SBOR II - Session Breakout and Reverse strategy (universal)

Intro:
This thread is dedicated to my only compact trading strategy called SBOR. Unlike many other trading strategies this is a statistical approach to making the right trading decisions.

The Idea:
If price breaks out from and later returns into a session range, few pips below/above the breakout candle's low/high, price will fall/rise further as much or more as the size of the breakout candle, if certain conditions are met.

How do we know what the conditions are for winning in this game? We register past session and candle OHLC data and register the best price before hitting our SL. By creating meaningful variables based on OHLC and session data and the relationship between these two, we can tell winning candle & session combinations from the loser ones. This we can do as we know which combinations were losers and which ones were winners in the past.

So this strategy is a reverse strategy (back into a session) that is based on candle and session analysis and probabilities based on past data. Also, I would like to emphasize that this strategy is neither a breakout strategy, nor a counter trade strategy, nor a trend trading strategy. It is a scalping strategy that aims to harvest pips in favorable market conditions.

SBOR history:
About 18 months ago I devised a strategy that did not raise much attention here at FF. Original SBOR thread (for EURUSD only): http://www.forexfactory.com/showthread.php?t=296776

Due to technical issues, I could never prove that my approach was right. Technical issues were lack of input data and an EA to back test and forward test the strategy. When I formed the strategy, the input data was entered by hand, data read from charts, giving me weeks of work. These issues have been partly resolved by now. Just one note before people start to have doubts: this strategy gives the near exact same results on back test and forward test, on live or demo account. Open price testing method is fast and sufficient.

This is how it works (overly simplified example, see picture):

1. red vertical line: candle broke through the session border (the previous session, not the one that it is in)
2. sell stop order is placed below this candle (assuming conditions are met) by x pips. We wait with the pending orders after candle close.
3. TP level is determined by probability (based on plenty of similar candles: what was the best average TP level? It works, although it can be fine tuned)
4. SL is above the BO candle by x (few) pips

Sessions to trade:
There are 4 sessions to trade [GMT+1, DST: GMT+2 spring and summer in Europe]. The EA comes with a check box for getting this right, however it is easier to manage this for those with brokers in Europe as changing times can disturb this strategy a lot).

1. Asian full from 1am-8am,
2. Asian late 5am-8am,
3. EU 8am-14pm,
4. Big dog 14pm-16pm

Eventually, in the full version EA anyone can define a session. These are the sessions I started with but there is a chance that the strategy works well with daily or any sessions a trader can think about.

Candles to trade (single and combined candles):
We trade either a single candle [the first candle that breaks through the session range by min. x (few) pips] or the two candles that follow the single breakout candle [only if the first candle does not qualify for trading]. The candles qualify for trading based on candle OHLC and session (high/low) data.

SBOR EA and Scripts:
I am forever thankful to Hanover who volunteered to code the EA for this strategy.

*** I am also thankful to Billbss who actually made the arrangement with Hanover to code the EA. I am thankful to Giraia_br as well, as his questions made me come up with plenty of explanations so the strategy has become more and more elaborated theoretically ***

Hanover (David Louisson) is by far the most professional and responsible coder I have ever worked with. His earlier and latest master pieces can be found here (many I use myself): http://www.forexfactory.com/showthread.php?t=280437

Unfortunately, due to his busy schedule he could not continue to develop the EA further but generously consented to share the EA on FF.

I have few ideas that should be implemented in the EA and the scripts to call the SBOR system complete. These are not endless ideas because the SBOR system is compact and a closed up system with fix processes. Hanover's original scripts are attached in a zip file, the rest I modified but I am not a coder so you can double check the differences.

Important note: any results of the current EA can not prove the system right or wrong (you will find it a winner from the time it was designed though with a BE activity for the past 5-6 months). The system is a framework thus the EA is a reflection on how the trader utilizes this particular framework. It will become obvious once one understands the full process of strategy creation. You can back test the EA but be aware of testing rules described in the documents. Notice however that, technically speaking, it is not a complete system yet. Some more coding is required before we can call it complete.

Test results of the current SBOR EA:

1. Ran between: 13.03.2012 - 13.07.2012 – tested for 4 months, ran non-stop on a server, all possible trades were taken
   
2. No. of trades: 121
3. Risked: 1% of equity
4. Max DD: ~14%
5. Max profit: ~15%
6. Current equity: ~BE,-137$, equals to 2 losing trades (BE does not mean this strategy can achieve BE only, I will explain this later)
   
7. On Demo and Live: nearly identical results, live did better but no significant difference experienced.

For now, I would like to draw attention to something important regarding the results: not one single indicator was used to achieve this and the settings are considered to be outdated. In the EA, I would like to use indicators only to filter out some bad trades.

Skills required:
Excel, including formula building (simple stuff that goes complex).
Coding skills to implement Excel formulas as mt4 code.

Files shared:
- SBOR-specs.doc
- EA and file specs.doc
- SBOR calculator only.xlsx
- SBOR EA - modified by Rewand.mq4
- SBOR_test_1_modified by Rewand.mq4

Hanover's original scripts
- SBOR EA.mq4
- SBOR_test_1 .mq4

Risk: is yours, as always. I do not take responsibility for whatever results the users of this strategy/EA achieve. If you do not know what you are doing, sooner or later you will fail in Forex.

Copyright: I maintain the right to sell/distribute any products/services related to this strategy, including the SBOR EA and SBOR scripts. I do not have such plan because I believe there are plenty of free systems to make money with but be forewarned that this thread has to be the only place for distribution.

What I am looking for:

I hope that someone will take interest in completing the scripts and work on that for some time. What has to be done is explained in the attached documents, what is not, I can explain later. All results of this work will be shared.

If you are a trader only, you can still familiarize yourself with the strategy and its elements but I have to draw your attention to the fact that this strategy requires the codes to be fixed/completed first.

Hi,
The full documentation of this strategy has been posted. I look forward to getting some valuable feedback.
Thanks,
Rewand

Dear Rewand!
I think you made big thing here. I red this thread and the old one twice and I think you catch the principles described in article I uploaded here. It was taken from X-Mans thread. If so I can imagine it can work on any volatile pair, but the best will be EU.

Unfortunately Im no coder so I cant help you to make this system (EA) complete.

Even when I red your article several times I do not fully understand some things. Maybe because english is not my native language..

Do I understand it right that you "program" all conditions that met your criteria in relation of size of the session box with size of the breakoutcandle into excel sheet?

So you made statistics from which you try to predict future situation where conditions are closest to conditions from past?

I would love to try your system but right now im not sure that i know how. I think I understand the principles but I do not know when will be "met the certain criteria" and how to recognise "certain criteria are met"

I know few coders can help you and I think I have a technical MM concept which can resolve the problem with TP (no need to compute breakout candle(s) length).



Be pips with you
With honor to your work
C.

Hi, thanks for the appreciation and for the understanding as well. I will check the docs once I will have some time. Unfortunately, I am too busy these days... Technical steps and Variables in the SBOR-specs doc describe the process nicely. "Conditions" are past values (or interval of values) of variables regarding session and BO candle OHCL data that resulted in pips. Therefore your second statement is right! Also, this is candle stick analysis, only we do not use our eyes while doing so.

There is no way to recognize when the criteria are met during trading (only we see that trades take place. Why, we would not understand, only understand the mechanism). Criteria forms when the strategy is formed using values of variables from the past. You would only know what you are doing when making up the formulas. Later on you would only know only that the EA executes the trades based on your formulas.

An example for the term "certain criteria": if the session is at least 25 pips wide and the BO candle is between 12-14 pips wide and its body is at least 75% of the total candle size and it breaks out at least 7 pips of the session and it is a continuation candle (meaning the candle is green when breaking out north) => then trade short, TP is x1.5 of the BO candle size. In the formulas you form such criteria. Such conditions should be met for the EA to trade. But these formulas are up to the trader. Does this formula make any sense? To me, it does not. I only know that when I formed this criteria it resulted in positive pips (and was based not only on few cases but numerous cases).

The TP issue - statistically speaking - is not a problem. The current approach works. Everything is based on statistics and TP is not an exception. However, if there is a better solution it is all for the better, I am opened to it. The basic concept though is that the TP should be at a minimum BO candle size distance from the high/low of the BO candle.

How you can try the system is that you can put the EA on the EU pair and watch what it does. I will post a template this week. Also you can backtest the EA as well. It should show profit for the past 3 years if calibrated well.
But this whole thing will start to be much more exciting if the scripts get beefed up a bit. This is really a must, otherwise traders will feel this EA is a black box even if it is not.

Hi.
Your formula does make sense to me and it is one of the pilon to understand better your strategy. If I understand it well, the excell sheet is not for entry calculaton but for exit (tp) calculation and SL will be placed on HL of breakoutcandle(s)? Im asking because I would do some manual trades on my live micro account. To be honest excel is not "my" program. Shame on me.

I will ofcourse setup a demo with your eas and watch them.
In our repulblic lives man named Karel Janecek. His firm RSJ is a market maker for Futures derivates, Oil etc. He studied Mathematics Bradley University in Peoria and Carnegie Mellon University in Pittsburgh.


He convert mathematical and stochastics scienece theory of probability into trading system algorithm which is making him a big amount of money yearly. It is billions USD.

In severeal interviews he told that this research of the algorithm was for about 10 years. And every event even in market conditions can be counted by the probability. In the computer there are programed all possible scenarios and what it does is to recognise when to trade and when the risk is too high to not to trade.

He has 50 empoyes which are controling the computer algorithm and news events (which cant be programed) If a really bad news happend they got about 11 seconds to close all trades. 11 seconds is the dealy between bad news released (by reuters) and actions on the market.

I think you are thinking very similar and I think that the way (hard way) to go. But it require high analythical mathematical brain.

Karel janecek Says that no one from small investors cant make money with short and mid term trading for a long period, because the big players and their computers simply beat and rulez the market. Thats why almost all here fail with their systems sooner or later.

Does anyone even think that he can beat those multi-million companies with super-computers and mathematical-analythical staff with rsi, Stoch and MAs cross for years?

On the excel sheet, when making the formulas, the trader decides which BO candles to trade. The SL is fixed, the TP is variable based on BO candle size (see TP multipliers).

If you want to do manual trades you will have to enter session and candle data into the Excel to get the trade advice. I suggest you do it on demo, as it does not make sense to do it on your live account. Also, just do it for the sake of getting familiar with the process and not to make pips.

Only because one spent 10 years to get the key, does not mean one can not get it even in months. Key means anything that exists and once one gets it, doors open. The problem in fx trading is that people can not find the key, although it is everywhere in all kind of forms. One who is not ready, will not find it, even if it is in front of his eyes. The plain truth about human nature that has nothing to do with fx trading.

The guy must be right with his strategy. However, although he has 50 people but in the end there is one person who has to push the button, meaning one individual can have it all and can take responsibility for the decision. The edge is that s/he does not have to spend time with nitty-gritty issues.

What I can say is to repeat that my system is a compact system. Any indicator and simple ideas can make it better but it is OK without it. The big ones have their logic and so do the small ones. I do not fight with the big ones, I fight with myself. I rarely feel that my broker or anyone is taking my money but blame myself if I lose. The big ones only move the markets due to their size but are not taking my money. Furher on they must be beating each other as well, not only me, to make money.

The Guy Trader story was a thought how can one man beat the market. He and his wife are on the black list and are not permited to play in cassinos. That was only a thought about a guy who can count probability, because he believes that almost every event in life can be counted..

It was no offence against you..
I really appreciate your work..

Thank you for your system, there is 30m TF to trade, right? I think TF is important for excel data input..
---------------------------------------------------------
My goal is to trade monthly or weekly tf. I have developed MM concept which can collect all the pips from trend. This concept was separatly coded
by generous mr Hopwood. The ea (its not a system just a tool) is somewhere here on FF but i wouldnt upload it here because it is not complete and later it was messed with ideas of contributors.

now I am looking good system where the entry has ambitions to catch weekly or monthly moves. And I think your system is very close to what I am looking for. I was somehow covinced that it must be a breakout system. I would call your system Postbreakout :-)

Let me give you an example of a trade with my concept of mm.

lets say we have a sellstop pending order placed on the low of breakout candle with SL for e. 20pips @1.22000 (SL 1.22200, no TP) Lets call this "safety trade"

20 pips under the "safety trade is pending sellstop (@1.21800) begining a grid of sellstop pendings with 10 pips TP and 10pips distance from each other (10 pips increments series: 1.21800, 1.21700... etc.) No SL.

If price goes against us by executing the "safety trade" @1.22000 we have 20pips SL.
If price goes in our direction it executes first sellstop from the grid @1.18000 and SL of the Safety trade is set to BE. Now we have and we will always have opened two trades:
a "safety trade" and executed sellstop from the grid.
if price reverses from any of the executed grid of sellstops the whole trade should always end with BE by positive pips of the "safety trade" and negative pips of a grid trade. more the price goes our direction the more room we have for corrections. This must be controled by EA.

Therefore fast breakouts are needed for this concept.

we can via the EA create 100 or 1000 sellstops to cover all the movement.

My goal is get on the boat and ride for 2-3 months.

I did an experiment with breakouts of round numbers and I made about 3300 pips in 6 weeks. That was of course coincidence. I had no system, no concept. I just set a pending grid 10 pip above (when price went from bottom up) or bellow the round number. All the roundnumbers I traded made a longer breakout candle so there was no DD at the beinnig. Simply I had a luck.

This concept is easy to use but not easy to understand when written by a person like me whos English is not native language.

Feel free to ask any questions or simply dont accept it.
i Will accept both. But try to understand it first..

Good luck and pips be with You..

Hi Carolco,

No offence, I did not take it like that and I do not have any opposing thoughts ; )
I am not a math guru but probability is an absolutely valid approach in life.

TF is M15 in SBOR. The approach can be extended to other TFs so if you are looking at higher TFs similar thinking can be applied, only you have to have enough cases to draw conclusions.

Postbreakout... sounds acceptable to me with with reverse and continuation types of trades. SBOR is the reverse type. I started dealing with breakout trades and they did not work (I would trade it based on SR levels, bounces) but I ended up with a reverse strategy. I believe however, that BO can be harvested with the similar approach. Just no time to deal with that right now but have plans to do that later.

The MM I somewhat understand but SBOR can not give you the direction of the trend. These are short term trades, similar to trading candle formations on low TFs. The best approach on trends I know is the NB10.2.

I try to keep things simple here, because I do not have time for trading and looking at things right now.. not even for posting much : (

Perfect! I understand completly. Looking forward to hook your ea on demo. for sake of order will do it next week.

All good

Hi,
I do not get much feedback, so I want to ask for opinion. Does the strategy look too complicated or hopeless to you guys? Or is it not that fancy to raise attention? In fact, a little bit of programming might cause a breakthrough... Just to repeat myself, this is a closed system, there is no space left for people improvising in it and this system does not need new ideas. Only have to learn the process that will never change. It has proved itself and those who put the EA on would see it is working rather reliably with input data that was created one and half a year ago. Must I promise the blue sky to raise a little bit of attention? Please, at least someone refute the idea or tell me it is plain stupid.
Regards,
A.

I'll give EA on EURUSD H1. If the demo account grow, so will the strategy studied in great detail. Then write your opinion.

EA does not open any trades.

Hi, thanks for trying, put it on M15 only. You can use the template from SBOR I. thread and do not forget to calibrate the EA so that you align it well with the sessions.

Hi, I gave EA the majors and EURJPY, M15. All transactions resulted in a loss. Calibration is OK.

Hi, it is good to read the manuals. It will not work without it. EUR/USD M15 only. For a very good reason...
Attached is what my EA did the past 10 days.

I am looking to trade the EURUSD.

EA is shareware trial version for 14 days?

As long as it is developed by people at FF it is free. The EA has no restriction in its current form. But do not use it on your live account until the strategy is coded completely into it. There is a copyright note however on the first page, I assume people to read it.
Anyway the EA has been running for 5.5 months non-stop, it is practically at BE without using a single indicator.

Thank you for the good news.

Just one more note on why it will always be free here at FF. It is because it is a think-set-forget EA and if anyone runs it without understanding how the decisions were made by others s/he is doomed to failure. The current settings are old, they worked for a year but now produce BE only. It can do better with few indicators but my main point is that it has to produce profit even without them. So new dataset is required to revise decisions. This is what I want from contributors. This EA and the strategy is fully transparent, just run a backtest and you will know why. Also, for validity check, you can put the EA on live and demo account with 0.01 lot and test the differences yourselves without much risk to lose. Technically, the EA works fine.