That is the question for me. Sometimes, I am late in identifying when the trend has reversed, and that takes away a good chunk of my profits.
I think I misunderstood. I was considering "taking large losses" at the end of a trade (as you wrote in your original post) and losing a
portion of your profits at the end of a trade as two different things. I think losing
some profit is inevitable if one is attempting to milk every penny out of a position that is humanly possible. It might all depend on what one views as "a good chunk."
To over simplify my system... as long as price is beneath the white moving average AND the measure is sloping downward, I am ONLY interested in shorting an asset, and only so long as the gist of the price flow is headed south.
Conversely, if price is above the white moving average AND the measure is sloping upward, I am ONLY interested in long positions, and only so long as the gist of price flow is headed north.
I should add however that from my perspective, the decision on WHICH measures to use is
crucial. The notion that there are no "best" moving averages to use when trading is not one to which I subscribe. Indeed, at the heart of my system is the use of
carefully selected baselines which I calculated in a manner it would take too much space to describe here.
And yes, I am fully aware that individuals such as Norm Fosback, the former head of the Institute for Econometric Research, believe there are no magic numbers in trend following, and that it should be a basic requirement of any moving average trend following system that practically all moving average lengths predict successfully to a greater or lesser degree.
I have also read that the moving average one chooses is not as important as getting familiar with the way in which price interacts with it. But with all due respect,
I regard both of these contentions as a bunch of baloney!
The way I see it, even Fosback's own statement suggests the possibility that there might indeed be "magic" numbers in trend following. For if practically all moving average lengths predict successfully to a greater or lesser degree, it follows that those which predict successfully to a
greater degree are the
better moving averages—which would in turn infer that the moving average which predicts successfully to the
greatest degree is the
best moving average of all.
So then, I simply waded through all the data I archived over the years to arrive at the two moving averages which I now regard as superior.
This is why the idea of applying the exact same standard settings to different time frames (i.e., the 10-, 20-, 50-, 100- and 200-period moving averages to five-minute charts, 60-minute charts, daily charts, etc.) has always struck me as
counter-intuitive.
Of course, I have had many contributors to this forum who possess far more experience and knowledge than me eagerly explain how it is
impossible to use moving averages in the manner I suggest.
According to them, the core beliefs on which I based my suppositions conflict with, and are therefore discredited by "the findings of practically every available objective, independent, systematic, statistically significant research trial ever conducted and published on the subject."
But as far as I'm concerned, that's quite alright. As long as I can make money by conflicting with the findings of every available objective, independent, systematic, statistically significant research trial ever conducted and published on the subject, I am perfectly happy to do so.