In this issue:
I. New Elliot Wave add-on from Bowfort
II. Peregrine Financial announces NeuroShell interface to their brokerage
III. Crispmind announces MTFeed, a data and trading interface for MetaTrader
IV. Commentary by Steve Ward, CEO
V. Optimization ended early because the genetic algorithm could not reproduce
VI. One way to stop this newsletter
*******************************************************
I. New Elliot Wave add-on from Bowfort
Bowfort Technologies announced a new indicator add-on at the recent NeuroShell Seminar in Tucson. Bowfort Wave contains a number of indicators for Elliott Wave analysis. This product finds the patterns, targets returns and provides probabilities of those returns, all ideal for feeding into a neural net.
Bowfort Wave is built around a unique sophisticated pattern recognition engine and analyzes each pattern individually to see how well this pattern performed historically throughout time.
Bowfort Wave identifies both Bull and Bear Elliott 5-3 waves. However it’s not just about Elliott, it can also identify other wave patterns and groups of patterns.
More information on this and their other products can be found at their website:
http://www.bowfort.com/Products/cartViewItem?sku=NSBWV
*******************************************************
II. Peregrine Financial announces NeuroShell interface to their brokerage
Peregrine Financial Group, Inc. (PFG) and NeuroShell will now interface effective with the release of NeuroShell 5.4. This new capability will allow automated trade execution from NeuroShell to PFG’s BEST Direct trading platform. PFG is one of the largest non-clearing U.S. futures commission merchants with customers, affiliates and brokerage offices in more than 80 countries. PFG offers a diverse range of futures and forex trading products and services, from full-service brokerage to self-directed accounts, in both simulated and live trading environments. The company is a leader in futures, forex, managed accounts, trader education and online trading. PFG sponsors a variety of live trading venues so that traders can hone their skills by practicing in a competitive, but no-risk venue. Try PFG commission-free for 2 weeks* and receive a FREE subscription to SFO magazine as a special gift. Please visit the following site to receive information:
www.pfgbest.com/neuroshell
or contact:
Herb Kral, Director of Automated Trading, 800.453.3290,
hk***@pf****.com
*Clients are responsible for the clearing and exchange fees during the commission-free period, and after the 2-week trial commissions are $5 all-inclusive (round-trip) for all trades in electronic markets.
*******************************************************
III. Crispmind announces MTFeed, a data and trading interface for MetaTrader
MTFeed is an add-on for automated trading and feeding historical and real-time data into NST/NSDT from any broker that uses MetaTrader as its trading platform. Being easy to install and use, it will expand your list of available data sources for processing with NST/NSDT and trading in sync with the actual price data.
MTFeed offers a wide range of options: end-of-day and intraday, historical and real-time bar data (Forex, CFD, Futures, and more); real-time tick/sub-minute/volume/range bars; automated trading with feedback. Yet it’s almost transparent to your day-to-day NST use.
The add-on and related information is available on the website:
http://www.neuroshell-addons.com/
*******************************************************
IV. Commentary by Steve Ward, CEO
1. Tucson. This was a really good seminar because we had a great group of interested users. They kept Marge and I on our toes with intelligent questions adding to the discussion, which is the way it should be. There were users from Singapore, South Africa, Australia, and Great Britain. We had two users from Hawaii. The attendees we spoke to said they got more than their money’s worth from the seminar, and one person wanted us to extend it another day.
We ran until about 6PM every evening, and then the informal conversations began. There were also a couple of great presentations from Bowfort and Peregrine.
2. Bowfort Technologies. Bowfort announced and demonstrated their Elliot Wave add-on at the seminar. I have been wanting an Elliot Wave capability for years but could never figure out how to do it. Thank goodness Mark Simpson from Bowfort has figured it out. See article I above.
3. Peregrine Financial. David Sheehan, Regional Manager for Peregrine, was the driving force behind Peregrine’s decision to build Peregrine’s new NeuroShell trading interface using our “Tradepump”. He demonstrated the brokerage and interface that will provide an excellent alternative to Interactive Brokers. They have a paper trading platform, and their interface worked flawlessly in 5.4 beta2. If you want to do integrated automatic trading of futures and FX with NeuroShell, I highly recommend them. See article II above.
4. Crispmind. This is a brand new company that didn’t attend the seminar, but has built a well-tested MetaTrader feed/brokerage interface for NeuroShell. I have had a number of conversations with the principal and I can tell you he is a very smart user of NeuroShell with a product that provides some very interesting alternatives – check out the article above.
5. Melbourne. The Tucson seminar was such a success, we are already CONSIDERING having another at the Hilton again in Florida in March or April. If you are seriously interested in attending, email us at
sa***@wa*********.com
and let us know so we can make our decision.
*******************************************************
V. “Optimization ended early because the genetic algorithm could not reproduce.”
There is a lot of confusion and disbelief when users get this error message in NeuroShell Trader Pro or DayTrader Pro. We thought the wording that comes with it was pretty self-explanatory, but a little more of a layman’s explanation might help. The genetic algorithm optimizer is essentially doing selective breeding, just as people selectively breed race horses and other animals and plants. Everyone knows that in-breeding is not a good thing because the gene pool gets pretty stagnant. The same thing can happen in our GA. If too many of the solutions being breed inside the computer result in exactly the same backtest statistics (e.g., the same profit) then reproduction cannot produce new solutions. The GA senses that and shuts down.
So what causes the in-breeding? Usually lack of diversity in the parameters that are being optimized, so that for some reason almost any combination of parameters the GA tries at some point produces the same backtest results. Your model is too restrictive in some way, but there are many ways this can occur. If you want to send us your model when this happens in a chart with data saved we will try to give you suggestions on how to broaden your model. However, if you just try backtesting manually with different parameters the optimizer might choose, you will probably see for yourself what the problem is.
*******************************************************
VI. One way to stop this newsletter
It is really easy. Just change your email address and don’t tell us.
*******************************************************