1 (edited by mentosan 2012-08-31 14:09:58)

Topic: is anyone making money with FSB&FST?

I have 2 years of looking for good strategies but still not so profitable. I'm not loosing anymore but I cannot say that I'm making money. What about you guys?

Re: is anyone making money with FSB&FST?

Hi,
I think it is not so easy to make money.
I have experience of one year.

At the moment I have a portfolio (not generated with fst) of 15 strategies.
The testperiod is two month. Some strategies produces money, but only one or two percent per month.

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

I'm testing also 6 strategies. Today for example I loose 120pips but I'm happy because the closing point is the same like in fsb. In fact, this was the problem with the 200 strategies what I tested in the past, the different close points between fsb and fst.

Anyway, I think the people are making money with forexsb because a lot are donating and this is an effect of the profit. I will make also the same, if I will have profit.

Re: is anyone making money with FSB&FST?

Hi mentosan,
what are the reasons that the 200 tested strategies have a different close points?
Are the indicators wrong? Or some settings?
What do you think?

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

for some strategies, the indicator was wrong, I opened topics about that, but at the end was no solution. I tried to optimize the strategies with that indicator but no good results. maybe it was also because of the FST. Anyway, the  different close point was the biggest problem for me. Now, I choose only the strategies on very big timeframes and the different closing points problem is not so interesting anymore for me because I can follow it manually.

Question: has anyone success with low timeframe stragies? (1min, 5min, 15 min)

Re: is anyone making money with FSB&FST?

mentosan wrote:

Question: has anyone success with low timeframe stragies? (1min, 5min, 15 min)

I have some strategies that work on the EURUSD 15min but I have not been able to make anything work below that. 30min works best but the lower the time frame the more potential trades.

Re: is anyone making money with FSB&FST?

Hi,
If we want to get the same closepoint, we need the real tickdata in the FSB
I think. But it is not possible at the moment in the FSB.

We can import Dukascopy tickdata. But at Dukascopy you don´t can use metatrader. They don´t support Metatrader.

The other problem is at dukascopy is, the have only a maximum leverage of 100.

You can use Oanda, but they have only leverage of 40. The other problem is that the dickdata has
a delay of 2 month.


If you use a different metatrader brocker, they don´t give you the tickdata.
You can write a small skript and save tickdata in metatrader, but you don´t can import this data into FSB.

The result is, you don´t get exactly the same closepoints at the moment.
thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

I agree that lower timeframes are complicated.

In my experience the 1h is the best timeframe. Only a few strategies can make money, approximately 1-2 of 10 backtested strategies will show positive forward testing results.

SpiderMan wrote:
mentosan wrote:

Question: has anyone success with low timeframe stragies? (1min, 5min, 15 min)

I have some strategies that work on the EURUSD 15min but I have not been able to make anything work below that. 30min works best but the lower the time frame the more potential trades.

Re: is anyone making money with FSB&FST?

zenoni wrote:

I agree that lower timeframes are complicated.

In my experience the 1h is the best timeframe. Only a few strategies can make money, approximately 1-2 of 10 backtested strategies will show positive forward testing results.

SpiderMan wrote:
mentosan wrote:

Question: has anyone success with low timeframe stragies? (1min, 5min, 15 min)

I have some strategies that work on the EURUSD 15min but I have not been able to make anything work below that. 30min works best but the lower the time frame the more potential trades.

I have couple of strategies on 1m vs. 5m. These are working well in bcktesting as well as in fwd test.

http://www.myfxbook.com/members/togr/experimental/305682 use magic to filter particular strategy (333,336,337,338 and 334)

I am to coward/smart to go live yet.

Re: is anyone making money with FSB&FST?

Hi togr,
1m_120620_opt looks good.

But the Broker FX_Primus has a very good Spread in EURUSD (0.3 Pips)

With what data you have calculated the EA?

Do you used 1minute-data from FX_Primus?
or
RealTickdata from dukascopy?
for calculation.

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi togr,
1m_120620_opt looks good.

But the Broker FX_Primus has a very good Spread in EURUSD (0.3 Pips)

With what data you have calculated the EA?

Do you used 1minute-data from FX_Primus?
or
RealTickdata from dukascopy?
for calculation.

thomas

Hi Thomas,

FXprimus has very low spread but it has $10 comission per lot opening.
I used Dukascopy (Jforex) tick data import to FSB

Re: is anyone making money with FSB&FST?

After how much time do you optimize these 1 min strategies?

Re: is anyone making money with FSB&FST?

Hi togr,
how long was your timeperiod of M1 data for your strategy?
Do you use 6 months of M1 data for your strategy? or less?

Made you a forwardtest?

How do your prevent curve fitting?

All my found strategies are curve fitted? It is a problem in FST. There is no walk forward test implemented
If I calculate a strategy too long, my systems are curve fitted.

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

mentosan wrote:

After how much time do you optimize these 1 min strategies?

I am not sure i understand what ...how much time... means.

But i spent several months by it. At the end i run optimization over 720 000 bars.

And them i tested period of 3 months for last 4 years since 1st JAN 2008.

The strategy with best result in FSB actually performed the worst in fwd test wink

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi togr,
how long was your timeperiod of M1 data for your strategy?
Do you use 6 months of M1 data for your strategy? or less?

Made you a forwardtest?

How do your prevent curve fitting?

All my found strategies are curve fitted? It is a problem in FST. There is no walk forward test implemented
If I calculate a strategy too long, my systems are curve fitted.

thomas

Hi Thomas,

i used 720 000 bars which is equal to 2 years i think.

I then tested much smaller periods like 1year, 6 months, 3 months to see if it fails in any period.

It did not wink

Re: is anyone making money with FSB&FST?

Hi togr,
thanks for the info.

Before this posting I thought I is not possible to find strategies in Timeframe M1 and M5.
I looked at your myfxbook statistics, I think it is possible. But on real accounts you have slippage. Only the test on a real account will give you an answer if the strategies work fine.

720000 bars is a big amount of bars. I used only 50000 bars in the past.

It is mass of work to find good strategies in M1 and M5 timeframe.
thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

Hi togr,
you said

"But i spent several months by it."

On what part of your work has you spent the most time?

Was this the testingtime on a demoaccount?
Or are some features on FSB missing?

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi togr,
thanks for the info.

Before this posting I thought I is not possible to find strategies in Timeframe M1 and M5.
I looked at your myfxbook statistics, I think it is possible. But on real accounts you have slippage. Only the test on a real account will give you an answer if the strategies work fine.

720000 bars is a big amount of bars. I used only 50000 bars in the past.

It is mass of work to find good strategies in M1 and M5 timeframe.
thomas

Thanks,

BTW i am running many strategies on 1 MT4 isntance/account to save system resources and later i analyse the results by magic number. It is not always easy to see performance of single strategy unless you use custom analyses or excel wink

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi togr,
you said

"But i spent several months by it."

On what part of your work has you spent the most time?

Was this the testingtime on a demoaccount?
Or are some features on FSB missing?

thomas

Well I am continuously creating new strategies in FSB and optimizing them , that's where i spent most of the time. Forward testing does not require so much attention from my side except the case when my VPS got restarted.

Thanks god and Mr. Popov and developers for autostart.bat wink

Re: is anyone making money with FSB&FST?

Hi togr,
how many computers do you use for your calculation of your strategy?
I have a AMD 8core system with 16GB memory.

It take a very long time to find a strategy on M5.
I used a datahorizont of one year (80000 Bars) searching a M5 strategy, but FSB don´t find a strategy in 5 hours.

Ok I used spread 2,6 and slippage of 1.
It is a little bit high for a M5 strategy.

The custom analysis of myfxbook is a fine feature. I use this tool too.
A automatic tool that take all the information from myfxbook and compare the results with the used strategies will be fine.

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi togr,
how many computers do you use for your calculation of your strategy?
I have a AMD 8core system with 16GB memory.

You may find that only one core is actually working on your strategy.

Re: is anyone making money with FSB&FST?

Hi SpiderMan,
I have 4 identical copys of FSB in different directorys. I can start FSB 4 times.

If FSB uses only one processor. The tool is very fast.
How much can it be if FSB uses 8 cores :-)

thomas

http://www.tradessystem.com/

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi togr,
how many computers do you use for your calculation of your strategy?
I have a AMD 8core system with 16GB memory.

It take a very long time to find a strategy on M5.
I used a datahorizont of one year (80000 Bars) searching a M5 strategy, but FSB don´t find a strategy in 5 hours.

Ok I used spread 2,6 and slippage of 1.
It is a little bit high for a M5 strategy.

The custom analysis of myfxbook is a fine feature. I use this tool too.
A automatic tool that take all the information from myfxbook and compare the results with the used strategies will be fine.

I am actually running 3 VPS and my computer at home. For this lower timeframes you need a lot of RAM for backtesting and optimizing but all my machines have 4gb at most. So I am thinking to buy IBM server with at least 16 GB. But i am not making any real money so far so this must wait wink
thomas

Re: is anyone making money with FSB&FST?

tnickel wrote:

Hi SpiderMan,
I have 4 identical copys of FSB in different directorys. I can start FSB 4 times.

If FSB uses only one processor. The tool is very fast.
How much can it be if FSB uses 8 cores :-)

thomas

Or to use graphics card for calculations...

Re: is anyone making money with FSB&FST?

togr wrote:
mentosan wrote:

After how much time do you optimize these 1 min strategies?

I am not sure i understand what ...how much time... means.

But i spent several months by it. At the end i run optimization over 720 000 bars.

And them i tested period of 3 months for last 4 years since 1st JAN 2008.

The strategy with best result in FSB actually performed the worst in fwd test wink

Hi togr,
"The strategy with best result in FSB actually performed the worst in fwd test"

It is possible that this best strategies are overoptimized. This can be the reason that the best strategies in FSB performed the worst in fwd test.

thomas

http://www.tradessystem.com/