RSI Trading Strategies

[Gorgesmith]

Hello Friends,
Which type of RSI Trading Strategies should be use in forex market? I am beginner in trading and want to make money through best RSI strategy or technique.

 

[howardbandy]

Pick the time frame you want to use. End-of-day or intra-day. Gather historical data for the issues you think you might want to trade in time bars that fit your plans.

Pick a development platform. Beware of platforms tied to brokers. Trades that are cleared through an independent agency are much safer for the trader than over-the-counter trading with a counterparty who sees your order, then executes it.

Write the RSI trading model you want to use in the language of that development platform.

Reserve the most recent portion of the data for a one-time out-of-sample validation test.

Use the earlier data to adjust the RSI model to give the best fit. Let the data determine which model is best.

Test the system -- model plus data -- on the reserved out-of-sample data. Analyze the risk of drawdown and the profit potential. Decide whether it is safe to trade and worthwhile to trade.

 

[minwa]

I wonder how much of that our beginner friend understands

 

[tech/a]

He did say he wanted to make money using RSI in Forex.

I note you haven't offered an alternate perhaps simpler solution?
Beginners should understand it's just not that easy.

If he does what Howard suggests what do you think he'd find.

Do you think he'd be the first or last test and make an informed decision
On trading Forex with RSI if he did test it.
Do you think Howard knows what he is likely to find?

 

[Quant]

He did say he wanted to make money using RSI in Forex.

I note you haven't offered an alternate perhaps simpler solution?
Beginners should understand it's just not that easy.

If he does what Howard suggests what do you think he'd find.

Do you think he'd be the first or last test and make an informed decision
On trading Forex with RSI if he did test it.
Do you think Howard knows what he is likely to find?

All good questions and if the OP actually goes through the process defined by Howard he may actually find a few home truths about RSI and whether as a standalone indicator it has any merit .

Naturally for a beginner this is a steep learning curve to get to the skill level required to complete this task but if the OP can actually accomplish the task defined by HL he will save a lot of money and learn some skills that may lead to successful trading because evidence is a very enlightening thing . Positive expectancy leads to a positive mindset and without the tools to define expectancy you are flying blind . For those that can define expectancy will cease to trade any ' methodology ' that has negative expectancy , to do otherwise is gambling and relying 100% on luck . Successful trading is very hard work and im glad of that because i work hard .

Howard hasnt told the OP that RSI as a standalone is crap , he has just put him on a path to determine whether it is or not .

 

[Quant]

RSI strat on 1 hour AUDUSD , looks fine on first glance and maybe this is the type of thing the OP has seen but its datamined BS , a trap for new players , i actually started this test on a dataset at a high drawdown point in previous sample data

 

[minwa]

Spend good money for quality data. Learn a programming language for tens to hundreds of hours. Raise 7 figures for prime brokerage to avoid OTC. Rest I don't understand.

Let's wait for OP to see how many of those he will look to accomplish.

edit: wrong quote

 

[Gringotts Bank]

RSI strat on 1 hour AUDUSD , looks fine on first glance and maybe this is the type of thing the OP has seen but its datamined BS , a trap for new players , i actually started this test on a dataset at a high drawdown point in previous sample data View attachment 70739

That looks like a strat from ProRealCode. I like the guys on that forum, but one needs to be careful about validity because many strats aren't walk forward tested, so they fall apart in real conditions.

OP, your question sounds simple enough but it's not at all. There's no easy answer for you. Oscillators oscillate. If you want to try a discretionary approach, buy when the oscillator is low, and sell when it's high. If you want to trade a system, you'll need to either buy one or create one.

 

[Quant]

That looks like a strat from ProRealCode. I like the guys on that forum, but one needs to be careful about validity because many strats aren't walk forward tested, so they fall apart in real conditions.

You obviously didnt get my point , its curve fitted rubbish , its has little validity going forwards , its basically buying RSI14 at 30 and shorting at 70 . im saying it will fall apart in real world at some stage , to go a step further i optimized it to the max and got something i could probably sell online for 1000 bucks ( i wont ) . 990% in 4 years and im sure i could literally sell 100's of it

 

[howardbandy]

In-sample backtest results always look good. We do not stop data mining and adjusting rules and parameters until they do.

It is possible that the rules do identify patterns that precede profitable trades and persist beyond the in-sample period. There is no way to tell from the in-sample results. Only independent tests performed on data that follows the in-sample period can give an estimate.