Attachments: Diversified Trend Trading Approach

Diversified Trend Trading Approach

#4,301
Aug 17, 2019 10:44pm Aug 17, 2019 10:44pm

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Time to chill by the pool :-)

Inserted Video

#4,302
Edited Aug 18, 2019 12:15am Aug 17, 2019 11:58pm | Edited Aug 18, 2019 12:15am

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Don't you just luv trading. Wouldn't you like to do this your whole life.....and extend that lifespan to enjoy it?

Thanks to our understanding of epigenetics, perhaps we are not so far away from this at all.

A little Intro:

Inserted Video

....and more lengthy discussions with David Sinclair.

Inserted Video

#4,303
Aug 19, 2019 3:09am Aug 19, 2019 3:09am

guido9
| Joined Oct 2016 | Status: Trader | 21 Posts

Hi Copernicus

I really enjoy reading your posts on my sporadic visits to forexfactory, and hope to be able to spend more time really going through some of these threads. I am at the stage where I really want to do my own thorough, sophisticated backtesting of some strategies and ideas, and was wondering if you could please advise me on the best toolset? I know you have lots of experience in this area.

I seem to remember in the past you mentioning Birt's Tick Data Suite which I am tempted by (I'd probably go for the Lifetime package as this seems the best value). Is this something you recommend?

I have purchased EA Wizard & QuantAnalyzer (I don't yet have SQ &/or EA Studio/Forex Strategy Builder because I want to test my own theories first rather than something generating EAs for me, although I am also interested in this, and am open to buying these too). I am not so worried about cost, and am happy to purchase multiple tools, as well as purchasing data if necessary.

Thus far I have done a bit of backtesting, using the relatively small amount of data that my current broker provides, and using MT4 Strategy Tester, testing some EAs and custom indicators. I have also exported OHLC and done some quite complex excel backtests (which has its limits but is good for purposes of understanding how indicators really work). But I need much more price data, and I can't wait to get stuck in to doing more sophisticated backtesting using best practices. I have quite an analytical mindset like yourself, but am much less far down the line than you.

Regarding TDS, I don't think I will be getting into scalping such as the TP & SL will be within the same M1 bar so I am uncertain whether I would actually benefit at all from tick data, but I would like to get as good data as possible going back as far as possible, and perhaps the other tools like slippage simulation, variable spreads etc make it worthwhile.

Very grateful for any advice you can give on optimum tools.

#4,304
Edited 9:20am Aug 19, 2019 8:08am | Edited 9:20am

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Quoting guido9

Disliked

Very grateful for any advice you can give on optimum tools.

Ignored

Hi G....thanks for popping in.

Mate....if you are not a fan of data mining....(which can be a tricky process as it's too easy generating curve fit results)...then the tools I would recommend for Forex are as follows:

1. If you are a discretionary trader and would like to backtest, then soft4fx provides a really great solution that uses your broker's data or Dukascopy tick data or TrueFX tick data. Apart from the initial cost (very reasonable for what is offered) then there is no additional costs for the data updates.

2. If you are fully systematic and like programming within MT4 (or via EA Wizard or Forex Strategy Builder) then backtesting on MT4 is fine...but fairly slow. I actually spend most of my time with this however (though I have a coding buddy that works with me on this for MQL programming which allows a bit more flexibility than the EA development tools for non-coders...however FXDreema is also a pretty good shortcut).
If you go down any of these paths then TDS is a great companion to these methods. Ultimately you need to test your strategies on your brokers data set so while Forex Strategy Builder et al are nice....there really is no better substitute than MT4 strategy tester itself. It includes current SWAP applied to your history and Commissions from your broker sources in the test and all you need to factor in is slippage.
Having data from multiple broker sources such as your prime broker and Dukascopy for data validation is highly recommended whether or not you intend to become a scalper. Tick by tick is a waste of time if you code off the open condition or close condition...so Dukascopy is preferred simply for the higher quality data it provides over the long term and for comparison with your brokers data feed. There is some variation in data between brokers, so one way to eliminate this for standardisation purposes is confirm your results across multiple sources.
Once TDS is installed, it sits in the background to MT4 Strategy Tester and can be activated when you want Dukascopy tick data or not activated when you operate off your brokers data. There is a monthly fee for TDS....but very much worth it in my opinion as it not only gives tick data feeds from Dukascopy, but many other broker sources as well such as Darwinex etc.

3. An additional companion piece to point 2) is Quant Analyzer which you have already. This saves a hell of a lot of time in performance analysis and imports trade results directly your MT4 Strategy Tester report into the program for portfolio compilation purposes. You do need to factor in slippage however if you trade the short term timeframes.....H4 and D1 and above is less of a problem with slippage.

That is really all you need mate to give you a great head start into the backtesting world....however you may also want to develop your own tools as well to add additional bells and whistles to this process....but that can come later.

There are great backtesting programs such as Trading Blox, Amibroker, Ninja Trader etc....for a broad range of other asset classes such as futures, equities etc.....but I am not a big fan of the off-the shelf solution as you need to commit to these and have less flexibility in your options as it is less clear to understand what is under the hood. The tools I have outlined give you broad capability in backtesting through any source that offers MT4...such as CFD/ETF providers and forex brokers which can give you broad coverage across asset classes.

I hope this helps mate.

Cheers

C

#4,305
Aug 19, 2019 9:27am Aug 19, 2019 9:27am

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Inserted Video

#4,306
Aug 19, 2019 11:25am Aug 19, 2019 11:25am

guido9
| Joined Oct 2016 | Status: Trader | 21 Posts

Quoting Copernicus

Disliked

{quote} Hi G....thanks for popping in. Mate.... I hope this helps mate. Cheers C

Ignored

Dude, thank you soooo much, that's a great help! Really appreciate your time there! Helps crystallise my thinking. I am interested in the data mining side of things but that is on my Someday list; first I want to hone my skills in the other discretionary and systematic approaches. I also have reliable coding partners who can plug gaps in what I can't achieve myself. Excited to take what I've done to the next level now...

Cheers

#4,307
Aug 20, 2019 7:45am Aug 20, 2019 7:45am

21vs7
Joined Dec 2010 | Status: Foook Bollinger-dr.Kegel knows! | 9,794 Posts

Quoting Copernicus

Disliked

Donchian Suite Tests continued.... Refer to system definition in Post 640 Strategy: D3T Type: Medium Term Trend Trader Timeframe: D1 Test Period: 1/1/2000 - 31/12/2015 Donchian Entry: 100 Period Donchian Exit: 50 Period Initial Stop: 50 Period ATR x 3 Note that all unrealised profit taken at the end of the test hence the significant profit taken on final bar of test. Note that when comparing against other methods, drawdowns usually are experienced upon portfolio commencement...

Ignored

Hey C,

Im reading your journey from the start. Love your approach. It really pinpoints your general personality i would presume

great orderly fella!

Wife must be happy lol..
Really good stuff

Ill go direct into

Have some general questions...this quoted test for 15 years and with 0,42R per trade in the portfolio. Do you trade that live going forward? What are results from the past 4 years with same strategy and portfolio settings?

From start.. you started with one strategy, then for testing purposes you extended to lots of strategies.. There are some D1vsH1.. Wait for 2,5R to move from 1,0 to 0,5 trailing lenght (primary strategy) which in your earlier stats you showed ROI for years 2013 and 2014 of 70 or 80%. Was this valid? What are the results for the past 4 years on this.?
I see Youve changed a bit on presenting the stats and i dont know if i understand it well. Im on 33 page right now

#4,308
Aug 20, 2019 3:26pm Aug 20, 2019 3:26pm

21vs7
Joined Dec 2010 | Status: Foook Bollinger-dr.Kegel knows! | 9,794 Posts

Quoting Copernicus

Disliked

{quote} When my war chest of unrealised profits starts getting low (reducing my margin of safety buffer) or when my drawdown starts building I start to further stretch the placement of my pending entries away from price. Clearly market conditions are unfavourable and the reasons for these symptoms appearing. So by placing my pending orders further away from current price I am reducing trade frequency during unfavorable market conditions (avoiding over-trading). This technique defends my capital base and allows me to manage drawdowns with more discretion...

Ignored

I dont understand this in rule based world actually

how would you define when entries should be moved as we do not know when the next big trade is gonna occur. Its right next one perhaps and if one misses one, we know the big ones occur rarely. How one puta this decision in the automated system?

"When my war chest of unrealised profits starts getting low (reducing my margin of safety buffer) or when my drawdown starts building I start to further stretch the placement of my pending entries away from price. Clearly market conditions are unfavourable and the reasons for these symptoms appearing. So by placing my pending orders further away from current price I am reducing trade frequency during unfavorable market conditions"

#4,309
Aug 20, 2019 6:23pm Aug 20, 2019 6:23pm

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Quoting 21vs7

Disliked

{quote} Hey C, Im reading your journey from the start. Love your approach. It really pinpoints your general personality i would presume great orderly fella! Wife must be happy lol.. Really good stuff Ill go direct into

Ignored

I pity you mate. This is more a journal of a desperate guy looking for a way to defeat discretion by systematic application.....but there are lots of lessons that I can take from this when I go back in time to reflect on the journey. You can also see the friends I made along the way including yourself that helped to shape my approach. That is why IMO a journal is so important for a trader. Not necessarily a journal in the strict sense of the word...but a diary of the ideas that shaped the 'now' and where that moving now is today.

Quoting 21vs7

Disliked

{quote} Have some general questions...this quoted test for 15 years and with 0,42R per trade in the portfolio. Do you trade that live going forward? What are results from the past 4 years with same strategy and portfolio settings? From start.. you started with one strategy, then for testing purposes you extended to lots of strategies.. There are some D1vsH1.. Wait for 2,5R to move from 1,0 to 0,5 trailing lenght (primary strategy) which in your earlier stats you showed ROI for years 2013 and 2014 of 70 or 80%. Was this valid? What are the results...

Ignored

Yep mate. The strategy took the form of a moving feast where it was reworked along the way. Sorry about not expressing it concisely enough....but it still stacks up. I don't trade it anymore but I do still track it in backtest land. It is a great system mate....but my ideas have now been sold on a new way forward.....a multi-model approach which I do through diversification and portfolio management.

It still could be used as an element of a broader range of systems....but given it's discretion, there is still too much subjectivity in it to keep me happy.

Here it is in all it's glory now.......I used it for demonstrating how to compile a portfolio. So....still works admirably.

#4,310
Aug 20, 2019 6:26pm Aug 20, 2019 6:26pm

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Quoting 21vs7

Disliked

{quote} I dont understand this in rule based world actually how would you define when entries should be moved as we do not know when the next big trade is gonna occur. Its right next one perhaps and if one misses one, we know the big ones occur rarely. How one puta this decision in the automated system? "When my war chest of unrealised profits starts getting low (reducing my margin of safety buffer) or when my drawdown starts building I start to further stretch the placement of my pending entries away from price. Clearly market...

Ignored

Yep mate...that was a bad discretionary rule that made sense at the time which has been ditched. You will pick lot's of these discretionary assumptions along the way that seemed smart at the time....but had no validity. A central reason for why discretion does not cut it in the long term. This brain of mine took too much of a lead in decision making back in those days....where I should have been listening to the data.

Cheers J :-)

#4,311
Edited 1:08am Aug 21, 2019 12:39am | Edited 1:08am

21vs7
Joined Dec 2010 | Status: Foook Bollinger-dr.Kegel knows! | 9,794 Posts

Quoting Copernicus

Disliked

{quote} I pity you mate. This is more a journal of a desperate guy looking for a way to defeat discretion by systematic application.....but there are lots of lessons that I can take from this when I go back in time to reflect on the journey. You can also see the friends I made along the way including yourself that helped to shape my approach. That is why IMO a journal is so important for a trader. Not necessarily a journal in the strict sense of the word...but a diary of the ideas that shaped the 'now' and where that moving now is today. {quote}...

Ignored

Yes. Its very nice to go through your journal. Lots of value there. Interesting also to see the process of thinking years back, to now.

MAR around 2 is of course top notch. Great achievement!

If I remember correctly, you have 35% win rate on the portfolio. If you have a info, what is your longest loosing streak in trade count and what is your average R per all trades in the portfolio? Would like to hear what is you maximum R your achived in one trade(one strategy one entry) just to get an idea what a beast are you riding Sir.

#4,312
Edited 2:24am Aug 21, 2019 2:07am | Edited 2:24am

Copernicus
| Commercial User | Joined Apr 2013 | 4,366 Posts

Quoting 21vs7

Disliked

{quote} Yes. Its very nice to go through your journal. Lots of value there. Interesting also to see the process of thinking years back, to now. MAR around 2 is of course top notch. Great achievement! If I remember correctly, you have 35% win rate on the portfolio. If you have a info, what is your longest loosing streak in trade count and what is your average R per all trades in the portfolio? Would like to hear what is you maximum R your achived in one trade(one strategy one entry) just to get an idea what a beast are you riding Sir.

Ignored

J...attached is a summary of major stats for the latest incarnation which was about as systematic as I could make it. The better performers in terms of MAR are highlighted.

Attached Image (click to enlarge)

Click to Enlarge

Name: EDTT M30.PNG
Size: 68 KB

Average Loss across the portfolio was ($260.30) on a $50K balance....so let's say R is ($260.30)

Max win was $9842.23/260.30 = 37.8 R. Went for a huge ride on XAUUSD back in 2011 when it went parabolic.

MAR came in at 1.77 and Pwin% at 42.0%

Average R for all winning trades was 638.44/260.3 was 2.45

Longest losing streak in trades was 14. Max winning streak was 10.

It was a great little beast mate....but more a momentum chaser than a trend trader.....but we did have a max hold of a trade for 165 days...but unusual times for that one (Brent). :-)

PS What makes me think that you are ingeniously crunching out a super strat that is going to smash this as a benchmark.*sniggers*. :-)

#4,313
Aug 21, 2019 5:35am Aug 21, 2019 5:35am

21vs7
Joined Dec 2010 | Status: Foook Bollinger-dr.Kegel knows! | 9,794 Posts

Quoting Copernicus

Disliked

{quote} J...attached is a summary of major stats for the latest incarnation which was about as systematic as I could make it. The better performers in terms of MAR are highlighted. {image} Average Loss across the portfolio was ($260.30) on a $50K balance....so let's say R is ($260.30) Max win was $9842.23/260.30 = 37.8 R. Went for a huge ride on XAUUSD back in 2011 when it went parabolic. MAR came in at 1.77 and Pwin% at 42.0% Average R for all winning trades was 638.44/260.3 was 2.45 Longest losing streak in trades was 14. Max winning streak was...

Ignored

Thanks C,
Really great results here. You should be proud.

Thanks for all the answers. I like those nimbers to see what you are dealing with.. Cons loses, wins,.. Average win, loss.. Average win on R and the outlayers. Good to see this numbers to try to feel what system nature looks like.

Privilige to see such a stable performance for that specific lenght of a fund. Remarkable. Probably lots of systematic traders out there (me included) who only watch and read this thread with hopes to achieve something like that going forward. This thread is really a golden apple to chew on.

Keep on!
PS, this time I wont pick a song for you, so probably just stay with the one which plays the good tune!