Artificial Intelligence Trading
by THIJS DEKKERS
What is the first thing that comes to your mind when you think about Artificial Intelligence (AI)? You’d probably think about a human-like robot army rebelling against humanity, but what if we would tell you that you are already making use of multiple AI based applications every day?
AI sounds like something that will play a big role in our lifes 10, or 20 years from now, but in reality Artificial Intelligence is already being used in numerous applications we use today, including voice recognition software such as Apple’s Siri as well as a lot of Google’s products such as Google Search, Gmail and Google Maps.
The term “artificial intelligence” dates back to 1956 and belongs to a Stanford researcher John McCarth:
“Artificial intelligence (AI) is the ability of a machine or a computer program to think and learn. The concept of AI is based on the idea of building machines capable of thinking, acting, and learning like humans”
One of the core beliefs of Trading Engineered is that if a human manages so be successful as a forex trader, an AI driven computer can do it better.
How do we use Artificial Intelligence?
Enough about the theory behind it, you probably want to know how we at Trading Engineered use it to come up with our trading strategies right? Our working process consists of the following 5 steps:
Input: we gather as much data of forex instruments as we can.
Leverage AI: we use different types of Artificial Intelligence such as Neural Networks and Genetic Algorithms. Most of the times we train a so called ‘model’ with the data we gathered in step 1 and let our AI come up with different kinds of trading strategies.
Manual evaluation: a lot of time is being spent in translating the outcomes of the step above into trading strategies ‘we’ understand. A lot of strategies are simply not usable in real life scenarios and that’s why we manually evaluate the outcomes of the strategies generated by our AI software.
Backtesting: we convert these so called ‘AI-strategies’ to strategies that can be backtested using historical data and then run backtests with them.
Initial selection: we make a selection of the most promising strategies based on the backtesting results. We use a lot of different variables to make our selection including statistical power, expected return, drawdown etc.
Forward testing: the next step is forward testing. To reduce the risk for our customers we ‘forward test’ these promising selected trading strategies and check if the results from the forward test are comparable to those coming from the backtests.
If a trading strategy meets all of our criteria we might consider a trading strategy as a new ArtiTrader for Trading Engineered. To reduce the risk of our customers we make sure that we thoroughly test every different aspect a trading strategy consists of. The steps outlined above make up a big part of the work we perform here at Trading Engineered on day-to-day basis.
Now that you know how our unique trading strategies are being developed, why don’t you give one of them a try? All of our systems are temporarily on sale. View our pricing here.