Chop vs. Trend
- Thread starter Yag
- Start date
Quote from kut2k2:
Use Perry Kaufman's efficiency ratio, the perfect trend indicator.
Use the unsigned version.
If it's greater than +1/3, you're in an uptrend. If it's less than -1/3, you're in a downtrend. If it's between +1/3 and -1/3, you're in the chop.
Is it your experience that with +1/3 you miss a portion of a move and do you find extreme readings of the ratio to be significant reversal points?
And do you use different number of periods for different time frames?
THANK YOU for introducing me to this gem.
I tested this indicator and have strong doubts about its usefulness. Using the indicator on sp500 and nasdaq, it shows that sp500 is more trending than nasdaq. Anyone else has similar conclusion?
You're welcome.
I confess that I haven't used ker much in its basic form. I imagine it will always miss a portion of a move because it is not a predictor of the next trend, it is a detector of the current trend. I personally would never to use it as a counter-trend indicator.
I used it primarily as a building block to simulate Jurik's CFB indicator. That worked more or less but the result was very noisy, not as smooth as the CFB. My current research is focused on noise reduction in time series (ala Jurik's JMA). I'm convinced that is the key to optimal TA. That may be wrong but I haven't gotten to the point yet where I can claim one way or the other with certainty.
As far as different time periods: to quote someone else (John Merchant?), "a trend is a trend is a trend". I imagine the same lookback periods would apply to all time frames, but again, that's not something I've tested so it's just a guess.
The efficiency ratio is a measure of trend quality, not rate of trend. In other words, all monotonically increasing curves will have a ker of +1, regardless of rate of increase, and all monotonically decreasing curves will have a ker of -1, regardless of rate of decrease. I hope this helps.
Great help! Keep on posting.
Thx.
You guys are still trying to find the perfect indicator. It doesnt exist. There is no magical way to spot a trend vs. a chop.
However, I always consider the news. A big piece of news or economic report is a good catalyst for a trend. The market is more likely to trend during these times.
Also, scaling into your positions will make your 'chopped losses' become cheap. Trailing stops and pyramiding will help you take advantage of the trend.
This (money management) is why you dont need the perfect indicator.
However, I always consider the news. A big piece of news or economic report is a good catalyst for a trend. The market is more likely to trend during these times.
Also, scaling into your positions will make your 'chopped losses' become cheap. Trailing stops and pyramiding will help you take advantage of the trend.
This (money management) is why you dont need the perfect indicator.