We are pleased to announce a two-month fully functional trial of DLPAL S, a software program that identifies trading strategies automatically from historical data and writes code for a variety of popular trading platforms.
We have decided to offer a two-month fully functional trial because quants nowadays need more time to develop strategies in a trading environment dominated by algos and there is simply no way of evaluating software potential without full code capability.
DLPAL generates code for strategies and systems for the Quantopian platform, Tradestation (EasyLanguage), Multicharts (EasyLanguage), NinjaTrader 7 and Amibroker AFL. Therefore, testing advanced trading strategies derived from price action using machine learning is just a copy and paste away.
We pioneered the concept of automatic synthesis of trading strategies and code generation in early 2000s and our articles were featured in several trading publications (For example, “Price Pattern Autopilot”, Michael Harris, Active Trader Magazine, September 2002, p. 70.) where we showed examples of strategies and code generation for Tradestation and other platforms.
After we introduced the concept in the early 2000s there have been several attempts to copy the idea, most unsuccessful. Only one developer was appreciative enough and had the courage to offer us credit in a blog post, although most, if not all, were motivated by our ideas presented in articles and books.
Quantitative trading has gotten more demanding in the last few years and quants must continuously update strategies and look for new edges. Although there are strategies developed by our algos many years ago that still outperform benchmarks, in general it is not recommended to rely on any strategy developed in the past because of high probability of failure. We recommend rebalancing strategies annually at the minimum. Therefore, a two month fully functional trail does not pose any risks to our business model any longer and does not compromise any edges. Furthermore, it takes at least two months to become familiar with the concept in sufficient depth so that it yields good results. Anyone thinking it is possible to click GO and get a profitable strategy from software without doing any work at all is naive and any vendor that advertises that is probably shady. There is a lot more into the process than having a program to data mine strategies, far more than those who are unfamiliar with the idea think. But this is a potentially rewarding endeavor if done correctly. Unfortunately, most start wrong and end even at a worse position. This is how one of our customers from Germany has summarized it:
“In my opinion people buy software and want immediate success without work and without thinking about the code of the computer generated strategy. That is the reason why most of them fail…” W. S., Germany.
Features and screenshots
Below is a screenshot of the results of a search in SPY historical data. DLPAL S found 17 strategies that fulfilled the user-defined criteria.
Below are backtest results of all the strategies combined:
This is the result of testing the robustness of the first strategy in the results to variations in exit levels:
Among many other useful functions, DLPAL S includes a price series analysis tool:
Randomization analysis assists in ranking the performance of strategies based on results from random strategies.
A two-month fully functional trial of DLPAL S can be requested by sending an email to: email@example.com or here.
Before downloading DLPAL S please take a few moments to read our disclaimer.
If you have any question you can contact the DLPAL S team at firstname.lastname@example.org