[kimchi]

I haven't looked at your tables in detail, but this looks like a channel breakout system using year-wide channels.

I've tested similar models using an adaptation of richard Dennis' sytem adopted by his 'turtles'. I didn't have reliable data for commodities going far enough back in history (where did you get your data?), so I primarily used Indices.

Are you trading EOD or does your timeframe allow you to use weekly data?

Many channel breakout sytems use an exit set at around half the entry channel. ie - maybe you use an exit of a breakdown over a 100 day channel for longs and 100 day breakout for shorts. Can you specify your exits?

Also, systems such as this which use a channel of around a month would use stops placed at around 3xATR. If you're trading longer term, how wide are your stops (I'd think up to 7xATR for a year-wide channel). This would mean your position sizes will be small, being limited by the larger initial risk exposure at your entry point.

If you're using a long time-fram such as a year, how do you manage adding to positions?

When I tested these systems I found a success rate of less than a third of trades, yet those fat tails produced 1 or more large winners each year which managed to turn the year into profit. This also led to large drawdowns and many many successive trades without a win making it a difficult system to follow.