Choosing an instrument and strategy

[JaybeeTrading]

Hi all, been creeping up to the markets a very long time, and want to get the ball rolling finally.

Spent a lot of time studying why the majority fail (no concretized strategy, then trading greedy then money mismanagement then losing then scared then bust) - and am hoping to be the Broker's nightmare, that 1-2% who did their backtesting and paper trading on one platform and transitioned into trading emotionlessly and accept inevitable (but still minimised) losses with the same cool as the profits.

There are a couple of quite basic things that still flummox me...as above, I'm MORE than sold on the idea of backtesting, but so far, all the vids/articles tell us you need to decide your strategy THEN backtest/forward test, but wait...how do you build the strategy in the first place?

You look at historical data, plot S+R lines (and imo not too much else, not to start a debate but I like the idea of the SLA, as it follows the rule of avoiding conflicting indicators), and - unless I'm very wrong, and please please tell me if I am - you go back in time, looking for movements that give you your desired number of pips/points, and THEN begin the process of creating a strategy by determining what (if any) common factors existed, when you should have gotten in and out, and going forward to see if your strategy has a favourable ratio of win/loss? What then, is the difference between "Backtesting", and creating a strategy via TA?

Secondly, picking a market/instrument - I want a high level of bounce in my market, with frequent setups - conceivably it will takes days, weeks to backtest the data, and that time is an investment I don't want to squander on a relatively flat market.

Thanks all,

JB

 

[dbphoenix]

If you're interested in trading price via the SLA, "How To Do It", attached to this post, may be helpful to you. If you're interested in backtesting beyond or outside of the SLA, "Developing A Plan", attached to the same post, will be helpful.

Db

 

[J_C_Anderson]

The main difference between backtesting and strategy creation is that to create strategy, you need just an idea, and you need strategy to backtest it.
For example:
- you had spotted something interesting on the markets
- you create the concept of the strategy, its main rules
- after this you can start backtesting to find out whether this idea profitable or not.
Backtesting is also necessary to check adjustments and improvments.Using special backtesting software you can test your ideas and than make your strategies better, backtesting the changes in each detail, changing them one by one

 

[JaybeeTrading]

The main difference between backtesting and strategy creation is that to create strategy, you need just an idea, and you need strategy to backtest it.
For example:
- you had spotted something interesting on the markets

Let me stop you there; the ONLY thing interesting in the markets are sharp movements (unless you like your money doing nothing), and to find those, you have to go back over historical data, and see if there are any repeating precedents, then go forward and see whether those conditions indeed repeat - so where is the difference?

 

[malaguti]

Let me stop you there; the ONLY thing interesting in the markets are sharp movements (unless you like your money doing nothing), and to find those, you have to go back over historical data, and see if there are any repeating precedents, then go forward and see whether those conditions indeed repeat - so where is the difference?

Your sharp movement is only an idea, as you go back and look at those “precedents” you are potentially identifying your entry criteria.
You’re defining what a “sharp movement” actually means, can this be repeated
It’s precedents as you say, are these repeatable. The combination of which might define your entry, possibly your exit in the other direction
You haven’t created a strategy yet as it takes more than just the entry and likewise you haven’t really backtested anything all you have done is come up with a potential entry/exit

Is that sharp movement noticeable on any particular timeframe, you may notice it hardly registers on a weekly timeframe for example, so now your potential strategy has a timeframe about it if it’s any good you’re beginning to notice it’s true across different timeframes adding now an element of robustness to this strategy you are just beginning to build
What other conditions exist that may be prevalent, such as a discernible trend that could also help shape this strategy that you could also test
Hopefully you’re beginning to see the difference between an initial idea and the building blocks of an overall strategy which you can then truly backtest to see if the combination of all of these ideas is actually going to make you any money, holds up against time, different market states, different instruments
Hope that makes sense and feel free to stop me right there🤚

 

[JaybeeTrading]

Your sharp movement is only an idea, as you go back and look at those “precedents” you are potentially identifying your entry criteria.
You’re defining what a “sharp movement” actually means, can this be repeated
It’s precedents as you say, are these repeatable. The combination of which might define your entry, possibly your exit in the other direction
You haven’t created a strategy yet as it takes more than just the entry and likewise you haven’t really backtested anything all you have done is come up with a potential entry/exit

Is that sharp movement noticeable on any particular timeframe, you may notice it hardly registers on a weekly timeframe for example, so now your potential strategy has a timeframe about it if it’s any good you’re beginning to notice it’s true across different timeframes adding now an element of robustness to this strategy you are just beginning to build
What other conditions exist that may be prevalent, such as a discernible trend that could also help shape this strategy that you could also test
Hopefully you’re beginning to see the difference between an initial idea and the building blocks of an overall strategy which you can then truly backtest to see if the combination of all of these ideas is actually going to make you any money, holds up against time, different market states, different instruments
Hope that makes sense and feel free to stop me right there��

Thanks for replying, but let me see if a rephrase will help;

You start out with thinking, "I want to make money", so, Phase 1: you look at a few flat markets, some wildly oscillating ones, you finally pick one that has lots of lovely neat and steep trendlines (and if anyone knows one, I'm all eyes!! ) then think, "That's the baby". Here's what I'm unsure about; at this point, to my way of thinking, the smart man moves to Phase 2, so starts drawing his S+R lines, trendlines, THEN sees repeated opportunities, builds a list of patterns seen, trigger points, and creates Entry/Exit rules (ie a strategy) for each pattern, dependant on his budget/timescale, then Phase 3, he moves forward into the next and presumably most recent block of time to see if those strategies/patterns/rules hold true (or hold true enough after refinement) to give him strong positive expectancy he can start paper trading them and move to full-money trading, naturally only after a month or two of seeing his expectations fulfilled.

That's MY perception; from research I've been doing, if we move back to Phase 2, you trade just ONE pattern - that could be a flag, could be a reversal, could be head+shoulders etc and then move to Phase 3….but wait, why not learn multiple pattern recognition and have all the more trading setups (so long as they give positive expectancy) ? Or is that generally considered overload?

Also - if the term for Phase 3 is called 'Backtesting', is there a common term for Phase 2 ?

 

[NVP]

i think what JT is trying to explain regarding his approach is similar to the famous quote below :-

The famous criminal Willie Sutton was once asked why he robbed banks, and his response was simple, eloquent, and humorous:

Because that’s where the money is.

JT is chasing volatility / momentum

 

[Bloodhound]

Let me stop you there; the ONLY thing interesting in the markets are sharp movements (unless you like your money doing nothing), and to find those, you have to go back over historical data, and see if there are any repeating precedents, then go forward and see whether those conditions indeed repeat - so where is the difference?

If you are interested in trading 'sharp movements' then you are very likely heading down the right path if you are looking to create an 'efficient' trading strategy - i.e. one where you can make as much as possible as quickly as possible. Take a look around the internet and see what you can learn about 'compression', especially in the context of supply and demand consumption. There is a reason why those sharp movements occur and they are predictable in their nature.

 

[dbphoenix]

Thanks for replying, but let me see if a rephrase will help;

You start out with thinking, "I want to make money", so, Phase 1: you look at a few flat markets, some wildly oscillating ones, you finally pick one that has lots of lovely neat and steep trendlines (and if anyone knows one, I'm all eyes!! ) then think, "That's the baby". Here's what I'm unsure about; at this point, to my way of thinking, the smart man moves to Phase 2, so starts drawing his S+R lines, trendlines, THEN sees repeated opportunities, builds a list of patterns seen, trigger points, and creates Entry/Exit rules (ie a strategy) for each pattern, dependant on his budget/timescale, then Phase 3, he moves forward into the next and presumably most recent block of time to see if those strategies/patterns/rules hold true (or hold true enough after refinement) to give him strong positive expectancy he can start paper trading them and move to full-money trading, naturally only after a month or two of seeing his expectations fulfilled.

That's MY perception; from research I've been doing, if we move back to Phase 2, you trade just ONE pattern - that could be a flag, could be a reversal, could be head+shoulders etc and then move to Phase 3….but wait, why not learn multiple pattern recognition and have all the more trading setups (so long as they give positive expectancy) ? Or is that generally considered overload?

Also - if the term for Phase 3 is called 'Backtesting', is there a common term for Phase 2 ?

As I pointed out a month ago in my reference to the pdf on "developing a plan", your "Phase 3" is forward-testing. As for "multiple pattern recognition", there is only one: the shift from buying pressure to selling pressure (and vice-versa). Everything else is only a variation of this fundamental dynamic.

Db

 

[NVP]

everyday i hunt momentum in the currency markets .......i look for the currencies that are displaying the most action and get on them ......

good luck JT