Reverse Engineering a Profitable System (REPS) | Page 13 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Very Profitable System - Roi system 355 replies

Reverse engineering order flow from naked chart 18 replies

MA, Gann and Ichimoku system (profitable system!) 60 replies

Fractal Reverse Engineering Journal 21 replies

METATRADER REVERSE ENGINEERING 3 replies

Attachments: Reverse Engineering a Profitable System (REPS)
Exit Attachments

Reverse Engineering a Profitable System (REPS)

  • #241
  • Sep 4, 2015 11:02am Sep 4, 2015 11:02am
awesome!

Attached Image (click to enlarge)
Click to Enlarge Name: Captura de pantalla de 2015-09-04 17:56:14.png Size: 79 KB


congrats!!
MY telegram channel: https://t.me/https://t.me/tnavieurusd
  • #242
  • Sep 4, 2015 12:09pm Sep 4, 2015 12:09pm
I am following this thread with great interest. I do not want to diminish the efforts, and it is difficult to disagree with the result: It looks lake a low-risk, low-maintenance and high gain method, which is a goal outlined in the first post.
With all this being said, I am sorry, I just do not see an edge here. OK, Donchian channel prevents stupid entries, as the buy and sell "hemispheres" are well defined and distinct. Still, there is little evidence that the long-period Donchian channel breakout implies a long trend. (Obviously vice versa is true.) However, if it is true, it should hold universally. Thus for example an analogous method should work on a short timeframe as well. I doubt that it does though. Something doesn't add up...
For example Graviton's method does an entry which is consistent with the fundamental/macro expectation. This is an edge.
All these "millipede" methods rely on the long tails of the return distribution - trends.
Is the trend such an edge that everything else is irellevant? Such that it is sufficient to trail the stops to breakeven (cut the losses), and, voilą, the rabbit is out of the hat, a consistent beating of the market is achieved?

k
  • #243
  • Sep 4, 2015 4:39pm Sep 4, 2015 4:39pm
Quoting howard
{quote} Hello spyderman Any reason for not taking the break of daily DC at 0.7047? 2 steps above the level you entered. Also is the DC on your 4hr chart a 120 period serving as daily?

Yes, the reason is that at that point price had not moved far enough from our initial entry to lock in profits (using our H4 s/l) to fund a new entry. Had to wait till price (and more importantly the H4 DC trailing stop) got further away.

I have a MTF DC indicator. The bluish is H4, the gold is Daily and the aqua is Weekly.
  • #244
  • Sep 4, 2015 4:43pm Sep 4, 2015 4:43pm
Quoting silverpuma
Things progressing well spyderman, still plenty of green How are you finding the emotional swings as the overall portfolio ebbs and flows with the significant number of live positions open. I'm especially interested in the addons trades and how you find these? What has been your biggest struggle (if any) in this area of managing your feelings and the urge sometimes to bend the rules. Keep up the good work and have a great weekend
So far so good. Not really dealing with large numbers yet...
I have confidence in the basis of the system. I know the floating is going to move around a lot. I take solace in the fact that with the addon trades I have profit locked in, even if they reverse. I just like to get to BE and then let the market decide.
Will any of these positions go on to be a major trend...just no way of knowing, but eventually we'll hit some.

You have a good weekend too.
  • #245
  • Sep 4, 2015 4:44pm Sep 4, 2015 4:44pm
Quoting tnavi
awesome! {image} congrats!!

Thanks
  • #246
  • Sep 4, 2015 4:55pm Sep 4, 2015 4:55pm
Quoting kprsa
I am following this thread with great interest. I do not want to diminish the efforts, and it is difficult to disagree with the result: It looks lake a low-risk, low-maintenance and high gain method, which is a goal outlined in the first post. With all this being said, I am sorry, I just do not see an edge here. OK, Donchian channel prevents stupid entries, as the buy and sell "hemispheres" are well defined and distinct. Still, there is little evidence that the long-period Donchian channel breakout implies a long trend. (Obviously vice versa is...

K...appreciate your thoughts.
I believe that the trend is the edge as you put it. I certainly claim no special edge in the entries, stops, etc. etc.
I do know that the parameters I've set up with the help of others here, would have traded recent trades (most notably the E/U template) very profitably. I don't feel that's an aberration, I'm confident it will happen again...repeatedly
I do think the MM we're employing is an edge of sorts but that same edge could be applied to any trading system.

I read a good book by Covel on Trend Following where he compared a simple moving average cross (which I've used often) over the last 25 years or so in a variety of markets. It out performed fund performance by a good margin.

The trick is to be consistent and limit risk so that you will be in place when the trend wave comes along. I did well with the Yen collapse trends back in late 2012 early 2013, but I didn't stay in it for the long haul (my MM at the time was not what it should have been). I'm sticking with this till it pans out or I can't afford it any more.
  • #247
  • Sep 5, 2015 5:50am Sep 5, 2015 5:50am
Quoting spyderman
{quote} K...appreciate your thoughts. I believe that the trend is the edge as you put it. I certainly claim no special edge in the entries, stops, etc. etc. I do know that the parameters I've set up with the help of others here, would have traded recent trades (most notably the E/U template) very profitably. I don't feel that's an aberration, I'm confident it will happen again...repeatedly I do think the MM we're employing is an edge of sorts but that same edge could be applied to any trading system. I read a good book by Covel on Trend Following...
OK, I hope you catch some good trends!
Cheers,
k
  • #248
  • Edited 8:50am Sep 5, 2015 7:12am | Edited 8:50am
How do you get:

Break of the 20 period Donchian Channel on the Weekly chart

So if the weekly is going up and hits the 20 period line at the top channel, it is the start of the trend? Meaning it will continue to go up?

So according to the picture when it breaks the 20, it should continue up.

Just trying to see the correct setup. Thanks mate!!!!
Attached Image
  • #249
  • Sep 5, 2015 9:16am Sep 5, 2015 9:16am
Quoting Gdi
How do you get: Break of the 20 period Donchian Channel on the Weekly chart So if the weekly is going up and hits the 20 period line at the top channel, it is the start of the trend? Meaning it will continue to go up? So according to the picture when it breaks the 20, it should continue up. Just trying to see the correct setup. Thanks mate!!!! {image}

That's the idea, yes. Basically the Donchian channel is just the high or low for the last (n) periods.
So the idea is price has just reached a new high for the last 20 weeks. That may well signify long term momentum in that direction...
Then again...it may not.
  • #250
  • Sep 5, 2015 11:14am Sep 5, 2015 11:14am
Quoting spyderman
{quote} That's the idea, yes. Basically the Donchian channel is just the high or low for the last (n) periods. So the idea is price has just reached a new high for the last 20 weeks. That may well signify long term momentum in that direction... Then again...it may not.
Thanks. Hope Aunt May is doing good.
  • #251
  • Sep 9, 2015 7:43am Sep 9, 2015 7:43am
So we've had several positions closed, some at BE some at partial loss.
I'm going to concentrate on the addons. You can refer to the TE for others if you want.

Following addons closed.

Aud/Usd - Nzd/Jpy - Eur/Aud all for a partial loss.
None of the original entries we're closed so obviously the opposite daily DC was not hit, so series is still open.
I've placed new pending orders on these using the 50% rule (decreasing the amount risked by the amount lost in the addon closure) so we'll see if PA swings back in our favor.
  • #252
  • Sep 9, 2015 8:04am Sep 9, 2015 8:04am
Nzd/Usd both entries closed so we're back to square 1 on that pair. 50% profit banked.
  • #253
  • Sep 9, 2015 9:38pm Sep 9, 2015 9:38pm
New 2nd entry addon position added to E/A.
Reduced risk amount by what was lost on the partial lost on the previous 2nd entry addon that closed.

Attached Image (click to enlarge)
Click to Enlarge Name: EURAUDiH4.png Size: 73 KB
  • #254
  • Sep 10, 2015 6:28am Sep 10, 2015 6:28am
Quoting spyderman
New 2nd entry addon position added to E/A. Reduced risk amount by what was lost on the partial lost on the previous 2nd entry addon that closed. {image}

Both entries closed. 50% banked
  • #255
  • Sep 13, 2015 7:07pm Sep 13, 2015 7:07pm
Hi SPyderman,

May i ask a few questions? Thanks!

1 Trade management
Trail s/l on the H4 DC until BE

How to trail?
For the first initial entry set, then when to set it to BE?How about the add-on dialy-entry/H4-aggressive-entry?

2 Money management

Does it mean there would be always just one position in risk?
For initial entry, it would be the .25%~2% range
For add-on entry, it would be count as the 50% profit already locked in.?
Outcome is not important at all, but correct technique is what only matters
  • #256
  • Sep 13, 2015 7:55pm Sep 13, 2015 7:55pm
Hey SZ

Quoting szpowerx
How to trail? For the first initial entry set, then when to set it to BE?How about the add-on dialy-entry/H4-aggressive-entry?
For both...once the entry s/l is set on the H4 Donchian, you trail the stop along that DC channel until your trade is at BE. So it's possible that every 4 hrs ( as the channel updates) you will move your s/l closer to your entry. Once it reaches BE then leave it there until you have to lock in more profit (for an another addon) or the Daily DC/Ichimoku cloud catches up. Now it may be that the H4 DC doesn't move every 4 hr segment depending on Price Action.

Quoting szpowerx
Does it mean there would be always just one position in risk? For initial entry, it would be the .25%~2% range For add-on entry, it would be count as the 50% profit already locked in.?
By "position in risk" we are referring to our actual trade account balance at risk. The only risk to that balance in any series (a series is a string of open positions on a given pair and remains active until all positions are closed by the daily dc/ichimoku cloud trailing stop) is the initial entry. Once that entry is at BE all subsequent entries in that series are funded by locking in 50% of the profits from all still open trades and using the other 50% of locked in profit to open the addons.

Hope this helps. If it's still not clear don't hesitate to ask and I can try and throw up some charts.
  • #257
  • Sep 13, 2015 9:24pm Sep 13, 2015 9:24pm
Quoting kprsa
I am following this thread with great interest. I do not want to diminish the efforts, and it is difficult to disagree with the result: It looks lake a low-risk, low-maintenance and high gain method, which is a goal outlined in the first post. With all this being said, I am sorry, I just do not see an edge here. OK, Donchian channel prevents stupid entries, as the buy and sell "hemispheres" are well defined and distinct. Still, there is little evidence that the long-period Donchian channel breakout implies a long trend. (Obviously vice versa is...
Valid points kprsa but I think timeframe does matter in relation to the use of simple breakout techniques such as the Donchian. What I have gleaned from trend following or indeed any simpler momentum strategy is that the edge associated with the technique lies in the ability to have a sustained price movement with general market consensus. Given the plethora of different market participants and styles of trading/investing, the only time we get a consistent consensus in price movement inevitably relates to when we have market divergence. These periods inevitably are associated with market 'crisis periods'. Most of the alpha generated in trend trading can be attributed to these discreet moments in time where market behaviour becomes consistent and unidirectional in nature (in periods of fear such as the 2008 GFC or more recently in times of market instability). Most of the success of trend trading therefore is attributed to what is referred to as 'crisis alpha' moments. That is why trend traders benefit from the fat tails of gaussian distributions.

Trend traders rely on market divergence as opposed to other trading stayles that rely on market convergence. The risk profile of a market convergent trader is one who takes plenty of small wins but ultimately is crippled by crisis events. Market divergent traders on the other hand like trend traders accept lots of small losses with the umtimate prize of huge wins when volatility becomes extreme.

The real purpose behind the Donchian on the higher timeframe is effectively to avoid trading in times of market stability (equilibrium) and only filter those trade opportunities that have the potential to turn into sustainable trends driven by market consensus or fundamental economics. It is not attempting to predict trends that is important. What is important is to keep the powder dry protectng capital while waiting for these unpredictable events to occur.....which has proven time and time again to occur with regular frequency and will inevitably become more and more extreme with persistent central bank intervention. That is why I find the Donchian is an excellent candidate to automate trend trading provided that the Donchian periods chosen is large enough to prevent overtrading which is an ultimate killer for trend trading.
  • #258
  • Edited Sep 14, 2015 12:43am Sep 13, 2015 9:48pm | Edited Sep 14, 2015 12:43am
Thanks Spyderman!

" So it's possible that every 4 hrs ( as the channel updates) you will move your s/l closer to your entry "

Let me try to understand it this way
Suppose an entry is of short entry,
with initial entry at 100.0
Initial SL at 150.0
After 4 hrs, DC-4-HR moved to 130, adjust SL to 130.0
The following 4HR, 130.0 got catched up, the initial entry stopped out with (130.0-100.0) loss.
Right? Thanks!
Outcome is not important at all, but correct technique is what only matters
  • #259
  • Sep 14, 2015 12:42am Sep 14, 2015 12:42am
Hi Spyderman,

Disturbing again, i have another questions

Considering the money managemnt, assume three entries would be considered, (No.1, No.2, No.3)
1)
THe time No.2 opened, now
- No.1 trade would be with 200 bucks in profit, then we locked in 100 bucks
- No.2 trade would be at risk of 100 bucks (Same as No 1's locked in)

2)
The time No.3 kicked in, we would have
- Given No.2 trade would be with 100 bucks in profit
- Given No.1 trade would be with 300 bucks in profit
We would set No.2 and No.1 locked in
- No.1 locked in 150 bucks
- No.2 locked in 50 bucks
And we would have 200 bucks as No.3' risk,

Please correct me if i am wrong, thanks!
Outcome is not important at all, but correct technique is what only matters
  • #260
  • Sep 14, 2015 6:21am Sep 14, 2015 6:21am
Quoting Copernicus
{quote} Valid points kprsa but I think timeframe does matter in relation to the use of simple breakout techniques such as the Donchian. What I have gleaned from trend following or indeed any simpler momentum strategy is that the edge associated with the technique lies in the ability to have a sustained price movement with general market consensus. Given the plethora of different market participants and styles of trading/investing, the only time we get a consistent consensus in price movement inevitably relates to when we have market divergence....
Hey Copernicus,
Thanks for your thoughtful answer. I understand what you say. And I do agree: It is reasonable that the larger scale events would be less prone to, say, false breakouts (stop hunts etc.), and that they should reflect better the economic divergences that ultimately drive the Forex market. Due to this, the simple trend following strategies might have an implicit edge. Additionally, the problem with lower timeframes is the (spread+transaction cost/movement) ratio, which I would expect to be negligible at higher timeframes, but not necessarily so at, say 5-15 min TF.
My current line of thought is to try to approach the market in a fractal manner. Sustained movements - trends occur at all timeframes, and I am trying to devise a system to follow trends at every timeframe -> basically caching momentum, and being direction-independent as much as possible. The Donchian by itself is not fit for this purpose, but I do have a couple of tricks of my own that could provide an edge at lower TFs.
Cheers,
k
0 traders viewing now