Trading Algorithms – The Basics

In trading, like in life, we try to make processes easier and more automated so that we can concentrate on the important things! A trading Algorithm is designed to take a complex set of rules, complex formulas and combine them with mathematical models and human oversight to make decisions that traders may take too long to complete.

Imagine that a futures trader (or perhaps forex, stocks or even crypto trader) has combined 12 points of control using lots of different trading tools and sub-charts with special settings that, when combined in a specific order, give high probability trading signals. This sounds great, but for the human brain to take in more than 5 or so inputs, it would be very difficult for this trader to take advantage of all the trading signals. This is where Trading Algos help to save time and ensure that all trading signals are captured.

Please don’t get confused with Artificial intelligence here. A Trading Algorithm is simply a lot of code that mimics a complex set of rules programmed and monitored by a human. A logic circuit if you like, for example:

“If point of control A is green after point of control B is green and if point of control C passes point of control D before point of control E is broken, then a medium buy signal is generated”

Now this would be a very simple example of a logic circuit (set of rules) within a Trading Algorithm, but you get the gist of how Trading Algo’s are developed.

In some circles a true Trading Algorithm is totally automated and takes the trade in, manages the trade with stops and take-profit orders. But there is a less complicated starting point for Trading Algorithms, which involve making all those complex calculations of rules, simply placing the signal on a chart and alerting the trader that it has occurred.

This simplified method of Trading Algos is how Paul Bratby first had his xBratAlgo developed. Imagine Paul’s experienced traders’ Brain as a blueprint for a trading algorithm. Then the desire to keep charts as clutter free as possible. What you see below is a blueprint of Paul Bratby’ s thought process but even he admits, he is getting too old to focus on all of these as we discussed earlier!

• 9 Decision Points before Paul considers grading the
trade a 3* trade (long or short)
• Then another special Decision Point will make it a 4* trade
• Another for 5* trade
• And finally, a total of 12 Decision Points being correct before he grades a trade 6* trade

And then simply printing them on a chart as the signal candle closes and simultaneously send an alert to the trader (even on his trading platform app push notifications on his phone). Sample chart image below.

As you can see, there is nothing else on the chart as all of those indicators and tools with special rules have been taken away! The xBratAlgo works all this out in the background and simply prints the long or short signals when rules are met and even grades the trading opportunity. A clean chart takes away distraction and allows traders to concentrate on the signals.

The next evolution for Paul then was to totally automate the xBratAlgo, which sounds easy, his developers already had the Trading Algorithm coded!

Well now it gets interesting as more code needs to be added now for more rules! Questions and rules to be included – What about trade execution? Where is the initial stop loss? When do we move stops to manage trade? When do we take off the stop and replace with a profit target limit order?

And then, the developers had to take into consideration the
restrictions of individual broker platforms and what could and could not be done to trade this xBratAlgo trading algorithm automatically with all the above additional questions and more.

The first platform used was TradeStation, a well-established auto trading platform and Paul’s first fully automated versions of the xBratAlgo were produced for Oil, Nasdaq and S&P500 futures and micro futures. Next was NinjaTrader AutoTraders with the xBratAlgo which, again, were not that simple as a different platforms have different code and different trade execution rules for auto trading algorithms.

The next evolution for the future is to build an Artificial Intelligent Auto Trader for the xBratAlgo that continually learns from signals generated, market conditions at the time, economic data reactions and so much more.