|  NOV 1991

Developing A System by John Sweeney

Developing A System by John Sweeney The great thing about many of the new trading support systems is that they facilitate the respectable practice of ""dinking around"": that process of tweaking every crazy idea that strikes one until something productive and, hopefully, profitable pops out. I've found this is best done on quiet Sunday afternoons. Now, this article is not for everyone. I'm going to take you through the grind-it-out process of developing a trading system. As you'll see, even with nifty software and fast computers, there's mostly sweat and dirt before you get to the gold—if you get any at all. Perhaps this is really for those new to the process, with stars in their eyes. Why show all the laborious detail? It's a reality check. Many people buy computers and software expecting their lives to get easier. However, my jaded experience is that the problems just change, old abilities fade and the new abilities take a while to become productive. An old boss of mine said: ""A new system never makes money. It just replaces lots of clerks with expensive systems analysts."" Even though I'm an advocate and a user of computer trading systems, I don't want anyone to get the idea that systems are a bed of roses. The initial gain is the elimination of wishful thinking that goes with visual inspection/analysis. Later, with practice, productivity is improved. Hence this cautionary note on the real world of trading systems development.

by John Sweeney