In this issue:
I. We need your vote!
II. NeuroShell tips (trading and non-trading versions)
III. NeuroShell Trader tips.
IV. Editorial Comment by Steve Ward, CEO
V. Ensemble estimators
VI. One way to stop this newsletter
*******************************************************
I. We need your vote!
No, not the US mid-term elections – we’re as sick of them as you probably are if you are in the US.
Ward Systems Group Inc. needs your vote for the Technical Analysis of Stocks & Commodities Readers’ Choice Awards. We know you’re getting hit with a lot of election ads this time of year, but think of it as your chance to vote for an Oscar. That’s pretty much what this contest means to software vendors, we just don’t have the red carpet, fabulous gowns, and extravagant gift bags.
But it also means something to you. We need new buyers to keep supporting the investment that you made in our software. Many prospective buyers rely on the TASC results for their buying decision.
Get your copy of the November Stocks & Commodities magazine and visit www.Traders.com. Look in the upper left hand corner of the web site for the 2007 Readers’ Choice Awards and click on the Vote Now! link.
On the page that is displayed enter your subscriber ID from your magazine address label (look at their example to see which numbers on the label are the subscriber ID).
After you are logged in, look for the Artificial Intelligence Software (Expert, Neural) category. Be sure to mark you ballot for NeuroShell Trader Professional (Ward Systems Group Inc.)
Do the same for the Standalone Analytical Software, $1000 and more category. Look for NeuroShell Trader. Please vote now before you forget.
The winners will be announced in the February 2007 Stocks & Commodities Bonus Issue.
*******************************************************
II. NeuroShell tips (trading and non-trading versions)
A. Unexpected problems at night with Interactive Brokers trading or long optimizations (including non-trading products).
Several users have reported unexpected logoffs from the IB TWS at night. The TWS forces a logoff once per day, and the default seems to be 10 PM. To set this logoff time in the TWS go to Configure->Global Configuration->Exit where you will be able to set the Auto Log Off Timer.
We do not know why the TWS forces you to select a logoff time. But the same dialog allows you to be prompted about the exit.
There are, however, other things that often happen at night that could affect your automatic trading or long optimizations, which you have some control over. Many of these can also affect long optimizations. These include but are not limited to:
1. Hibernation
2. Shutoffs of disks because of power options
3. The operating systems performing maintenance activities
4. Your virus detection performing scans and/or maintenance
5. Your ISP performing maintenance or timing you out
6. Problems with the exchange you are trading (usually reported by IB with a flashing red bar, but you may not be watching)
In view of all these things, please do not contact us about unexplained problems at night until you have completely researched all the above.
*******************************************************
III. NeuroShell Trader tips.
Getting the correct symbol and exchange for IB trading
The symbols Interactive Brokers uses are often not the same one your data feed uses, especially if you are trading something other than stocks. Often the exchange is different, too. Therefore, when you set up automatic trading, you need to “map” the data feed symbol to the IB symbol. In other words, NeuroShell translates the symbol for you when it places a trade. However, it is sometimes difficult to find out exactly what to tell NeuroShell to translate too.
For example, suppose you want to trade E-mini Dow futures, whose symbol is YM and it is on the CBOT Exchange. If you want the December 2006 contract, you would choose YMZ6 on most data feeds. That doesn’t work with IB, unfortunately, so you’d have to set up a mapping in NeuroShell (Tools -> Options -> Trading Orders -> Ticker & Exchange Mapping…)
But what should it be mapped to? The key to the correct symbol and exchange discovery is in the IB Trader Workstation (TWS). The first thing you have to do is insert the symbol there. The default way is to insert YM in the “Underlying column” and then press the enter key to get a dialog going to let you choose that you want Futures and the December 2006 USD contract. When you are done you do see that the exchange is ECBOT (the electronic CBOT), but all you see for the symbol is YM.
You can clarify things a little by going to the TWS View menu and select “Symbol Mode” instead of “underlying mode”. Now you can see the correct symbol which is “YM DEC 06”.
But wait! How many spaces are between YM and DEC and how many between DEC and 06? Turns out IB wants that to be exactly right. When we looked we thought there were two and one respectively. We never dreamed there were THREE spaces between YM and DEC, but there are, and you need all three to make IB happy!
There is a quick way you can make sure you have the right symbol before you put it in NeuroShell’s mapping. While still in Symbol Mode, go to the Ticker Menu and select “Add Symbol” . If you can get the symbol accepted by TWS there, then TWS will accept that symbol when NeuroShell sends it.
In the end we had to email IB tech support to find out there were three spaces, because we never dreamed we’d have to enter more than two! So the correct symbol is YMbbbDECb06 where each b is a blank space. If you have to email IB like we did, just tell them you are having trouble getting the symbol accepted in their Ticker -> Add Symbol dialog. If you mention NeuroShell, they will just assume it is a NeuroShell problem.
*******************************************************
IV. Editorial Comment by Steve Ward, CEO
We frequently hear how good our technical support is. We appreciate that, because frankly we take pride in the fact that everyone participates in providing you that support so that we all know your needs. It also means that all of our sales people know what they are talking about, because those same people help with tech support, including me. We have no plans to ship our support overseas, like so many companies are doing now, even though it costs a lot, and it isn’t always work that highly technical people like to do.
We want to keep bringing you that great support, as well as ongoing enhancements to our software. Right now we’re working hard to get our products running under VISTA. You can help us help you by voting for us in the Technical Analysis of Stocks and Commodities contest (see Article I above). Winning that contest convinces new people to purchase our software, which keeps us healthy so we can continue to pay dividends to you on the software investment YOU made when YOU purchased. Many of you made that purchase long ago, and you’re still getting free support because new people keep buying. You are still getting free incremental updates, and fairly low cost major upgrades. Let’s keep it that way.
*******************************************************
V. Ensemble estimators
Although the following article is written with NeuroShell Trader mechanics in mind, the concepts described are equally applicable to the NeuroShell Predictor, NeuroShell Classifier, and NeuroShell 2. In our old NeuroShell 2 seminars we often taught Ensemble techniques.
For more advanced users, the NeuroShell Trader can easily implement “ensemble estimators”, wherein several predictions or estimations are then fed into a higher level prediction or estimator for the final answer. In this way, the higher level estimator acts like a supervisor who queries his/her employees for their projections before producing his/her own.
To implement an “averaging” ensemble prediction, first use the Prediction Wizard to build three predictions, all based upon different inputs. Then use the Average3 indicator to average the three results:
Average3(prediction signal 1, prediction signal 2, prediction signal 3)
To implement a “supervisory” ensemble prediction, again use the Prediction Wizard to build three predictions, all based upon different inputs. Then use the Prediction Wizard to build a new prediction whose inputs are the three prediction signals.
To implement a “consensus” ensemble trading strategy, build three rules (made up of predictions or a combination of predictions and indicators). These are usually true/false indicators (classifications if you are using NeuroShell Classifier), and the final result is the best two out of three.
In the Trading Strategy Wizard, insert three rules for long entry, e.g.:
A>B(Prediction signal 1, 0)
BB High Breakout(Close, 20, 2)
CloseAvg>OpenAvg(Open, Close, 5)
In the Trading Strategy Wizard, specify that the buy signal should be generated if two of the conditions are true. Make similar rules for long exit, again specifying signal generation upon two of the three rules being true.
If you have the NeuroShell Trader Professional, you can optimize the three rules above.
For short entries, you may want unanimous agreement, so specify that all conditions be true.
It is easy to prove that consensus models work better than any of their parts if you think of your models at any time as being either right or wrong, as say in a reversal strategy where you are either long or short. This proof depends on the models being completely independent of each other, too.
Let’s say you have three models (A, B, and C) and each one has a probability of 0.6 of being correct at any given time. In other words, the models are 60% right. To figure the probability that the average (which is the consensus) is right, think of it this way. The average is right if 2 or 3 of the 3 are right. Now,
probability of A right, B right, C wrong = 0.6 x 0.6 x 0.4 = 0.144
probability of A right, B wrong, C right = 0.6 x 0.4 x 0.6 = 0.144
probability of A wrong, B right, C right = 0.4 x 0.6 x 0.6 = 0.144
probability of A right, B right, C right = 0.6 x 0.6 x 0.6 = 0.216
The sum of the probabilities is 0.648 so the average will be right about 64.8% of the time.
If you assume the models are 55% right, the average of the 3 is 57.475% right.
You can simulate this in an Excel spreadsheet and the numbers should come out pretty close.
*******************************************************
VI. One way to stop this newsletter from coming
It is really easy. Just change your email address and don’t tell us.
*******************************************************