Reverse Engineering a Profitable System (REPS) | Page 5 | 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)

  • #81
  • Aug 11, 2015 7:13pm Aug 11, 2015 7:13pm
Quoting Copernicus
{quote} Spidey You might be interested in applying an anti-martingale progression to your trends such as the Snowball
I'll definitely look this over. I've used the snoball EA before but as a stand alone. I'll do some research.
  • #82
  • Aug 11, 2015 7:41pm Aug 11, 2015 7:41pm
Nice thread!
If you are good at something, never do it for free--Joker
  • #83
  • Aug 12, 2015 12:43am Aug 12, 2015 12:43am
Quoting spyderman
{quote} I'll definitely look this over. I've used the snoball EA before but as a stand alone. I'll do some research.
Hi guys,

I'm following this thread since its beginning. I really like spyderman's trade management. Long time I was searching for a pyramid scheme which doesn't increase the risk. I studied 7bit's snowball up and down. It looks great on paper. Unfortunately the asymptotic performance hides an awful linear drain on the account. This linear loss depends on the future volatility and can get bigger than the quadratic profit at the target. I mean that at the target exit your "winner" may be in a net loss! A truly negative RR! This DD is clearly seen on the backtest by Copernicus. Remember it is a cherry picked example. With Spyderman's way the risk is known at the beginning and after the first addition a small profit is guaranteed. Even if no trend emerge we may get a small profit or a well controlled loss. In the backtest, the balance DD is approx 50% yet the peak floating profit is half of the realized profit in the example given by Spy'. No contest.
No greed. No fear. Just maths.
  • #84
  • Aug 12, 2015 1:03am Aug 12, 2015 1:03am
Quoting PipMeUp
{quote} Hi guys, I'm following this thread since its beginning. I really like spyderman's trade management. Long time I was searching for a pyramid scheme which doesn't increase the risk. I studied 7bit's snowball up and down. It looks great on paper. Unfortunately the asymptotic performance hides an awful linear drain on the account. This linear loss depends on the future volatility and can get bigger than the quadratic profit at the target. I mean that at the target exit your "winner" may be in a net loss! A truly negative RR! This DD...
Good advice PipMeUp. Thanks for that. I do see how the Snowball fails on less volatile trends as it can't get ahead of it's break even. :-)
  • #85
  • Aug 12, 2015 3:12am Aug 12, 2015 3:12am
Welcome JMN...Pipmeup

Update on E/C trade. Moved to BE Now we just let the market decide...and of course, root for a climb.
  • #86
  • Aug 12, 2015 3:14am Aug 12, 2015 3:14am
After some revisions to Post #1, we're back in the "Trading Systems" forum.

Thanks Twee
  • #87
  • Aug 12, 2015 3:48am Aug 12, 2015 3:48am
Quoting spyderman
Moved to BE Now we just let the market decide...and of course, root for a climb.

And...we're out at BE.
  • #88
  • Aug 12, 2015 6:38am Aug 12, 2015 6:38am
Quoting spyderman
{quote} And...we're out at BE.
When you entered the position your SL was 230 pips. This gives room for the trade to breath. But as soon as the price reaches 50 pips profit the SL gets very close to the PA. No more breathing. Additionally 50 pips on E/C are not the same thing as 50 pips on E/G. It isn't related to the volatility. What about using a DC 20 on H1 and trail the SL @BE when the channel is fully above/below the entry?

Coincidentally for the E/U example this would have been around 50 pips. But for your E/C you would still be at risk (good or bad time will tell)

Your thought?
No greed. No fear. Just maths.
  • #89
  • Edited 7:08am Aug 12, 2015 6:52am | Edited 7:08am
Quoting PipMeUp
50 pips on E/C are not the same thing as 50 pips on E/G.
Exactly. Also 50 pips on E/U in 2015 are not the same thing as 50 pips in 2012. Another way to solve this problem could be by basing your SL off an ATR value to account for the volatility of different pairs.
Confluence is king.
  • #90
  • Aug 12, 2015 12:32pm Aug 12, 2015 12:32pm
Quoting PipMeUp
{quote} When you entered the position your SL was 230 pips. This gives room for the trade to breath. But as soon as the price reaches 50 pips profit the SL gets very close to the PA. No more breathing. Additionally 50 pips on E/C are not the same thing as 50 pips on E/G. It isn't related to the volatility. What about using a DC 20 on H1 and trail the SL @BE when the channel is fully above/below the entry? Coincidentally for the E/U example this would have been around 50 pips. But for your E/C you would still be at risk (good or bad time will tell)...

Yeah, I've been thinking about it all day. I realize I was too quick on the BE. At the very least I could have waited for some structure to form support.

We had mentioned the H1 Donchian earlier. So you're saying just use it as a volatility indicator and move stop to BE when the H1 channel has cleared price?
That would seem to be more in keeping with the flow of the series in general.
  • #91
  • Aug 12, 2015 12:35pm Aug 12, 2015 12:35pm
Quoting Liquidity
{quote} Exactly. Also 50 pips on E/U in 2015 are not the same thing as 50 pips in 2012. Another way to solve this problem could be by basing your SL off an ATR value to account for the volatility of different pairs.

Yes, that would work. I think I like the idea of sticking with the DC's. Looking to hear back from Pipmeup to make sure I understood him correctly.
  • #92
  • Aug 12, 2015 3:21pm Aug 12, 2015 3:21pm
So here's another wrinkle to work out.

What do you do about entering (or not) a trend once it's been on it's way for awhile.

For example here on the Gbp/Nzd, it might look like the trend should be exhausted.

Attached Image (click to enlarge)
Click to Enlarge Name: GBPNZDiWeekly.png Size: 60 KB


Not so much when you zoom out though

Attached Image (click to enlarge)
Click to Enlarge Name: GBPNZDiWeeklya.png Size: 73 KB


I'm thinking perhaps we should start a new "leg" when the weekly DC gets hit again. Obviously if we had been closed out of a series on the daily that would mean that price has come down/up off the Weekly extreme, thus perhaps setting up another trade series.

Any thoughts?
  • #93
  • Edited 5:23pm Aug 12, 2015 4:19pm | Edited 5:23pm
Quoting spyderman
So here's another wrinkle to work out. What do you do about entering (or not) a trend once it's been on it's way for awhile. For example here on the Gbp/Nzd, it might look like the trend should be exhausted. {image} Not so much when you zoom out though {image} I'm thinking perhaps we should start a new "leg" when the weekly DC gets hit again. Obviously if we had been closed out of a series on the daily that would mean that price has come down/up off the Weekly extreme, thus perhaps setting up another trade series. Any thoughts?
Morning from here Spidey

Treating each trade as a seperate series on a trend isn't a bad idea.

For example with the earlier post 66 using multiple timeframes on EURUSD different timeframes can be deployed to ride a single trend. The advantage of this rationale is that as volatility changes along a trend, each trade therefore adapts to the changing volatility and also the various timeframes used ensure a staggered approach to entry and risk is configured to each timeframe providing the required breathing room associated with each timeframe's associated volatility range. While post 66 provides a way to operate 6 different trade series on a trend (assuming 6 timeframes used), your approach can deploy the same rationale if you treat each trade as a new series and use say ATR as a basis for position sizing. Doing so ensures that each trade adapts to changing volatility along the length of the trend.

Also I have low risk tolerance by nature and if it was me, I would tend to get progressively more risk averse as the trend extended and establish conditions such as reducing position sizes by decreasing trade risk% of additional entries along the way to dilute the risk at the moving end (right hand side) of the trend. This way my effort is made up front and therefore time in trend becomes your friend.....but that's just me....and others would strongly differ in that view.

Good to see that this thread is getting more contributors. Many heads are better than a few and it is great seeing a strategy unfold before your eyes. :-)
  • #94
  • Aug 13, 2015 2:02am Aug 13, 2015 2:02am
Quoting spyderman
{quote} Yes, that would work. I think I like the idea of sticking with the DC's. Looking to hear back from Pipmeup to make sure I understood him correctly.
I'm sure you got it. Here are two pics
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSD-H4.png Size: 18 KB
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSD-H1.png Size: 20 KB
No greed. No fear. Just maths.
  • #95
  • Edited 4:34am Aug 13, 2015 4:22am | Edited 4:34am
Quoting PipMeUp
{quote} I'm sure you got it. Here are two pics {image}{image}



So I think PipMeUp has a good idea on the BE move. We'll try moving our stop to BE after the 20 period H1 DC has cleared our entry point.

In regards to trend continuation, my goal here is not to predict price movement, but to follow it.
Therefore I can't really know if one break of the weekly Donchian Channel is more promising than another, so I will treat each as a separate trading series opportunity.
It's going to produce a lot more trading opportunities and time will tell if that's a good thing or bad. This is definitely a work in progress, so we'll monitor as we go.
  • #96
  • Aug 13, 2015 4:28am Aug 13, 2015 4:28am
We reverse engineer a perfect trade but no position survived between 1.25 and 1.29. It isn't that perfect! The reason is obvious: the trend wiggled. Is it possible to hook a few trades in there without increasing the risk or aborting the series?

I wondered how we could enter the retracements with the tools we have, i.e. the DCs. I got an idea. What do you guys think about this: when the DC on D1 is still going down but the DC on H4 is going up we enter at the B/O of the H4 channel. Picture: (green entries survived, yellow taken @BE, no loser)

Attached Image (click to enlarge)
Click to Enlarge Name: EURUSD-H4.png Size: 28 KB
No greed. No fear. Just maths.
  • #97
  • Aug 13, 2015 4:49am Aug 13, 2015 4:49am
Quoting PipMeUp
We reverse engineer a perfect trade but no position survived between 1.25 and 1.29. It isn't that perfect! The reason is obvious: the trend wiggled. Is it possible to hook a few trades in there without increasing the risk or aborting the series? I wondered how we could enter the retracements with the tools we have, i.e. the DCs. I got an idea. What do you guys think about this: when the DC on D1 is still going down but the DC on H4 is going up we enter at the B/O of the H4 channel. Picture: (green entries survived, yellow taken @BE, no loser) {image}...
Hmmm. Definitely aggressive stacking. I'll look through the chart for the trend and see where that leads.
We greatly increase profit potential but at the same time obviously increase the chance of hitting our s/l at loss and loosing our trade progression momentum.
  • #98
  • Aug 13, 2015 4:52am Aug 13, 2015 4:52am
Quoting spyderman
{quote} So I think PipMeUp has a good idea on the BE move. We'll try moving our stop to BE after the 20 period H1 DC has cleared our entry point.
Fully agree. This H1 DC clearing the entry point should give the very first position enough room ...
MT4/MT5 EAs/Indicators/Alerts coder
  • #99
  • Aug 13, 2015 5:01am Aug 13, 2015 5:01am
Quoting FerruFx
{quote} Fully agree. This H1 DC clearing the entry point should give the very first position enough room ...
Thanks for the input FerruFx. Much respect
  • #100
  • Aug 13, 2015 5:04am Aug 13, 2015 5:04am
Quoting spyderman
{quote} Hmmm. Definitely aggressive stacking. I'll look through the chart for the trend and see where that leads. We greatly increase profit potential but at the same time obviously increase the chance of hitting our s/l at loss and loosing our trade progression momentum.
So taking a quick peek through our series chart it's seems we'd add the following entries. All would have got to BE. Now some may not have been taken due to trade spacing from previous entries and some would of course get taken out at BE, but definitely would have added some positions. I even checked the end of the trend where price eventually turned and closed the series, as I feared we may have a loss down there...but we didn't

Nice find PipMeUp.

Attached Image (click to enlarge)
Click to Enlarge Name: EURUSDiDaily.png Size: 65 KB


EDIT: There were a couple more entries down there in the junk at the bottom of the series that I forgot to circle. Both went to BE and then stopped out.
0 traders viewing now