Mkcky's GBP System. | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

GBP/AUD, GBP/CAD, GBP/CHF and GBP/NZD thread 126 replies

The System Behind the System, by Merlin Jeffries 16 replies

Another great Breakout GBP/JPY System (Dachel's Simple system) 62 replies

Knife system, Vegas 4h system related questions. 4 replies

Would you trade this system? New system #2. 10 replies

Attachments: Mkcky's GBP System.
Exit Attachments

Mkcky's GBP System.

  • #41
  • Oct 17, 2004 1:27am Oct 17, 2004 1:27am
Hola Mario and everybody...

I have been looking for good charts able to see historical data in 1hr... to do the backtest... i talked to Mario via Yahoo Messenger, and the backtest is maded in Daily Charts... i think if we see the results in Daily Charts maybe will be very good but in my opinion, in Daily Charts we can not see if the stoploss was touched or it wasn´t... so, we can imagine than in a good month the result was around +600 pips, but testing this strategy in 1h charts i only got +200 or something like that... so, i only could test 3 months in MarketScope FXCM Fee... today i tried to start again with the backtest but i could open historical data, i sent a email to support from FXCM Charting asking them about if the data of MarketScope Charts is good and reliable, well i attached here only 3 months of my backtesting... i will start again with Feb 2003 when i get able to see the historical data in MarketScope...

2 enero: +122

3 enero: -30 +106

6 enero: -30 -2

7 enero: -30

8 enero: -30 +57

9 enero: +53

10 enero: -30

13 enero: +17

14 enero: -30 -30

15 enero: -30

16 enero: +41

17 enero: -30

20 enero: -30

21 enero: -30 +89

22 enero: -30 -30

23 enero: +66

24 enero: -30

27 enero: -30 -30

28 enero: -30 +66

29 enero: -30

30 enero: -30 +37

31 enero: -30 -30

ENERO 2003: +22

1 septiembre: +53
2 septiembre: -30 -30

3 septiembre: -30

6 septiembre: +30

7 septiembre: -30 +4

8 septiembre: -30 +61

9 septiembre: -30 -30

10 septiembre: -30

13 septiembre: -30 -30

14 septiembre: -30 -30

15 septiembre: +145

16 septiembre: +139

17 septiembre: -30 -30

20 septiembre: +16

21 septiembre: +91

22 septiembre: -30 -8

23 septiembre: -17

24 septiembre: +78

27 septiembre: -30 +14

28 septiembre: -30

29 septiembre: +87

30 septiembre: -30 +68

SEPTIEMBRE 2004 +221

1 octubre: +143
4 octubre: -30 +30

5 octubre: -30 -30

6 octubre: -30 -30

7 octubre: -30

8 octubre: +93

11 octubre: -30

12 octubre: +9

OCTUBRE 2004 +65
  • #42
  • Oct 17, 2004 4:57pm Oct 17, 2004 4:57pm
Sorry, i had a mistake in Juanuary 2003, the last trade was +81 and not -30... i started the backtest again in 30min charts, i got the email from FXCM Support and they told me about MarketScope... the charts and the historical prices are ok, but i could get "blackholes" in NY trading time, so, i will be very careful looking for those mistakes... so, in Juanuary 2003 was +133, today i will post will the backtest of Feb and March...

Thanks
  • #43
  • Oct 17, 2004 9:24pm Oct 17, 2004 9:24pm
So, just so I understand, when I am just about to enter the trade, I go to check what the ATR is CURRENTly right, and then if it is under 100, I do not go in, otherwise I put the trade in right? Does not matter if I use 1hr or Daily charts to check the ATR, because what is important is what the ATR is at now right?

Can someone confirm?
  • #44
  • Oct 17, 2004 10:03pm Oct 17, 2004 10:03pm
Quoting jesse_breaker
So, just so I understand, when I am just about to enter the trade, I go to check what the ATR is CURRENTly right, and then if it is under 100, I do not go in, otherwise I put the trade in right? Does not matter if I use 1hr or Daily charts to check the ATR, because what is important is what the ATR is at now right?

Can someone confirm?
Must be daily ATR.
Regards.
mariovides
  • #45
  • Oct 17, 2004 10:17pm Oct 17, 2004 10:17pm
Quoting arrieta
Sorry, i had a mistake in Juanuary 2003, the last trade was +81 and not -30... i started the backtest again in 30min charts, i got the email from FXCM Support and they told me about MarketScope... the charts and the historical prices are ok, but i could get "blackholes" in NY trading time, so, i will be very careful looking for those mistakes... so, in Juanuary 2003 was +133, today i will post will the backtest of Feb and March...

Thanks
As talked with arrieta by yahoo messenger results could vary for many factors in a backtest, one difference could be the open hour for trades, i checked some days for september in the 30 min charts that never touched one stop loss when i open trades at 24:00 EST (hour of my backtest) and touched a stop if opened at 23:00 EST or viceversa, then the backtest is specific in each case, the important is that with so many backtest we have broadest point of view of the system and better understand, i invite other forum colleagues to confirm our results.
Regards
mariovides
  • #46
  • Oct 18, 2004 12:29am Oct 18, 2004 12:29am
Would it be possible to post a chart showing which bar of the daily you look at?

Or perhaps can you confirm if what I circled in red is what I should be looking at when I go to netdania and look at the daily charts with atr14?



Quoting mariovides
Must be daily ATR.
Regards.
mariovides
Attached Image (click to enlarge)
Click to Enlarge Name: daily_gbpusd.JPG Size: 36 KB
  • #47
  • Oct 18, 2004 8:04pm Oct 18, 2004 8:04pm
Quoting jesse_breaker
Would it be possible to post a chart showing which bar of the daily you look at?

Or perhaps can you confirm if what I circled in red is what I should be looking at when I go to netdania and look at the daily charts with atr14?
Yes in the chart you sent, the below value is the ATR (112).
  • #48
  • Oct 28, 2004 3:14pm Oct 28, 2004 3:14pm
Mkcky, On the first part of this thread, you said that this system works Sunday-Thursday @ 11pm EST. You said that you close all open positions and open new ones.

What about Thursday's position? If we do this @ 11pm Thursday, the market is closed by the time 11pm Friday roles around.

Should we open any new trades Sunday-Wed, and close any open trades Monday-Thursday?

not clear on this point of the system. Any comments would be helpful
  • #49
  • Oct 28, 2004 11:52pm Oct 28, 2004 11:52pm
Quoting Mkcky
Hello Everyone,

Here is backtesting results sent my mariovides its zipped excel file.


Here is brief summary

Year 2002 Total profit 1091 Pips
Year 2003 Total profit 5259 Pips
Year 2004 (uptill Sept) Total profit 9936 Pips

Year 2004 results are fantastic because ATR was very high in begining of Year. There is no loosing month in 2003 and 2004 . In year 2002 ATR was below 100 system had negative performance for four months. So I would suggest not to trade if ATR falls below 100.

Thanks mariovides for the Backtesting.

Mkcky
When you refer to the ATR what time frame are you using to look at the ATR and what values do you have it set to, as well as on what software do you use?

-Fleetwood
  • #50
  • Oct 29, 2004 2:33pm Oct 29, 2004 2:33pm
I gather the trades need to be placed before the opening of the Europesn session and the time quoted is 11pm EST. Could I place trades earlier- lets say about 8pm EST?
Any comments?
thanks.
  • #51
  • Oct 29, 2004 3:13pm Oct 29, 2004 3:13pm
Quoting snook13
I gather the trades need to be placed before the opening of the Europesn session and the time quoted is 11pm EST. Could I place trades earlier- lets say about 8pm EST?
Any comments?
thanks.
Normally market is very slow between 4 PM to 12:00 PM and does not hit 30 pips channel before begining of Europen Session. So you can place trade between this time. so you can place the trade at 8 PM just stick to same time everyday.

Mkcky
Lets Make Money!!!
  • #52
  • Edited 6:47pm Oct 29, 2004 4:21pm | Edited 6:47pm
Any one trading this system? Can someone post the results of OCT.

My demo resuts were -15 pips for OCT. The GBP Reviersal System gave a profit of +291 Pips for Oct. I will post that in seprate Thread.

Mkcky
Lets Make Money!!!
  • #53
  • Oct 29, 2004 5:14pm Oct 29, 2004 5:14pm
Quoting Mkcky
Any one trading this system? Can someone post the results of OCT.

Mkcky
Did you see my question about what value you have the ATR set to, and on what time frame you are looking at it?

- Fleetwood
  • #54
  • Oct 29, 2004 6:02pm Oct 29, 2004 6:02pm
Quoting Fleetwood
Did you see my question about what value you have the ATR set to, and on what time frame you are looking at it?

- Fleetwood
hmm I didnt, but let me answer your question right away.

ATR 14 on Daily time frame.

Mkcky
Lets Make Money!!!
  • #55
  • Oct 30, 2004 1:02am Oct 30, 2004 1:02am
Quoting firedave
Hai Mario,

First I like to thank you for you backtesting, such a promising system. I just like to add some condition to the backtest. This condition apply on descending / ascending bar with 30 pips above and below the opening price.

What happend if price move forth and back the opening price and hit both entry but also hit both SL ?

Example : Opening price 1.8000, price go to 1.8050 hit first entry. Then go down to 1.7950 hit first SL and second entry. Then it goes up again to 1.8000 where it hits second SL. After that all the way down to close at 1.7900. Vice versa for ascending bar.

This example will give -60 pips rather than 1.8000 - 1.7900 - .0060 = 40 pips profit.

So for this kind of strategy, we can't use only daily chart, but also need a hourly chart at least.

IMHO

I do not mean to rag on Mariovides or insult him in any way, and actually express my appreciation for his taking the time to back test and share the results with us. However, the results that he shows are completely false, because they were done using daily charts. Firedave is 100% correct. If you think about it, if you try to back test this system on a daily chart you have absolutely no way of knowing what the movement is WITHIN the daily bar you are looking at, and it's not possible to tell whether trades within that daily bar are stopped out or if they make it to the closing trade time with a profit. I did some back testing of my own and the results are not even close to as good as what was posted by Mariovides using daily charts. 221 for Setpember is about right. This system does seem to have some merrit however, even though the results are not as good as Mariovides back testing. It definitely did not make anywhere near 9900 pips this year, not even close. I will find out how well it did do so far this year.


I actually got 353 pips for September, but arrived at this number by looking at the system a little bit differently than Mkcky described. I am looking looking at exiting trades at 5 pm, instead of 11 PM the next day. The reason I am doing this is, is to avoid the rollever charges from the broker at the close of each day. Rollover charges will kill you long term, so it will perform better without incurring these rollover charges, and these also are not calculated into the back testing anyone is doing, I'm sure. For the couple months I've looked at it, you normally get better exit levels with higher profit exiting at 5 instead of 11 PM, as you can see from my results for September of 353 pips. October was NOT a good month for this sytem, but not terrible, because it didn't result in a loss, but there were an enormous amount of individual losses. I show 33 trades but a gain of only 82 pips, which is not good at all for so many trades. There were many days where both the buy and sell orders were triggered and both resulted in a loss. This system only makes money on days where the market moves a lot in a single direction, or only two directions, without countertrend spikes that are too big. If the market bounces around in both directions this system will normally lose 30 pips twice, and can do that for days in a row, so pretty large drawdowns are in order inbetween the days that the market breaks out and has some big moves. The good news is that the days with large moves tend to overshadow the groups of losses that occur inbetween.


This system is extremely simple and just based on price, so it's possible to see exactly what the results would be by back testing with I feel a very, very small margin of error, so if we can back test for a couple years, I think it would be much better than demo trading for a few months, but let's do it correctly. A daily chart is useless for back testing this system, we need to use a time frame somwhere around an hourly. Obviously it would be the easiest and quickest if someone can do the programming and turn this into a TradeStation strategy and TS can give the results of a back test in seconds, the same as would take us days and days to do it manually.


- Fleetwood

  • #56
  • Oct 30, 2004 8:12am Oct 30, 2004 8:12am
[quote=Fleetwood]

I do not mean to rag on Mariovides or insult him in any way, and actually express my appreciation for his taking the time to back test and share the results with us. However, the results that he shows are completely false, because they were done using daily charts. Firedave is 100% correct. If you think about it, if you try to back test this system on a daily chart you have absolutely no way of knowing what the movement is WITHIN the daily bar you are looking at, and it's not possible to tell whether trades within that daily bar are stopped out or if they make it to the closing trade time with a profit. I did some back testing of my own and the results are not even close to as good as what was posted by Mariovides using daily charts. 221 for Setpember is about right. This system does seem to have some merrit however, even though the results are not as good as Mariovides back testing. It definitely did not make anywhere near 9900 pips this year, not even close. I will find out how well it did do so far this year.



Dear Feetwood.
Your observations are right, the results of my backtest using daily bars are not reflecting the real situation, i encouraged all of you to do it in the right way using at least 1 hour charts or 30 min charts, but at the moment of reply your message an idea comes to my mind: if backtesting with daily bars the results was so excellent its reflects that probably using a modification of MCKY system could be great to obtain the results that my backtest reflected i think in the following modification:

For months with ATR +100, opening trades at the same level of 30 pips up or down the 11:00 p.m or 12:00 a.m. EST, but with wider stop losses and opening only with the trend bias using stop buy or sell orders (i only know about pivot points method for trend bias forecast, and support-resistance trendlines, some of you should help us with other technique) and with the specific money management rules according with the range of the stop loss used.
The above is and idea for not make over trading and only open high probability trades.
What do you think Junior and Senior traders?
Regards
mariovides
  • #57
  • Edited 9:55am Oct 30, 2004 9:51am | Edited 9:55am
[quote=mariovides]
Quoting Fleetwood

I do not mean to rag on Mariovides or insult him in any way, and actually express my appreciation for his taking the time to back test and share the results with us. However, the results that he shows are completely false, because they were done using daily charts. Firedave is 100% correct. If you think about it, if you try to back test this system on a daily chart you have absolutely no way of knowing what the movement is WITHIN the daily bar you are looking at, and it's not possible to tell whether trades within that daily bar are stopped out or if they make it to the closing trade time with a profit. I did some back testing of my own and the results are not even close to as good as what was posted by Mariovides using daily charts. 221 for Setpember is about right. This system does seem to have some merrit however, even though the results are not as good as Mariovides back testing. It definitely did not make anywhere near 9900 pips this year, not even close. I will find out how well it did do so far this year.









Dear Feetwood.
Your observations are right, the results of my backtest using daily bars are not reflecting the real situation, i encouraged all of you to do it in the right way using at least 1 hour charts or 30 min charts, but at the moment of reply your message an idea comes to my mind: if backtesting with daily bars the results was so excellent its reflects that probably using a modification of MCKY system could be great to obtain the results that my backtest reflected i think in the following modification:

For months with ATR +100, opening trades at the same level of 30 pips up or down the 11:00 p.m or 12:00 a.m. EST, but with wider stop losses and opening only with the trend bias using stop buy or sell orders (i only know about pivot points method for trend bias forecast, and support-resistance trendlines, some of you should help us with other technique) and with the specific money management rules according with the range of the stop loss used.
The above is and idea for not make over trading and only open high probability trades.
What do you think Junior and Senior traders?
Regards
mariovides
I feel that the reason this strategy works is because it is so incredibly simple. There is always a temptation to try to put all these filters and stipulations on a system to try to improve it's historical results, but what this generally does is over fit the system to past data and kill the future and long term performance of the system. Enormous stop losses would have to be used to try to avoid the stop outs and achieve similar results as back testing with a daily chart, and it would be very difficult to gage what size is best. And trying to base the trades only on the trend is much too subjective, and I would be willing to bet any method to try to achieve this would fail and would kill the results, causing you to miss many good trades. If one could reliably predict what the upcoming trend will be, or accurately forecast whether or not the current trend could continue one wouldn't even need this system to begin with, because they would already know the next market movement.

Remember that some of the greatest success traders have had LONG TERM were with systems that were almost deceivingly simple. The more complicated a system the worse it's results will be long term, and the less reliably it will work in the future compared to past results.

- Fleetwood
  • #58
  • Oct 30, 2004 11:19am Oct 30, 2004 11:19am
OK, the clocks are going to change on Sunday. How does this change when we enter and close the trades? Is it still 8pm or is it now 7pm?
  • #59
  • Oct 30, 2004 6:30pm Oct 30, 2004 6:30pm
Quoting Fleetwood
I feel that the reason this strategy ... works is because it is so incredibly simple. There is always a temptation to try to put all these filters and stipulations on a system to try to improve it's historical results, but what this generally does is over fit the system to past data and kill the future and long term performance of the system. Enormous stop losses would have to be used to try to avoid the stop outs and achieve similar results as back testing with a daily chart, and it would be very difficult to gage what size is best. And trying to base the trades only on the trend is much too subjective, and I would be willing to bet any method to try to achieve this would fail and would kill the results, causing you to miss many good trades. If one could reliably predict what the upcoming trend will be, or accurately forecast whether or not the current trend could continue one wouldn't even need this system to begin with, because they would already know the next market movement.

Remember that some of the greatest success traders have had LONG TERM were with systems that were almost deceivingly simple. The more complicated a system the worse it's results will be long term, and the less reliably it will w... ork in the future compared to past results.

- Fleetwood
Then, we only must confirm if really "works", i encourage everyone to make a manual backtest using 30 min charts, or software backtest, some members start the backtest and never ended, we need to see the real results.
  • #60
  • Nov 1, 2004 3:36am Nov 1, 2004 3:36am
11:00 PM EST could you tell me please what time that by GMT !

And also did you backtest it on GBP/JPY, CHF, EUR/JPY?
Men are born to succeed, not fail !!!
0 traders viewing now