# Sounding Board and Random Brainstorming

#### arsh22g

##### Well-Known Member
Just starting a separate thread to bounce off ideas. All ideas pertaining to trading welcome, and a little bit of others too.

#### ankur07090

##### Banned
I have one idea.
I need any AFL (MA crossover, Stoch, rsi....) which will give 70-80% loss (more is better).

Now I will revert all the signals generated by that AFL.

#### arsh22g

##### Well-Known Member
I have one idea.
I need any AFL (MA crossover, Stoch, rsi....) which will give 70-80% loss (more is better).

Now I will revert all the signals generated by that AFL.
Ha ha...I tried it once. You will find such at lower TFs, but # of transactions will be high. Reversing profit may get washed out by transaction cost.

#### vijkris

##### Learner and Follower
I have one idea.
I need any AFL (MA crossover, Stoch, rsi....) which will give 70-80% loss (more is better).

Now I will revert all the signals generated by that AFL.
oh great... this principle is already used in my afl posted in lonewolf thread. I thought I m a crazy fellow but it worked for me...
:clap:
thanks for the company

#### extremist

##### Well-Known Member
in regards of ur code in the above post i would like to suggest few things.

as Vijaykris spotted coding flaws in it code might be working but for better results just try following.

1.> as i see in ur code ur watching 3 min timeframe for trading. so to get better results of hourly averages just multiply the average periods by the TF factor i.e. 20 here (60/3 = 20)
and to get Hrly averages plotted on 3M tf use 260 ema instead of 13 and 1280 instead of 64 ema.
to get the most accurate result for any EMA or SMA of higher TF multiply the period by TF factor.
it works better coz the ema plotted for Higher TF will be based on the candle closed in lower tf so as the candle is closed, it bears better accuracy.

2.> And while backtesting such strategy back test it on NF spot data coz it will not have Expiry - New series gaps in it

#### arsh22g

##### Well-Known Member
in regards of ur code in the above post i would like to suggest few things.

as Vijaykris spotted coding flaws in it code might be working but for better results just try following.

1.> as i see in ur code ur watching 3 min timeframe for trading. so to get better results of hourly averages just multiply the average periods by the TF factor i.e. 20 here (60/3 = 20)
and to get Hrly averages plotted on 3M tf use 260 ema instead of 13 and 1280 instead of 64 ema.
to get the most accurate result for any EMA or SMA of higher TF multiply the period by TF factor.
it works better coz the ema plotted for Higher TF will be based on the candle closed in lower tf so as the candle is closed, it bears better accuracy.
Yup, tried this a little unclean method too. :thumb:

#### arsh22g

##### Well-Known Member
but at least u will get more relevant out put in backtesting reports.

coz there will be no interference of unfinished candle of HTF and expandfirst or last issue.
I discovered 124, 1240 crossover on 1 -min using this method. It gives very good swing trades in NF.

780, 6240 crossover on 1-min gives good swing trades in Crude.

#### Tavnaz

##### Well-Known Member
Ok
I have an idea of session based trading for non 24 hour market.
I havn't been able to test it,maybe you guys already doing it
Here it goes so basically we break the day into sections of volatility and non volatility.
So during volatile sessions we assume a trend would form and not volatile session would form range which can be resolved either way up or down when next volatile session comes.
So we break down the trading period into chunks of these volatility and non volatility and classify the day into types.
The time of volatile periods depends on global market timings couple them up with the time in the market you trade.
So we could either have full stack bull/bear day.
Or we could have a indecisive day turning during second session.
Or we could have a range day based on first session trading inside previous session.
So basic idea is your same gauging market move relative to previous day but in this we are gauging market moves as boxes on top of each other or trading inside each other.
For finding time of range we draw volatility map.based on news for the scrip or maybe a specific time when our trading instrument always acts odd in.
I am talking about market rythym at exact time on exact day.
Because it's basically it's institutional traders placing orders as a part of their job clearing their order book or getting order fill. At the time when there is some particular time at home or abroad.
It's a basic idea at this moment.
It can become a trading method I know cause I use it somewhat.

#### extremist

##### Well-Known Member
I discovered 124, 1240 crossover on 1 -min using this method. It gives very good swing trades in NF.

780, 6240 crossover on 1-min gives good swing trades in Crude.
i found 375 and 1125 good for NF
there r lot different combos.