Practice

Trading systems can be of two kinds: profitable, and the rest.

Throughout my career I have encountered quite a few systems. Some worked a little, some worked very well, but the vast majority of them were (for the lack of a better word) terrible. Trust me, if the system is available for free, it probably doesn’t work. But not all trading systems are useless, and there is always something to learn by studying a system.

By their nature, trading systems are like Lego sets. With the essential principles in mind – just as an architect would consider the law of gravity when designing a building – you build a system from the building blocks that are readily available. The building blocks of a trading system are not plastic bricks, of course, but rather the moving averages and other indicators that you will find in most charting packages.

Your job is to find the right combination of building blocks that work for the specific market that you wish to trade.  What works for gold may not work for oil; what works for oil may not work for coffee beans.  The nature of the market is the same, the principles are the same, but the specific combination of tools necessary is a little different in each case. 

As an example of a profitable system, let’s consider the system we designed for trading gold. This system has shown some very respectable results for the last three years over which it has been traded. It has performed so well because it is based on the core principles of supply, demand and an understanding of mass psychology rather than being based on a statistical inefficiency in the market.

The first important criterion is the win-to-loss ratio. Below are the trading results from one year’s worth of trading. As you can see in the table, the system made a total of 46 positive and 58 negative trades.   

profit loss  

At a first glance it doesn’t look too good, until you notice that the average win is twice the size of the average loss! Thus the system has an overall positive expectancy, and you can read more about ‘expectancy’ in the “characteristics of a winning system” section of this web site.

Below are the trading results for a period of 361 days. These are computer-generated results based on actual trades.  (Note that your ‘real life’ results would likely be slightly worse by about 10% due to operator inefficiency and slippage.)

indices

Take a moment to reflect on what you’re seeing here: a 130%+ annual gain from a system that makes more losing trades than winning ones!  While each individual trade may offer no better odds of success than a coin flip, over a series of 100 trades you can come out well ahead.

The system performance is based on a 3% risk control rule.  As you can see in the first table, the highest loss on any one trade is -$11678, which means that even in a worst case scenario the maximum loss was never more than the 3% risked initially on a trade.

To round off, take a look (below) at the equity curve for the system. The horizontal axis shows the number of hourly bars (5963 bars or 361 days), and the verticle axis shows the growing account size.

equity curve

As you have read in the “discipline” section of this web site, such results are only possible if you strictly follow your system. If you used your own discretion to choose which trades to take and which to avoid, then – believe me – the results would be very much different.

I strongly recommend you read Curtis Faith’s book “Way of the Turtle”. The author describes in great detail the psychological mechanisms that will help you follow your system with minimal effort.