Billy
08-03-2011, 04:22 AM
9616
As unfortunate as a series of losing trades may be in any probabilistic mechanical system, these tend to happen in major intermediate term trend transition stages. Indeed, most of the time, the market is progressing into a relatively clear dominant trend, and is transitioning from one trend to another only a small percentage of the time. Therefore, the win/loss ratio of a good system will perform well most of the time because the statistics of high probabilities and expectancy are backtested on trending (up, down, sideways) market periods a majority of the time. The backtest covers only a small sample of transitioning trends and therefore the system’s statistics become less reliable and even misleading during actual trend transition stages.
The major lesson that can be learned from these successive losses is that the market has transitioned and entered a new major intermediate term downtrend. The robot should now quickly find itself back into its “normal” statistical environment and recover from the last blows.
The temptation is high to change some rules after a series of drawdowns, but because these changes will reflect more the transitional statistics that happen only a minority of the time, we must be absolutely sure that any change won’t hamper the robot’s performance under normal trending circumstances.
Ironically, the 20 DMF is now in cash because oversold, and looking for a new long entry based upon the LT/ST edges found by the robot.
The new limit long entry at 74.48 would only be executed in case of an extremely oversold situation. The resistance/support cluster ratios are all very bearish and are in favor of a final flush out in this downleg. It is far too late to find any good reward-risk setup for shorting and it is better to wait for a robot short signal after a significant bounce.
9615
GDX is in cash. Please refer to Pascal’s comments.
I will be out must of the day and back late today.
Billy
9614
As unfortunate as a series of losing trades may be in any probabilistic mechanical system, these tend to happen in major intermediate term trend transition stages. Indeed, most of the time, the market is progressing into a relatively clear dominant trend, and is transitioning from one trend to another only a small percentage of the time. Therefore, the win/loss ratio of a good system will perform well most of the time because the statistics of high probabilities and expectancy are backtested on trending (up, down, sideways) market periods a majority of the time. The backtest covers only a small sample of transitioning trends and therefore the system’s statistics become less reliable and even misleading during actual trend transition stages.
The major lesson that can be learned from these successive losses is that the market has transitioned and entered a new major intermediate term downtrend. The robot should now quickly find itself back into its “normal” statistical environment and recover from the last blows.
The temptation is high to change some rules after a series of drawdowns, but because these changes will reflect more the transitional statistics that happen only a minority of the time, we must be absolutely sure that any change won’t hamper the robot’s performance under normal trending circumstances.
Ironically, the 20 DMF is now in cash because oversold, and looking for a new long entry based upon the LT/ST edges found by the robot.
The new limit long entry at 74.48 would only be executed in case of an extremely oversold situation. The resistance/support cluster ratios are all very bearish and are in favor of a final flush out in this downleg. It is far too late to find any good reward-risk setup for shorting and it is better to wait for a robot short signal after a significant bounce.
9615
GDX is in cash. Please refer to Pascal’s comments.
I will be out must of the day and back late today.
Billy
9614