@Bob . . Do you want to know the answer ? .. if you are price actions /day trader or ATS .. is different .. in any case .. both need years of testing and learning ..this is "my trough " never buy any "miracle software " = don't exist @Aurel

It heavily depends on what's your appetite (and depo size), for some people 12-15% annual is plenty, for some 30% monthly is poor. Reasonably good strategy in my view should fit following criteria:

(1) you can explain it to a 10-year-old kid (maybe 15-year-old ))

(2) close all trades relatively fast (to sleep tight at night ))

There is not one answer to this question, but many.

No one will tell you if they have a well working approach - they may give hints, but not a manual of how to do.

So you have do do it your self, best advice is to be smart, do your desk research long time before you start writing a single line of code, of how to approach this using as little time and resources as possible - if you should at all write a single programming line is one of the decisions - if you should at all be involved in this another.

@Søren, pls extend your advice little further:

(1) better focus on Forex, or Equities, or both, and what's your personal preference?

(2) programming can be jungle and scary away good trader with good trading plan, why not outsource coding part to freelance expert?

@ Oleksandr - we could be interested in a profile like you for developing custom code for our clients, feel free to contact me if you are interested.

@Bob in the last 6 year's I have tested more than 800 strategy's.. 50 %.. are the moust know around net RSI/MACD/FISHER... GRID.. DD was more than you can sustain..

... . rest of the strategy's.. I have developed by my self.. from all.. just 2 are working great.. Also in this period of time I have worked with strategy developer's from around the world .. my question to this guy's.. was ::::***,,,.... Have you ever made a wining strategy (except grid)? . answer WAS = NO... in the end I decided to make my own team ..

Be careful.... all this guru's.. witch are around.. LinkedIn.. never make money in consistent way..

@ Aurel, this is not how it always works - the institutional path is:

An experienced trader, who knows by experience what works and what dosnt, has an idea to a new trading strategy or he has an existing he want to automate -> he write down this in a project manual ---> he ask a programmer/team, internal or external, to implement this ---> ........... months of coding, testing, coding, testing ..... ---> optimizing ... weeks .... ---> the final result is deployed if the final result was as anticipated.

@Soren ... How many people can afford to pay a full team to develop a stregy?.. It will need minimum 1 year.. 3 person + resources...!.. this is the reason why 70 %.. are losing.. I know only 4 person.. (from > 1000) *... are making money from full automatic trading ..

@ Aurel - very good question, and a very good conclusion.

If you check our website, you will see the headline "Custom Financial Trading", which is what institutional players, and the few like you, use today. The rest use "Fixed Financial Trading".

is above my league too.. but i am the developer / owner .. and I don't want to sell :) ask your self why >!

I and a colleague have developed a simple algorithm to tactically allocate between equities and Treasuries based on Risk On - Risk Off (RoRo) sentiment. It uses 3 ETFs: SSO, XIV, and IEF and is long only. Over a 5-year backtest (time-limited by XIV), it earns 139.8% and has a Sharpe Ratio of 1.62. It tolerates systemic stress well and has an out-of-sample Bayesian predictive distribution that is favorable for future returns. http://goo.gl/co9YlF

Seriously, You're asking this question? And what exactly do you think you're going to get as a reply?

Let me ask you something. Would you give away the location to the gold well if you were to find one? Especially if you have found it after wandering around for ages? A good trading strategy is like discovering gold.

@Khan . . . every day .. i have an idea of a system to trade or improving the existing one .. but never . . . that felling ...of . . the golden grail system.. unfortunate me :)

@Aurel Nice to make your acquaintance. Question, when improving existing strategies how often you realize the improvement is simpler than the original method?

It depends on what time frame you are looking to trade. On the long term I would look at a monthly momentum strategy, its been shown to perform well in bull and bear markets. On a shorter timescale you can look at stat arb strategies looking to minimise your market exposure. As some have mentioned there is no holy grail when it comes to what parameters or indicators you use. BUT the holy grail does exists its just sadly out of your reach. If you dont believe it just look at Virtu performance over the past 5 years or so... 1 losing day!

Fuad.. I know same person from... Vertu... before 3 years... we talked.. they don't suceed any profitable algorithm on currency.. Is good in HFT.. Stocks .. as far I know... no any person or company successful made HFT on forex.... regarding losing = is part of the game.. in 4 year's.. I am losing 3 months... every day will be one finding a new way... = evolution...

Bob, if you are really interested I can test the 2nd example: http://etfhq.com/blog/2013/05/04/market-timing-through-market-dominance-transdow/

The logic seems simple enough.The problem might be getting the correct data.

On the first link I just see results but not the logic of the strategies.

@Aurel

Regardless on the timeframe you trade, it is my strong believe that 5 years of backtesting is not enough to make a valid prediction that most possible situation or market movements have been covered.

@Volker... i have run the 25 years in the beginning (before 6 years) back-test .. i just find out after 2 years on forward (live) and 6 month (live) trading . . that the market profile is not the same and 25 years has no any add value in my test .. on my opinion the issue came from decision on entry and the management and portfolio exposure. .

. . .in any case i am not using time as entry definitions .. my system is not as any usual i have find .. over the 4 years research .. we can kip one order open > 1 h --> 4 months ---> 2 years .. depend on the symbol or instrument :) thanks to you all for the nice chat ... I will be in holiday for 2 weeks (no internet / no phone ) for any question/ comments ..@email :)

@Aurel, they made $42m in global currencies in 3 months... http://ir.virtu.com/releasedetail.cfm?releaseid=911120