Topic: Cant make a single strategy work????

No matter what I do....my buy and sell signals are completely backwards.....I will be 100% honest...If I were to start the automatic execution and sit all day infront of the computer and do EXACTLY the opposite of what the system is trying to execute....I will make money! And simply just reversing the strategy wont work! It still doesnt execute a "good" order! No matter if it is trying to execute a long or a sell.....it drops IMMEDIATELY from where it executes....so my logic is manually override the system and we will profit! HAS ANYONE BEEN ABLE TO ACTUALLY come up with a winning strategy and is ACTUALLY making profits with an atomated system? Also, you cant exactly do "back testing" because all the data is from the past!!!! We are betting on the fact that the market trades consistantly, and it doesnt! So "back testing" is useless........someone please help me understand this if I am wrong!

Re: Cant make a single strategy work????

Yes, you are right.
I never found a good strategy, because of your reasons.
But I know that there are some automated systems, which work. But they are secret.

You have to update your strategy on a regular basis. And use a fitting strategy for the current market, down, up or no direction.

Re: Cant make a single strategy work????

I have developed over 40 strategies with successful backtesting results. In the real world I have only 3 strategies that finished forward tests on demo account. It is possible to build such a strategy that will provide moderate profits.

Re: Cant make a single strategy work????

I've almost given up on the FSB, until I can find some info that hints that I'm not wasting my time for something worthwhile might actually exist. I don't like Indicators, so I tried to focus on the actual betting system. The latter can't be designed and optimized in FSB,  so I decided to make my own tester in Excel, where you can implement as many ideas as you want. After playing for days, I came to the concusion that it has a lot to do with the spread. I usually start off with a spread of 0 (so I can see the exact effect of my equity / balance graph while I optimize my betting strategy variables. It's not difficult to create a straight line balance heading for the stars (even if the winning / losing ratio is 50%) but then... I set the spread to 3 (for EURUSD) and bam, the equity graph dances like someone walking over a bed of burning coals. I'm very discouraged at the moment and starting to doubt if there is ANY system that actually works on the long run. Good luck for now, if you find something interesting and worthwhile investigating, please let me know. If you have any betting system ideas, I am willing to test them in my spreadsheet.

Re: Cant make a single strategy work????

It's impossible to foretell the future. What we can do is to get data from the past & analyze them only & try to guess more reliable possibility for the future.

Current market is greatly influenced by the fundamental matter & it disturbs the technical analysis. That's why there are plenty of day-traders waiting for the announcement at certain time & try to scalp.  They don't think about the meaning of the announcement at all. Up or Down only. Because of this, the market price is NOT reflecting the real value, I think. And our automated system cannot catch up the sudden movement at the time.

Nowadays it's difficult to catch the trend by the technical analysis. But what we are doing here is to make such system. I made so many models however i haven't got any promised one so far.

Now I think that closing logic & reverse logic are very important, especially when your strategy doesn't work & market price goes opposite way. We cannot win all the time. So if your strategy collapse in real market, we need to cut loss quickly to save our fund.
I wish FSB & FST can have more function is closing logic, such as trailing stop system or logical group in FST etc.
It'll be nice if we can set the conditions for Add, Reduce, Reverse & Close independently from the opening logic as well...

Re: Cant make a single strategy work????

I think a trend following system is always a good idea, but you need a indicator which indicates markets without a trend and this is the problem.
Do you know how to detect a market without an up or down trend?

Re: Cant make a single strategy work????

Let's discuss trend following system.
Which indicators do you think are useful?

I feel that DeTrend & Laguarre are worth studying rather than normal MA or stochastic.

Re: Cant make a single strategy work????

Alem Bert wrote:

... I came to the concusion that it has a lot to do with the spread. I usually start off with a spread of 0 (so I can see the exact effect of my equity / balance graph while I optimize my betting strategy variables. It's not difficult to create a straight line balance heading for the stars (even if the winning / losing ratio is 50%) but then... I set the spread to 3 (for EURUSD) and bam, the equity graph dances like someone walking over a bed of burning coals ...

I have similar results as you. This is what I found with a random strategy, with fixed limits and stops, run over a data set of prices 10,000 times:
With 0 spread, and equal stop and limit sizes, if the win ratio is 50%, the ending balance average is usually very close to the starting balance. Changing the spread to 1, the ending balance average goes down some; 2, it goes down more; and so on as the spread increases.
The pattern seemed to be the average ending balance would go down somewhere close to the spread times the average number of trades (eg spread = 2, average number of trades = 67 --> average ending balance would go down by around 134). Further, setting the spread to a negative number would increase the average ending balance, again by the same formula of spread times average number of trades.
The conclusion I drew is that my disadvantage on a cross is directly related to the spread size. And, the win ratio itself is not important, but if it exceeds the breakeven ratio, and the breakeven ratio should include the spread.

9

Re: Cant make a single strategy work????

Alem Bert wrote:

I don't like Indicators, so I tried to focus on the actual betting system.

I've been a patterned day trader on the US equities exchange for about a year now.  I'm young so I can't sit here by myself all day and trade alone (Jhovany is one of my buddies), so I often sit in stock chatrooms filled with new traders who don't have $25k for a ptd account and only trade pennies.  I have watched countless people fail over and over again and I must say that most of them refused to learn technical analysis.  They claim that indicators don't work although they couldn't even tell me what MACD stands for, so I watched them all make terrible trades over and over and over again.

I've recently started to switch from intraday swinging large caps to trading Forex and the two are very much alike.  I use a combination of two indicators, moving averages, and 3 different intraday timed charts to make my trades and as Jhovany will tell you, it is deadly accurate.  Often times when I call out a trade over the mic in those chat rooms, I get to hear the newbies going "how did you know to short then???" when it was entirely obvious that the entire stock was overbought and ready to come down, if you knew the indicators.

The difficulty with putting a strategy such as mine into an automated system, is that I do not trade by the bar, and I do not trade using one time frame of reference; more or so human intuition is still required.  I've devised a pretty solid strategy using the 1 minute time frame that seems to be working out and after another week I'll be ready to go live with it, but it looks to take advantage of the bigger pictures rather than small scalp trades.

The issue with doing this (or at least to me about posting successful trading ideas) is that when the systems begin to be used by many people, it sends the same signals to all the people using it at the same time which then all sends market orders.  The integrity of the entry is compromised.  This is what I believe a lot of these bots and EA's for sale on the market look to take advantage of.  It wouldn't surprise me if they are selling a moderately successful Expert Adviser while running a very similar version of it with an entry right before the one they put up for sale.  This guarantees them buying or selling volume for every trade they wish to make, leaving them in profits every time no matter what and the people who bought the program bewildered as to why it entered in 5-10 pips above the signal price.  Genius if you ask me, I wouldn't mind taking advantage of that situation myself. lol

10

Re: Cant make a single strategy work????

NicholasK wrote:

Let's discuss trend following system.
Which indicators do you think are useful?

I feel that DeTrend & Laguarre are worth studying rather than normal MA or stochastic.

Moving averages and stochastics in my opinion are GREAT and very useful for live day trading, however I also agree that they are not reliable enough to solely go on for an automated system.

The MACD (Moving Average Convergence Divergence) although it involves two moving averages, would be the most basic trend spotter, when the faster moving average is increasing or decreasing quicker than the slower it identifies the upward or downward trend.  The issue is that the typical (12,26,9) settings are often slow to spot the bottoms and tops or switches, and there are many moments of indecision, double bottoms, or double tops.

The TRIX (Triple Exponential Average) operates much like the MACD although it responds much faster, especially when set to (5,3).

The ADX (Average Directional Index) spots the trend with an ADX line indicating the strength of the trend, and a green DI (Position Directional Indicator) and a red DI (Negative Directional Indicator).  Obviously when the green is over the red the trend is up and when the red is over the green the trend is down.  When the two lines are diverging the trend is increasing in strength and the ADX line goes higher, and when the lines are converging the trend is decreasing in strength and the ADX line goes lower.  The ADX however would be extremely hard to take advantage of for an automated system though, imo, although if you're inclined to try it out why not.


Honestly, once a trend has been spotted it is probably time for it to reverse so I personally don't believe following "the trends" is the way to go.  What you need to do is spot the bottoms and tops, the reversal points for entry and exit.  This not only maximizes profit but also insures that you aren't getting into a trend late, such as Jhovany's problems were.  It seems like after his program confirmed the trend it enters him in, although the trend is stale and fresh for a reversal or pullback.

The way I'm devising my system is off of how I make decisions intraday with my live trading.  The only difference is I'm only taking positions when the currency pair is absolutely oversold or absolutely overbought.  When something is absolutely oversold or overbought, the only way it can continue in the same direction is if it goes flat for a period of time and allows all the indicators to drop back down to oversold or back up to overbought, in which case your losses become minimal or even can exit with a small profit.

11

Re: Cant make a single strategy work????

If anyone's familiar with a "pinch play," it is a classic example of spotting a bottom or reverse pinch for a top.  You put the PPO (Percentage Price Oscillator) over top of the ADX, and when the PPO comes down to the bottom and "pinches" with a high ADX line in a negative trend, it spots the bottom and shows that it's time for a reversal.  Adding a full stochastic overlay to the PPO adds a 3rd indicator of absolute oversold to the table.  These are typically safe trades because they often point to a direct spot where the price must move back up or remain flat for some period of time, unless outside influence such as news or an event occurs.

Indicators of overbought/oversold are:
RSI (Relative Strength Index)
CCI (Commodity Channel Index)
Williams %
Stochastics (although not useful for automated trading imo)

Re: Cant make a single strategy work????

amanda24 wrote:

I've spent almost a thousand dollars and none of the strategies i got work, at least not long term. sad

You posted invisible casino link in your post, go somewhere else you spammer!

13

Re: Cant make a single strategy work????

acharyadas wrote:
amanda24 wrote:

I've spent almost a thousand dollars and none of the strategies i got work, at least not long term. sad

You posted invisible casino link in your post, go somewhere else you spammer!

hahahha!  wow!  maybe cause his trades are bets not trades wink

Re: Cant make a single strategy work????

jhovany ive been playing around with FSB myself and found exactly the same problems as you. i have developed a relatively simple scalping strategy to use on 5 min charts, to take advantage of the markets during after market hours when the market is usually in a very small range.

my initial SL was set to 20 pips and TP as 10 pips. with 0 charges the strategy performed spectacularly but as i input the charges of a 2pip spread... BAM! my balance plummets. so i played around for hours trying different combinations until it hit me, if i do as you say an simply take the reverse position as the signal given by the strategy and flip my SL and TP values it should be profitable. and just as you found its not as simple as reversing the entry criteria. so i exported my failed strategy to excel and manually reversed the positions i.e. where i went long and my SL was hit for a 20 pip loss, i went short with a TP of 20 pips and SL of 10 pips. the results were very good but i am wondering if my logic process was correct. if anyone is interested let me know an i can email you the excel sheet showing the original and the reversed positions to look over it an let me know what they think? i cant seem to attach it to this message.

thanks looking forward to hearing a response from someone!

Re: Cant make a single strategy work????

Jhovany,
In real live trading doesn't matter how you get results, you cheating is permitted. you can fool, market brokers, whatever/whoever, the most important is to get profit.
There is not right or wrong way.
If you can do what you just said then, you can use account copier on MT4 and just reverse TRADES,
not reverse STRATAGY, but only TRADES, and indeed sometimes it works, if spread is not an issue.
So trading long time frames. Is the key, longing for more than 50pips a trade.

Forex market has patents, the same as everything has, it has seasons the same as year has, the trick is to see them and exploit them for your benefit. This is what all backtesting about.
There is not guaranty that it will work, but there is a high chance that if it works on backtest it will work on live, if we use all steps listed in Footon list of strategies evaluation see topic #11 [by Footon] : http://forexsb.com/forum/topic/1651/sharpe-ratio/

So cheer up and just continue your resource, it is difficult task but it is achievable.





jhovany wrote:

No matter what I do....my buy and sell signals are completely backwards.....I will be 100% honest...If I were to start the automatic execution and sit all day infront of the computer and do EXACTLY the opposite of what the system is trying to execute....I will make money! And simply just reversing the strategy wont work! It still doesnt execute a "good" order! No matter if it is trying to execute a long or a sell.....it drops IMMEDIATELY from where it executes....so my logic is manually override the system and we will profit!

Re: Cant make a single strategy work????

mikhail86 wrote:

... if i do as you say an simply take the reverse position as the signal given by the strategy and flip my SL and TP values it should be profitable. and just as you found its not as simple as reversing the entry criteria. ...

If you haven't tried this, it sounds like this FSB feature can reverse the signals for you, if you go to Tools > Custom Indicators > Export the Strategy as a Custom Indicator. Save it, and open the indicator on a new strategy. It's got a check box at the bottom for "Reversed Signals".

I've found the same as you, that flipping a strategy does not flip the balance/equity chart. I think this could happen if, from a strategy's entry points, the price is more likely to vary up and down by the stop amount, instead of making a straight move of the limit amount. In this case, it doesn't matter if you are long or short, if it wavers up and down by the stop amount both long and short would lose.

Re: Cant make a single strategy work????

Krog,
you posted right after I posted the answer, please read my comments.
I hope it makes sense.
Do not reverse strategy, reverse signals. and it is not the same from my experience.


krog wrote:
mikhail86 wrote:

... if i do as you say an simply take the reverse position as the signal given by the strategy and flip my SL and TP values it should be profitable. and just as you found its not as simple as reversing the entry criteria. ...

If you haven't tried this, it sounds like this FSB feature can reverse the signals for you, if you go to Tools > Custom Indicators > Export the Strategy as a Custom Indicator. Save it, and open the indicator on a new strategy. It's got a check box at the bottom for "Reversed Signals".

I've found the same as you, that flipping a strategy does not flip the balance/equity chart. I think this could happen if, from a strategy's entry points, the price is more likely to vary up and down by the stop amount, instead of making a straight move of the limit amount. In this case, it doesn't matter if you are long or short, if it wavers up and down by the stop amount both long and short would lose.

Re: Cant make a single strategy work????

Alem Bert,
If you don't mind me asking:
how many backtests you have run?
how many stagiest fall under Footon #11 conditions http://forexsb.com/forum/topic/1651/sharpe-ratio/
How many you tested on Demo account?
How many you put on live nano account?

If you can provide this information we might help you.




Alem Bert wrote:

I've almost given up on the FSB, until I can find some info that hints that I'm not wasting my time for something worthwhile might actually exist.

Re: Cant make a single strategy work????

fxwinner,
Number of backtests: Difficult to say the exact number, but I have spent at least 100 hours generating strategies, optimizing, back testing with data from various sources (different time frames and various instruments)

Strategies fall under Footon: All my potentially profitable strategies take all the typical measures of success into account, those that fail even one condition is not even considered for the final back testing phase.

Tested on Demo account: None, I don't waste my time with demo accounts, cos they are not similar to live accounts. If a strategy seems reasonably worthwhile based on accurate data used in back testing, then I go live almost the next day, and test for a couple of weeks. If it does not produce a positive equity curve within a month, I scrap the strategy and go back to the drawing board. I have about 4 live accounts at different brokers and run 1 to 3 strategies at each. Overall, the account balances are slightly above or below my original deposit amounts. Forward testing on demo account takes way too long and the temptation to manually interfere is way too much. If back testing over lots of different time horizons and various instruments prove ok, then it's actually better than forward testing on a demo account.

On live nano account: Don't have nano account, all my accounts (mentioned above) are standard accounts with starting balances of $1000.

No need to help me with "how to evaluate" a profitable strategy. I know how to evaluate a strategy, hence my frustration towards not being able to find one (that meets MY criteria), and so far no one have seemed to be able to produce something reasonable (that works in Live environment for more than a year)

Re: Cant make a single strategy work????

Alem Bert,
Please do not be offended by my questions, I cannot judge your trading experience only by nine posts, thus, I asked these simple questions. Now I have clearer picture what you are going thru.
If you don't like demo trading or don't see use of it you can Use IBFX , they have nono lot 1c/pip.
Thus testing becomes very easy. as we don't risk much then if you are happy with trading you just increase risk per strategy.  At least this is how I do that.
it is not common to share successful strategies on this forum. we probably should create a close section where we can share this experience and work together on end results, however it is not my call, but Popov's. I can only suggest we do that for FSB and our sack.
Let me know your thoughts.

Re: Cant make a single strategy work????

fxwinner,

Regarding the nano lots, yes, it is definitely the better option over a demo account.
However, I have some standard accounts though, and main reason was to get the sign up bonuses, so I play the minimum lots to test some strategies and manage to stay alive long enough to get the required volume to earn my bonuses in cash. So it is a completely "unconventional strategy" I suppose, but at least I have made more money that way than any "normal strategy" have done for me. Especially towards the end of the year, many brokers offer a 100% first deposit bonus. I realize I can not ride this wave forever, but luckily there are so many of them, and they seems to pop up like mushrooms, so I think I can probably go on like this for at least another year (hopefully I will find something more sustainable by then)

Closed section would be nice, but in the meanwhile if you really have the need to or think it would be more beneficial for the members who want to contribute by putting in some real hours, then perhaps one can create a Google Wave for it, and "invite" certain members depending on what they can bring to the table. However, I don't mind sharing all experience on this forum since we all have similar interests and no information regarding successful strategies shared will ever make a difference to what happens in the market. Only the BIG players (like banks and governments) have the "limited" ability to influence the price. FX traders merely "play" against the brokers , and it will take one extremely successful strategy posted on a forum, if thousands of traders implement it exactly to the specifications, many many months if ever, before brokers might change the rules if their profitability becomes endangered. (I know you probably know this, I am just giving my opinion towards the guy that said a good strategy might compromise the whole system)
Anyway, but since the interaction in a forum is very limited, a Google Wave might be ideal for working together on a specific project where each member contributes towards a potentially successful strategy. Only reason I prefer the "invitation" process is that too many people on the same project will actually cause it to not be successful, since then you would have to start managing people, instead of focusing on the project and tasks at hand. I have managed projects for many years, don't need to do another (especially if people are all over the world and most of them don't even know each other)

BTW, did Popov write the FSB ? Just wondering, cos I need to know who to speak to regarding features and future development, cos FSB is cool, but it can be improved in so many ways, so I was wondering if there aren't perhaps an open source version, or a different software package all together, where one has access to the code. I am aware of the "Feature Request" section, it's just that I have so many, that by the time anyone decide to implement them, I could have probably coded it myself already.

Anyway, your thoughts... ?

Re: Cant make a single strategy work????

Alpari does it all the time up to 20% you can get, no need to be a new customer.
it is 20% on a new deposit.
I know what you mean, I used this tick as well, the best way to do is to scalp. Scalping can generate pretty good volume, which one cannot do with "normal" long term trading.

Alem Bert wrote:

Regarding the nano lots, yes, it is definitely the better option over a demo account.
However, I have some standard accounts though, and main reason was to get the sign up bonuses, so I play the minimum lots to test some strategies and manage to stay alive long enough to get the required volume to earn my bonuses in cash. So it is a completely "unconventional strategy" I suppose, but at least I have made more money that way than any "normal strategy" have done for me. Especially towards the end of the year, many brokers offer a 100% first deposit bonus. I realize I can not ride this wave forever, but luckily there are so many of them, and they seems to pop up like mushrooms, so I think I can probably go on like this for at least another year (hopefully I will find something more sustainable by then)

Re: Cant make a single strategy work????

Yea, 20% is ok and might we worth it even in the long run, but those 100% once off bonuses are really nice, the last one I did was with ForexYard December last year. And yes, scalping gets you the volume in no time if you trade currencies, commodities not really worth it.

Please let me know your thoughts on the Google Wave and the open source strategy generator / optimizer if you know of one, or to whom I might be able to discuss possible involvement of future development of FSB, if Popov the original programmer ?

Re: Cant make a single strategy work????

your suggestions are very valuable.
I have noted about Google Wave.
Yes, Popov is the solo author of FSB  alone, and it was his genius idea of FSB itself, it is an unique approach. And this is why I'm here --->> to enhance it. Don't be concerned about silence on the forum, it is temporally until he comes back from his business trip and we'll rock, I agree with all your points here. Please feel free to post your ideas here:  http://forexsb.com/forum/topic/1549/fsb … -features/
Popov will be very active from beginning of September, we will summarize the wish list and implement what the TOP ideas, so there is high chance yours will the on next release.
So please continue be enthusiastic, we will work hard to satisfy your needs as an active trader.

Re: Cant make a single strategy work????

you can post your ideas, and if you would like to be involved the task list will be enable in September.
FST is an open source, however due to the fact that FSB possesses an unique optimisation approach, it won't be open source.


Alem Bert wrote:

Please let me know your thoughts on the Google Wave and the open source strategy generator / optimizer if you know of one, or to whom I might be able to discuss possible involvement of future development of FSB, if Popov the original programmer ?