Quote from efficiency:
Hmmm, independent testing. Let's see, way back in 1984, when you were dreaming about getting your finger wet, Futures Magazine, not to be confused with John Hill's Futures Truth, tested (12) systems upon (12) commodities over (7) years.
There have been dozens to hundreds of studies by everyone from amateurs to institutions. Picking out a 23 year old study which is neither defining nor the only one has almost no relevance. If you are trying to show that this is what things are based on, you have much learning and research to do.
It's debatable whether seven years is statistically significant
This demonstrates a serious lack of knowledge on your part.
a) 30 data points in statistics is the beginning of statistical significant. 7 years of trades with many RSI trades would be considered the final nail in the coffin, it would have such a strong confidence interval. But every study must be placed within the universe of all other studies to have relevance, such as factoring out intended or unintended biases in the study, such as cherry picking instruments, timer intervals, indicators, or other things to place something in a chosen light.
b) Show me ANY well-known vendor system that has a 7 year successful track record with Tech Analysis. Once they have 6 onths people start flocking to them as if they are God.
... Any oscillator will transcend to an over-bought (or oversold) state. Stochastics. CCI, %R even a Bollinger Band. Problem is , over-bought can remain over-bought for extended periods. Scant predictive qualities. RSI is basically a normalized oscillator. As such, it has a defined ceiling (and floor) regardless of periods employed. Accordingly, it will eventually diverge from the underlying price.
none of the above has any statistical relevance. Things morphing into other unproven indicators adds little value.
The RSI "works" when used as Wilder intended, DIVERGENCE. The wider the divergence, the more meaningful....
Well, now I am convinced. Your exhautive testing included with that statement really made it evident ... NOT
Ask any carpenter (or mechanic) about using the right tool for the job. Of course with your number of posts (vs duration), it doesn't appear you ask anyone anything.
Given your tunnel vision, I haven't seen you have anything to say. Throwing up an old article and attaching some viewpoints does nothing to disprove the original thought or to prove your point. Opinions have no relevance as to whether an indicator works.
Hmmm, independent testing. Let's see, way back in 1984, when you were dreaming about getting your finger wet, Futures Magazine, not to be confused with John Hill's Futures Truth, tested (12) systems upon (12) commodities over (7) years.
There have been dozens to hundreds of studies by everyone from amateurs to institutions. Picking out a 23 year old study which is neither defining nor the only one has almost no relevance. If you are trying to show that this is what things are based on, you have much learning and research to do.
It's debatable whether seven years is statistically significant
This demonstrates a serious lack of knowledge on your part.
a) 30 data points in statistics is the beginning of statistical significant. 7 years of trades with many RSI trades would be considered the final nail in the coffin, it would have such a strong confidence interval. But every study must be placed within the universe of all other studies to have relevance, such as factoring out intended or unintended biases in the study, such as cherry picking instruments, timer intervals, indicators, or other things to place something in a chosen light.
b) Show me ANY well-known vendor system that has a 7 year successful track record with Tech Analysis. Once they have 6 onths people start flocking to them as if they are God.
... Any oscillator will transcend to an over-bought (or oversold) state. Stochastics. CCI, %R even a Bollinger Band. Problem is , over-bought can remain over-bought for extended periods. Scant predictive qualities. RSI is basically a normalized oscillator. As such, it has a defined ceiling (and floor) regardless of periods employed. Accordingly, it will eventually diverge from the underlying price.
none of the above has any statistical relevance. Things morphing into other unproven indicators adds little value.
The RSI "works" when used as Wilder intended, DIVERGENCE. The wider the divergence, the more meaningful....
Well, now I am convinced. Your exhautive testing included with that statement really made it evident ... NOT
Ask any carpenter (or mechanic) about using the right tool for the job. Of course with your number of posts (vs duration), it doesn't appear you ask anyone anything.
Given your tunnel vision, I haven't seen you have anything to say. Throwing up an old article and attaching some viewpoints does nothing to disprove the original thought or to prove your point. Opinions have no relevance as to whether an indicator works.