The idea here is to let a simple, objective, mechanical algorithm draw a trend line and then to be LONG when prices are above the trend line and SHORT when prices are below the trend line. Here is an example, using SSO from the above table:
SSO Daily Trend Model
These Trend Models have beaten the pants off of my own analysis this past month, which says something about both my analysis and the Trend Models. Forecasting is easy when we are in sync, but much more difficult when we disagree. Looking back over the past seven months, the Trend Models win hands down. It looks from here that identifying a trend is a whole lot easier the forecasting one, a concept that from the looks of these results, can be taken to the bank.
Past performance does not guarantee future results.