How to Choose a Commodity Trading System

<   

Curve-Fit System to Fit the Current Market Cycle

All markets move in cycles and these cycles fluctuate continuously. Some cycles last only minutes; others for weeks, months, years, even decades. On all cycles, the plotted market data is seen to move between two, curved parallel lines. On longer cycles, these lines appear to be straight lines (trendlines), but they are in fact curved.

Long cycles are sporadic, but many developers will find a period in a market where one - or several - long, flowing cycles allow a system to produce fantastic results. The key word is sporadic as, again, things are unlikely to continue the same into the future and a programmed system will produce completely different trade performance than historic data

Similarly, most markets are in different cycles at any one time. I believe the basic system should work on all markets, in all time frames.

Why Would Anyone Sell a System?

It's a common question, and a valid one. If a system works and can make a large amount of money, why would anyone sell it? I can't see why anyone would sell a programmed system for a few hundred (or a few thousand dollars if they are really cheeky) if it actually worked. Apart from getting in a few thousand dollars a year from the people who'd buy from them, of course.

However, if you trade your system for other people, then it can be worthwhile financially to do so. For example, if you had your own system and $100,000 in trading capital, you would need to make 100% return a year to make $100K salary. It is possible, but unlikely as most CTAs will be pleased with a return over 30% a year. Instead, you could trade accounts for other people. Say you got 20 accounts of $100K, you have $2 Million in equity. You make a 25% return ($500,000) and charge 20% management fee; you have made $100K salary with the more realistic 25% return on account. And without the pressure of trading your own money and HAVING to make 100% return for a salary!