Kevin Davey – Creating an Algorithmic Trading System

I'll make the answer VERY simple, even though "algorithmic" sounds super complicated.

Think of an algo as "rules to trade by."

If you want to buy 52-week highs and sell 52-week lows, those are rules that can be turned into a computer algo.  Maybe you add a stop loss in there or even a profit target. Basically, you program the rules you want (the algo), and then test it on whatever stock, forex pair, futures instrument, or crypto that you want.

An algorithmic trading system, then, is just a set of rules to control your trading decisions.  These rules can include buy/sell decisions, position sizing rules or even concepts like equity curve trading.
​

Most retail traders trade one of the following trading systems:

Full 100% Algo - ALL rules for entry and exit are determined by the computer.  No intervention by the human trader.  Can be easily historically tested (known as a backtest).

Hybrid Algo - Rules are defined, but sometimes the human trader overrides the computer rules/decisions.  Many traders fall into this category since they like rules, but they also think they are smarter than any algo they create.

Discretionary - Trading is conducted via guidelines rather than hard and fast rules.  The trader uses a lot of judgment and intuition to make trading decisions.  Most traders start out with this approach since they feel they can easily outsmart the market (usually a bad assumption!).
​

This is a complex topic, for sure.
​
I could write a book on how to create an Algo trading system - in fact, I actually did!  My first book is called Building Winning Algorithmic Trading Systems.  It was released in 2014, and in many trading circles, it is now considered a classic in trading system development.

My approach to building trading algos is a multi-step process I call the Strategy Factory®.  It is an objective process, built to help traders avoid common mistakes like over-optimization, curve fitting, etc.  The best part?  This process is what I use to develop trading strategies AND it works for traders I have taught it to.

I've also written 4 other algo trading books.  They are all best sellers at Amazon.  Here is a time when 3 of them were in the top 5 bestsellers for Futures Trading!

If you are thinking about Algo trading and creating your own algos, there are certain skills and experience you need:

Math & Statistics Knowledge
Programming Experience
Trading Platform Knowledge (how do you automate live?)
Mental Discipline (to not override your algos)
Ability To Deal With Frustration (Algo trading is NOT easy)

I've created this simple interactive quiz that will tell you pretty quickly where you stand relative to algo trading.  I also provided tons of free info at this site - KJTradingSystems.com - to get you rolling on your algo trading journey!

​Also, take a look at my Ultimate Guide To Successful Algo Trading.  It will help you get started.

And if you are looking for a detailed comparison of some of the best algo trading courses, check out my "best algo trading courses" guide.

This is the goal.  And what is a "successful" strategy?  A winning Algo trading strategy PRODUCES GOOD RISK ADJUSTED PROFITS IN REAL TIME.  That is the goal.

This is a VERY important point.  Traders new to algo trading mistakenly think the goal is creating a great-looking historic backtest.  WRONG!!!  In fact, creating a great backtest is many times the exact wrong thing to do.  Don't fall for "educators" who claim you can optimize your way to profits.  You can't.

There is a paradox here, though.  Optimizing is a bad thing if you do too much of it.  BUT, if you optimize properly, it can really boost your algo trading.  This is kind of like the Goldilocks story - the porridge is not too hot, and not too cold, but is just right!

There are a ton of choices to make when setting up your algo trading.  Many traders find my "best of trading" compilation very helpful in finding their way around.

If you want to succeed in algo trading, you need a process to develop algos.  Ideally, the process will be objective, rigorous, and comprehensive.  I've created such a step-by-step process for my own trading.  I can take a trading idea, develop a strategy based on it, test it and determine if it is worth trading.

The problem is most ideas fail!  The large building process I use routinely discards 99 to 99.9% of strategies I test.  And that is OK by me - I only want to trade the best algorithms.  I'll let losing traders trade the others!

Many traders like to say "show me the money."  So, on this site I give away an algo for the mini S&P.  It has returned over $50,000 in hypothetical profits per contract in the past 5 years - not too shabby!

It can be yours free, just sign up for my e-mail list at the top of the page.