Discussion about Renko : (from Varun's thread)

[Happy_Singh]

Sorry for long post, but could not resist myself from sharing my view guys. The second post, please bear with me (hopefully the last one).

What is trading? It is 'Dhanda'. Business. Earning is difference between Entry and Exit + Direction being in your favour. The only thing (trust me), the only thing that holds good in trading always, is S & R --> Support And resistance. If you really want to build something, try this (You can use anything, CCI, OBV, STI etc):

1. Is there a disagrement between the price & your indicator (which is supposed to trade in direction of the price)
2. Is it happening near a known point of S & R (FCB is beautiful tool, EMAs can also be used, like Varun does)
3. Is the failure point clearly identified (the place where the SL should be)
4. Most important part --> the exit conidition, the biggest problem of code based system. You can try RSI ,50 cross to exit buy etc, but it is sub optimal. This is the most critical part. If you can address this. you are done. Entering the market is not difficult, the exit is where we get the 'Pitaji ki Maafi' from Market.

I am coder (my league hates to use this word, so they calls me technical specialist) for some 14 years. And probably explored all the avenue of code based trading I could have done. So I am aware of all the drawbacks I suffered.

The job is trading is "predict the future" & Repeatedly. And among all setup I have played, only 2 concepts shows maximum repeatibility:

1. A pullback during a strong trend, turning in direction of the trend, the pullback should stop pulling back further at the known areas of S&R, you enter and enjoy the wave3 of the trend

2. Divergence - Price should not agree with something - it could be volume(OBV, WADV), strength (RSI), Momentum (Stoch), Rate of Change (ROC, CCI) or Volume + Strength (MFI)

Babu Rao: Ye Tuna. Bas kar re baba. Aur ye baki log, next time mera system mera system kiya na, to back test ka result daalo. Live trade dikha raha bolke discreationary trade ko mechnical trade mat banao. On hinds eye everything is rossy, the real result of the code is in its executions. So banao AFL aur chalo Back test. Jab dil shaant ho jaye ki bhai bahut hua tab rok na.

This is what you will find (let Tuna Predict):

1. A system with high win rate & high frequency (more trades), will have low RR (lower profit on winning trade compared to loss on the losing trade)

2. A system with high RR and High Win rate will have low frequency (matlab, you need a job buddy or huge amount to play with so that even 3-4 trade a month can take care of your monthly expense)

3. A system with high frequency and high RR will have pathetic win rate (2 out of 10 trades are winner)

Ideal system (mechnical, if anyone disputes it, dont give excuse of trades on bloggs, show backtest result, in mechnical world, we are mehnical - no middle path plz):

Code based system (which can be back tested + 0 discreation) with High win rate / High RR/ High Frequency - does not exist.

Prove me wrong - I will quit coding and sell banana in front of laalbag.

Nice Post

but why Kela, better things to sell in Lalbag


Happy

 

[timepass]

@primitivetrader , one could say the same about the other technicals

Is anyone really making money using RSI, MACD, Stochastics, ADX ?? Intraday, I mean. With minimum risk ? There are traders who will swear by simple candlestick pattern, those by the ticker alone... and there are those who are trying Renko.

 

[marimuthu13]

Sorry for long post, but could not resist myself from sharing my view guys. The second post, please bear with me (hopefully the last one).

What is trading? It is 'Dhanda'. Business. Earning is difference between Entry and Exit + Direction being in your favour. The only thing (trust me), the only thing that holds good in trading always, is S & R --> Support And resistance. If you really want to build something, try this (You can use anything, CCI, OBV, STI etc):

1. Is there a disagrement between the price & your indicator (which is supposed to trade in direction of the price)
2. Is it happening near a known point of S & R (FCB is beautiful tool, EMAs can also be used, like Varun does)
3. Is the failure point clearly identified (the place where the SL should be)
4. Most important part --> the exit conidition, the biggest problem of code based system. You can try RSI ,50 cross to exit buy etc, but it is sub optimal. This is the most critical part. If you can address this. you are done. Entering the market is not difficult, the exit is where we get the 'Pitaji ki Maafi' from Market.

I am coder (my league hates to use this word, so they calls me technical specialist) for some 14 years. And probably explored all the avenue of code based trading I could have done. So I am aware of all the drawbacks I suffered.

The job is trading is "predict the future" & Repeatedly. And among all setup I have played, only 2 concepts shows maximum repeatibility:

1. A pullback during a strong trend, turning in direction of the trend, the pullback should stop pulling back further at the known areas of S&R, you enter and enjoy the wave3 of the trend

2. Divergence - Price should not agree with something - it could be volume(OBV, WADV), strength (RSI), Momentum (Stoch), Rate of Change (ROC, CCI) or Volume + Strength (MFI)

Babu Rao: Ye Tuna. Bas kar re baba. Aur ye baki log, next time mera system mera system kiya na, to back test ka result daalo. Live trade dikha raha bolke discreationary trade ko mechnical trade mat banao. On hinds eye everything is rossy, the real result of the code is in its executions. So banao AFL aur chalo Back test. Jab dil shaant ho jaye ki bhai bahut hua tab rok na.

This is what you will find (let Tuna Predict):

1. A system with high win rate & high frequency (more trades), will have low RR (lower profit on winning trade compared to loss on the losing trade)

2. A system with high RR and High Win rate will have low frequency (matlab, you need a job buddy or huge amount to play with so that even 3-4 trade a month can take care of your monthly expense)

3. A system with high frequency and high RR will have pathetic win rate (2 out of 10 trades are winner)

Ideal system (mechnical, if anyone disputes it, dont give excuse of trades on bloggs, show backtest result, in mechnical world, we are mehnical - no middle path plz):

Code based system (which can be back tested + 0 discreation) with High win rate / High RR/ High Frequency - does not exist.

Prove me wrong - I will quit coding and sell banana in front of laalbag.

I admire for all your posts , especially on non-technical aspects..Great going bro..

 

[monkeybusiness]

you can not do that, the essense of Renco is the brick size have to be same. Now if you use static brick size, on TF (1 min, 3 min, 1 HR etc) completion, the past Renco won't repaint. But if you use a dynamic formula for calculation of the brick size, then it will repaint.

Anything other then this, you can develop a system which will show the chart with Renco like bricks, but that is anything but Renco.

Not static time based brick, brick should be of price only say bricksize is set at 5, reversal at 5, that is from Point A to point B, reversal should only happen if price moved for 25 points and then retraced back for 5 or more points to plot reversal brick. That way all plotted bricks can become static only last brick under construction will be dynamic.

 

[Tuna]

Not static time based brick, brick should be of price only say bricksize is set at 5, reversal at 5, that is from Point A to point B, reversal should only happen if price moved for 25 points and then retraced back for 5 or more points to plot reversal brick. That way all plotted bricks can become static only last brick under construction will be dynamic.

MB ,
can you backtest this for the chosen TF for max period of data you have ?

 

[monkeybusiness]

MB ,
can you backtest this for the chosen TF for max period of data you have ?

Which AFL Tunabhai, the one posted by Doss186 today?

 

[Tuna]

Test the latest updated one, dont know who posted it. Lets pick the one from Doss

 

[monkeybusiness]

Test the latest updated one, dont know who posted it. Lets pick the one from Doss

TF, Brick size, Instrument?

 

[Tuna]

TF, Brick size, Instrument?

Try Nifty (Index, as future continous data may be a problem) and BankNifty (index) for 5 years.

TF:
3 Min
15 min
1 Hour
1 Day

This should take care of both volatile and nonvolatile instruement accross low to high TF. If you have less than 5 years data, test it for the max period you have. Thanks for all the effort in advance.

 

[timepass]

Well, maybe a case of small mouth, big talk. But I have been looking at Renko trying to combine it with something else, using Kite.

There that ema(36) in blue to indicate the major trend. Then there is ema(5) in white and ma(5-time series) in yellow. Theoretically the crossover of the two 5ma gives you the trade signal, in the direction of Renko.