RSI indicator - page 73

krzysiaczek99
298
krzysiaczek99  

RSI vs RSX vs RSX on Jurik smooth

Here I made fast analysis of RSI versus its derivatives RSX and RSX on Jurik smooth.

You can clearly see here that e.g. RSX on Jurik smooth makes substantial overshoot,

price makes lower low, RSI lower lows but RSX on Jurik smooth made double bottom.

Both of them made also made higher high even RSI made lower high and also

overshoot 50 level when RSI found resistance there.

So RSI on raw data is the best !!!

Comparing RSX vs RSX on Jurik smooth perhaps RSX is better as it overshoots less.

Files:
rsx.jpg 418 kb
rsx.mq4 4 kb
Mladen Rakic
139663
Mladen Rakic  
fajst_k:
Here I made fast analysis of RSI versus its derivatives RSX and RSX on Jurik smooth.

You can clearly see here that e.g. RSX on Jurik smooth makes substantial overshoot,

price makes lower low, RSI lower lows but RSX on Jurik smooth made double bottom.

Both of them made also made higher high even RSI made lower high and also

overshoot 50 level when RSI found resistance there.

So RSI on raw data is the best !!!

Comparing RSX vs RSX on Jurik smooth perhaps RSX is better as it overshoots less.

The purpose of using something like that in RSX was to prevent the flattening effect of any kind of rsi calculation for longer periods (as you have noticed by now, the longer the rsi/rsx period is, the flatter (closer to 50 values) are the rsi/rsx going to be). That renders static levels useless (because they are going to entirely depend on the calculation length of the rsi/rsx)

So, that is the purpose of doing something like that : to enhance the rsi/rsx without having to adjust the overbought/oversold levels all the time (not talking about self adjustable levels this time). And it seems that the whole process does what it is meant to do in the first place

krzysiaczek99
298
krzysiaczek99  

Yes but overall structure, proportions and patterns of original data should stay. Instead it falsifies the most important points of it....

Mladen Rakic
139663
Mladen Rakic  
fajst_k:
Yes but overall structure, proportions and patterns of original data should stay. Instead it falsifies the most important points of it....

Which points are falsified?

Peak and valleys positions are not falsified nor is the momentum falsified. And taking into consideration how many false signals are we going to get using the "raw RSI", weighting what is more useful in any type of trading should be obvious

krzysiaczek99
298
krzysiaczek99  

RSI raw made lower low

RSX on jurik made double bottom

RSI raw made 1st high high than lower high (see two peaks of RSI)

RSX on jurik and RSX made 1st lower high than higher high

RSI raw hit resistance at neutral 50

RSX on jurik and RSX overshoot it

that falsification of main pattern of RSI

in another tests i simple tried smoothing methods of RSI (Ehler way, and others)

and RSI raw, when RSI levels were used as an exit of strategy which generates

thousands of trades so results are significiant and PF was always lower for smoothed

RSI than for raw RI so sorry but im 100% sure about it.

making and watching the indicators without proper backtesting in the strategies

which results are significiant (so hundreds or thousands not correlated trades)

is a bit disconnected from reality as simply there is no significiant proof that they work.

Mladen Rakic
139663
Mladen Rakic  
fajst_k:
RSI raw made lower low

RSX on jurik made double bottom

RSI raw made 1st high high than lower high (see two peaks of RSI)

RSX on jurik and RSX made 1st lower high than higher high

RSI raw hit resistance at neutral 50

RSX on jurik and RSX overshoot it

that falsification of main pattern of RSI

in another tests i simple tried smoothing methods of RSI (Ehler way, and others)

and RSI raw, when RSI levels were used as an exit of strategy which generates

thousands of trades so results are significiant and PF was always lower for smoothed

RSI than for raw RI so sorry but im 100% sure about it.

making and watching the indicators without proper backtesting in the strategies

which results are significiant (so hundreds or thousands not correlated trades)

is a bit disconnected from reality as simply there is no significiant proof that they work.

This part I agree 100%

making and watching the indicators without proper backtesting in the strategies

which results are significiant (so hundreds or thousands not correlated trades)

is a bit disconnected from reality as simply there is no significiant proof that they work.

and i am stopping taking part at this meaningless dispute based on subjective visual back tests only

happy trading

efendy
202
efendy  

hello, any body have or can code rsi level in All Time frame like ALLRSI indicator..and added more pair in one windows ?

Mladen Rakic
139663
Mladen Rakic  
f3n_dy:
hello, any body have or can code rsi level in All Time frame like ALLRSI indicator..and added more pair in one windows ?

Would you mind explaining a bit more?

Mladen Rakic
139663
Mladen Rakic  

Updated version of RSI adaptive EMA with alerts : rsi_adaptive_ema_nrp___mtf__alerts_nmc.mq4

Original was posted here : https://www.mql5.com/en/forum/178733/page44

efendy
202
efendy  
mladen:
Would you mind explaining a bit more?

i mean rsi all tf like the image and add some alert for tf we choose for alert show..

and also can add pair we choose more than one..so i dont need finding pair with oversold or overbought rsi manual draging pair..

Files:
gabungan.jpg 279 kb
gabungan_2.jpg 152 kb