Trailing TP - page 22

 
Maxim Kuznetsov #:

ATR (as it is) is a poor indicator. It is averaged internally, by high/low without taking their moments into account. So it "lags" and also lies a little bit

So it can be used, but only in some sketches.

Which indicator is better?

 
JRandomTrader #:

Which indicator is better?

you start with the ATR, then you think about what you want from it and what is wrong with it, make yourself a truer

and so on with all indicators. They are hospital average and for yourself they require rethinking and substitutions

just the ATR - it's indicative, on period 14 it's 7 bars behind reality. But it is cyclical (intraday) and it can be moved 24 hours-7 bars to the right. Take away the noise in the past and compare it with the current one. It will already be better.

 
Maxim Kuznetsov #:

ATR (as it is) is a poor indicator. It is averaged internally, by high/low without taking their moments into account. So it "lags" and also lies a little bit.

So it can be used, but only in some sketches.

I disagree. ATR - in my opinion, a great indicator of volatility, to which you can "bind" literally all indicators in the TS.

But, I should immediately note that we are talking about long, "position" trade with an average holding period longer than a day. ATR in this case is taken at watches and higher.

The ATR is probably not the best indicator for scalpers and news jitters.

 
Maxim Kuznetsov #:

you start with the ATR, then you think about what you want from it, and what is wrong with it, make it right for yourself

and so on with all the indicators. They are hospital average and for yourself they require rethinking and substitutions

just the ATR - it's indicative, on a period of 14 it's 7 bars behind reality. But it is cyclical (intraday) and it can be moved 24 hours-7 bars to the right. Take away the noise in the past and compare it with the current one. It will already be better.

Here is just a good illustration. ATR(14) "slows down" only if we use it for intraday trading. Then the time difference starts to play a role, and the lag appears. And if we are going to hold a position for a week or two, the lag disappears. In this case, the period can be three or five times longer. Or even switch to H4 or D1.

 
And yes, the coefficient at ATR can not only be variable, but also non-linear.
 
Aleksei Stepanenko #:

Valery, I've tried it, so far it's not improving. This search for the correct ranges, gives different sets of options, which introduce even more uncertainty in further actions.

Also tried it, the result was not encouraging. But purely intuitively I like it. But the ranges are really difficult to calculate that I would catch it immediately on a loss-making move and wait for the reversal in profit. Most probably, there should be many algorithms to calculate ranges depending on price behaviour. The channel width, rate of change, frequency of changes, regularity of changes, the same volumes. Such problems cannot be solved out of the blue)

 
Georgiy Merts #:

I disagree. ATR is, in my opinion, an excellent indicator of volatility, to which literally all indicators in the TS can be "tied".

But, at the same time, I should immediately note that we are not talking about long, "position" trading with an average hold on trades longer than a day. ATR in this case is taken at hours and above.

In scalpers and news jitter, probably the ATR is not the best indicator.

Read and think about the arithmetic of ATR calculation.

What kind of one on the hours when it averages and with regular and cyclical strong differences in hourly candles does it get @PA?

In its raw form or in the lower timeframes or already in the days - only there it shows something like that. On M30-H2 it's a finger in the sky, too big are the natural differences at the beginning and end of the atr period.

It's good to get something like that. Either in days its or in minutes-5 minutes, where natural cycles are not so breaking

PS/ In general all classic indicators and strategies are made for D1 and the place of their work is there (they even have default parameters for D1). For intraday work on M15-H1 they have to be corrected and replaced.

 
Maxim Kuznetsov #:

read and think about the arithmetic of calculating ATR.

What kind on watches, when it averages and with regular and cyclic strong differences in hourly candles does it get @PA?

In its raw form or in the lower timeframes or already in the days - only there it shows something like that. On M30-H2 it's a finger in the sky, too big are the natural differences at the beginning and end of the atr period.

It's good to get something like that. Either in days its or in minutes-5 minutes, where natural cycles are not so breaking

PS/ In general all classic indicators and strategies are made for D1 and the place of their work is there (they even have default parameters for D1). For intraday work on M15-H1 they have to be corrected and replaced.

No man has yet been born capable of proving anything to him
 
Maxim Kuznetsov #:


PS/ In general all classic indicators and strategies are made for D1 and their place of work is there (they even have default parameters for D1). For intraday work in M15-H1 they have to be corrected and replaced.

So I kind of said that we are not talking about scalping and micro levels on units of pips of five digits.

And ATR(14) on D1 - it doesn't behave much different from ATR(300) on watches.

I don't see where your words contradict mine... Same thing, only from the side.