouver (see our fourth article in this issue). Knowing Jeff, this conference should be a well organized affair.
3. With regard to Marge’s upcoming Technical Analysis of Stocks and Commodities article (see article I in this newsletter), we have been very pleased with the sales of the Cybernetic Analysis addon. Marge has actually used the addon for some trading systems that we use in-house and others that we have built for institutions. John Ehlers, who is probably the most famous engineer in trading, actually did the initial NeuroShell code for the indicators of the addon, for which we are very appreciative. This addon is very easy to read about before you purchase, because John’s book is the “manual” and “help file”. Just go purchase “Cybernetic Analysis for Stocks and Futures”, published by John Wiley & Sons, ISBN: 0471463078.
4. Some of you have heard me brag about how a large company that makes vehicles uses our neural net technology to make investment models for their pension fund assets. I could never name that company, and still cannot. However I now feel it is necessary to tell you who the company is NOT – it is not General Motors! Perhaps if their pension fund did utilize our technology, they wouldn’t be in the pickle they are in now.
*******************************************************
III. Jeff Parent, Director of the Canadian Society of Technical Analysts
Jeff Parent has had success in modeling the US stock markets using the NeuroShell Classifier. He collects technical analysis-based data taken at the outset of an outlier event and uses the data as inputs for a classifier net. Many of the trained nets correctly identify close to 70% of the outcomes from the validation set. This is a good result considering the notoriously noisy stock markets. Parent has been analyzing stocks for over 12 years and has authored articles for Technical Analysis of Stocks and Commodities Magazine.
Model Creation
Many analysts and academics do not believe it is possible to make profitable, short-term predictions of the stock markets. It is often said that markets are efficient in adjusting price to reflect changes in value thereby providing few trading opportunities. In general, Parent agrees with this statement. However, by focusing on outlier events only, he has been able to increase the odds of making successful predictions. At the beginning of an unusual price movement, more so than in a normal market, he believes there are clues to a potential outcome embedded in the technical factors of a stock. Working with a classifier net is natural because the results he is looking for are in the ‘event will/will not likely occur’ form.
An example of an outlier event is a stock trading at an unusually low level from a previous day’s average price. The outcome to be determined is whether or not the stock will revert back to the average price on the following day. As one possibility, Parent defines a low level as a stock dropping in value by a percentage of its average volatility. Technical indicator values like RSI and relative volatility and general market conditions are collected on the day before the drop. Knowing the condition of the markets puts the stock’s movements into perspective. Equally important is knowing how the stock has acted over the past several days, so recent values of the same technical indicators are included. In all, about 10-12 values are known at the moment before the event. The final piece of data collected is the outcome. After the ‘dip’, did the stock revert back to the average price or not?
Parent describes the process, “Up to ten years of data from actively traded stocks are scanned to net about 40,000 ‘dip’ events. Once collected, the event data must be divided into training and validation sets. Events are sorted by date and separated into somewhat randomly selected chunks of 6-12 month segments. One portion is set aside for validation. The balance is used to for training purposes. Having equal amount of both outcomes in the training set is necessary to ensure the network finds a non-trivial solution. Therefore, equal amount of data from positive outcomes (the stock reverts to the average price) and negative outcome are added together to complete the training set.
Validation and Usage
He continues, “I use the Classifier’s TurboProp2 method. It trains faster and I intuitively believe it is more likely to be aligned with the task. To be considered a success, I look for low false-positive rates and consistency between training and validation results. Normally, the ‘dip’ buyer strategy requires setting price alerts or placing limit buy orders on up to 500 stocks each day. Due to capital limitations, positions are taken only in the first 10-20 stocks hitting their targets. Using the trained network the same list is screened. A smaller watch list is created and profit on each position taken is potentially better. I have been very happy with the results. The Classifier is easy to use and very powerful.
*******************************************************
IV. The CSTA welcomes the world Nov. 3-5, 2005
As host of the 18th Annual International Federation of Technical Analysts Conference, the Canadian Society of Technical Analysts (see article above about CTSA’s Director, Jeff Parent) is proud to invite all IFTA Colleagues and other technical analysts to meet in beautiful Vancouver, British Columbia, for three days of learning, networking, socializing and enjoying the natural beauty and cultural diversity of Canada’s gateway to the Pacific. For more information, please visit this website:
http://www.csta.org/events/conferences/ifta2005.html
*******************************************************
V. Don’t let your emails to us get lost!
A good deal of your tech support and sales questions come to us by email to support@wardsystems.com and sales@wardsystems.com. Like many of you, we have to use spam filters to keep ourselves from getting buried by mass emailers. Sometimes, though, legitimate email finds its way into the junk folder, and we might not see it if you use really brief or general subjects. The best thing to do is make sure the product name is in the subject line like “NeuroShell 4.7 question serial number 540133121”. If you use a subject like “Hello” or one like “Question”, then your email could look like spam. Help us help you by taking a little more time to write your emails.
*******************************************************
VI. One way to stop this newsletter from coming
It is really easy. Just change your email address and don’t tell us.
*******************************************************