Sugi Pulla System - UK Breakout Strategy: 370 pips in 2 weeks!

[Garbanzo]

Sugi Pulla System - UK Breakout Strategy: 370 pips in 2 weeks!

Hello,

I've been working on a strategy based on the 5min 7 am UK candle for the GBP/USD pair.

Let's call it the sugi pulla system.

Entry Rules:

You are going to need 2 accounts for this otherwise you will have issues with the entry orders - they will be like stop orders if they are for the same amount.

Place buy order 5 pips above the 5 minute 7am UK time candle and a sell order (on the second account) 5 pips bellow the 5 minute 7am UK time candle.
Place stops at the other end of the candle in both cases (do not include the 5 pips when placing stops).
Place take profit orders for both at 100 pips.
Close your trades if they havent reached either stop loss or take profit before 23:59 EST.

Comments:

I've backtested this strategy using 5 minute charts on FXCM,thus these results would reflect real world conditions and does not have the same problems that backtesting via other means has.
You can certainly backtest this successfully with 5 minute data.

You will often get whipsawed,either on both legs of the trade (long and short) or just on one leg.
Sometimes when you get only 1 leg profitable,it will not reach your take profit but will be profitable when you close it out at 23:59 EST.

I have backtested this with 2 sets of take profit targets,one with a 50 pip take profit and the other with 100 pips take profit.
The second backtest included instances where one leg of the trade was profitable but did not reach take profit while the other was a loss.
The estimates for these trades are not precise but id say it's -+10 pips at most.

Assumptions:

I assumed the 7 am 5 min candle was 20 pips + 5 pips entry = 25 pips.
So I automatically assumed each 1 legged whipsaw was 25 pips and each double whipsaw was 50 pips.
Most of those candles however where less than 20 pips so there's some extra profit margin to be squeezed out of there but I was just being conservative.

Keep in mind this is just 2 weeks of backtested data and the market may have just been favorable to this system over this period.
But also take into consideration the foundations of the system,it can be logically deduced that this system will work in the long run simply because its safe to assume that every now and then the pair in question will travel from its open 100 pips in one direction or whipsaw once and then travel in that direction.
The issue is the number of whipsaws and this can certainly change but I still think its going to be profitable.
If you can do more long term backtesting reliably,be my guest maybe I am just an optimist because this is my pet theory.


Results:

Take profit: 50 pips

8/7/2009: 25 pip profit
9/7/2009: 25 pip profit
10/7/2009: 50 pip whipsaw
13/7/2009: 50 pip profit
14/7/2009: 50 pip profit
15/7/2009: 50 pip whipsaw
16/7/2009: 50 pip profit
17/7/2009: 50 pip whipsaw
20/7/2009: 50 pip profit
21/7/2009: 50 pip profit
22/7/2009: 25 pip profit
23/7/2009: 50 pip whipsaw
24/7/2009: 50 pip whipsaw


Total P/L: 325-250= 75 pips profit


Take profit = 100 pips

8/7/2009: 75 pip profit
9/7/2009: 100 pip profit
10/7/2009: 50 pip loss
13/7/2009: 75 pip profit
14/7/2009: 60 something pips profit
15/7/2009: 25 pip loss
19/7/2009: 35 pip profit
20/7/2009: 100 pip profit
21/7/2009: 50 pip loss
22/7/2009: 100 pip profit
23/7/2009: 50 pip loss

Total P/L: 545-175= 370 pips profit




Warning,unlike my daily breakout strategy (http://www.trade2win.com/boards/for...0-20-pips-every-day-95-profitable-trades.html)
This strategy has a much lower success ratio (the number of times it goes your way) witch statistically implies there's a higher likelyhood of hiting a prolonged period of bad trades that can mess up your account really bad.
So I urge you not to leverage to the hilt with this strategy.
Try and keep those maximum loss potential per trade of 50 pips to 2-3% of your account equity.


PS: I haven't traded this live yet but I am planning to do so this week - this doesn't change the real backtested results in any way.

Good luck.

 

[shadowninja]

I'd say forward test before committing real money. A sample of 10 trades is not big enough. Thanks for sharing it, though. Straddling is an old system that generally doesn't reap good rewards (or everyone would be doing it).

 

[Garbanzo]

Hmmm...
I think disagree with you on that and that certainly feels weird when you obviously have more experience than me!
In my opinion,unlike in options trading,there is no increase in implied volatility that would make spot FX more expensive to trade if somehow it was known by the market that a significant move is ahead in either direction.
Pairs with large daily ranges like GBP/USD are inherently more volatile and there's simply no implied volatility that will suddenly make the transaction costs increase like in options.
So as long as there is volatility (150 pips average daily range for the cable) it seems obvious to me that there is always going to be the likelihood that price will move in one or two directions decisively after the open more often than not.
This is not straddling in the options market meaning of the word,here you don't need an informational advantage of better analysis to succeed - the volatility is there,always.

 

[Garbanzo]

Also for everyone out there thinking about the hurdle of having 2 accounts to trade this,you could probably just make EA to handle the long/short issue.

 

[shadowninja]

I don't mean straddle options. I don't do options.

I just mean straddling a price by putting a buy stop and a sell stop in place.

 

[Garbanzo]

I don't mean straddle options. I don't do options.

I just mean straddling a price by putting a buy stop and a sell stop in place.

Yeah I understand that,I was just pointing out that its true,in options,it probably will have bad results unless you have inside info because the implied volatility will take care of it.
In spot forex,there is no such thing as implied volatility increasing your position overhead,and also the strategy doesn't involve straddling news witch is extremely dangerous imho

 

[shadowninja]

Well, good luck (although luck shouldn't be needed!) and let us know how you get on!

 

[azwhaley]

Hi Garbanzo,

How's it performing 3 weeks later ? I tried a similar competing strategy on the FTSE100 but it was rubbish but it was based on the previous day's mid point value.

- Andrew

 

[phreddyinbg]

By the lack of posts, are we to assume that it didn't work Garbanzo?

 

[pseudo straddle]

I'd say forward test before committing real money. A sample of 10 trades is not big enough. Thanks for sharing it, though.

Straddling is an old system that generally doesn't reap good rewards (or everyone would be doing it).

Says it all, really!

 

[shadowninja]

Well, I cunningly use the term "generally". I'm sure it's possible that there are people out there making a living from it.

 

[phreddyinbg]

Originally Posted by shadowninja Straddling is an old system that generally doesn't reap good rewards (or everyone would be doing it).

So do we assume then that nothing works (or everyone would be doing it), Oh wise one?

 

[phreddyinbg]

Sir, you have my total respect!!
Anyone who can answer a question before it is asked...My msg 1:21 Your answer 1:20

I am coming to your blog instantly but then you have probably already greeted me.

 

[shadowninja]

I knew you'd say that.

On a personal level, my experience is that I cannot scalp profitably despite years of trying. Does this mean nobody can scalp profitably? My answer: I cannot swim the English channel, nor can I do the Four Minute Mile.

 

[phreddyinbg]

Getting back to the meat of this dysfunctional thread. I am interested in breakouts and feel that they could make money if treated right.
Just looking at a day's movement the price will leave any point selected and travel down or up from there. So we select a point and a target - simple.
Problem comes with price not going where it is eventually going directly.
It wanders around back and forth taking us out of our stop loss then going where it was required in the first place.
What then is the answer?

I have seen many schemes for the starting point ranging from:
the first hour of the day - this leads to lots of pages where people discuss just where midnight GMT is on their charts;
First four hour candle - with the same discussion;
Overnight, the asian session - mustn't use overnight it is only overnight for Europe, it is daytime for me where I am now but being English I still call it overnight;
Anything from 15 mins before opening time Frankfurt to two hours before.

I will stop here in case someone wants to discuss this, otherwise I rather feel that I will be talking to myself. Mind, that does allow me to assemble my thoughts on the subject.

 

[shadowninja]

I've played with them. The thing is, breakouts tend to end up as fakeouts (whipsaws) so you'll take a few losses then have to let the winner run to cover the losses. A breakout should be immediate (otherwise is it a breakout or a general meandering in your direction?) so if it doesn't break out of a range immediately, should you close for minimal loss/profit. Another aspect is identifying a breakout vs a fakeout. I suppose you could look at the "price action" (whatever that means ).

Oh, someone on here suggested that you only take trades in the direction of the trend as you're more likely to achieve target.

 

[phreddyinbg]

Another aspect is identifying a breakout vs a fakeout. I suppose you could look at the "price action" (whatever that means ).
Oh, someone on here suggested that you only take trades in the direction of the trend as you're more likely to achieve target.

I think that breakout is a misnomer anyway. The market doesn't tend to behave that nicely for us. I am thinking about trading when price has left the nest.

Some people suggest a buffer of 5 to 10 pips my experience is that price still peeps above the line, triggers then retreats again.

The 'with trend' idea is always good if price is going to behave when Frankfurt then London opens. I have seen it expressed that Frankfurt sets the direction and London follows. I don't know the data on that but directly I heard it, it turned out true. Next day it reversed.

Another idea is to use a moving average and trade when that 'leaves the nest'.

Phil Newton talks about a false breakout and trading the next move out.

Of course, all this leads on to the stop loss - how large bearing in mind that it might be hit with a false BO as the price goes the other way and triggers the opposite trade.

Too late for me to sit up any longer, time for bed.
Thank you for your contribution.

 

[shadowninja]

Yes, the trick is to figure out which is the fakeout and avoid losing money on that.