Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Attachments: Daily95pips
Exit Attachments

Daily95pips

Trading doesn't start for IBFX on Sunday until 6pm EST. The EA reacts strange for me on Sundays opening double pending orders (12 for each pair).

I have changed my strategy for testing as follows:

Demo 1: IBFX 22-17 (19 pairs)
Demo 2: IBFX 13-17 (19 pairs)
Demo 3: IBFX 08-17 (19 pairs)

Demo 4: NF 00-17 (19 pairs)
Demo 5: NF 15-17 (19 pairs)
Demo 6: NF 10-17 (19 pairs)

22-17 is start hour of 22GMT (5pm EST) looking back 17 hours.
13-17 is start hour of 13GMT (NY open) looking back 17 hours.
08-17 is start hour of 08GMT (London open) looking back 17 hours.

00-17 is start hour of 22GMT (5pm EST) looking back 17 hours.
15-17 is start hour of 13GMT (NY open) looking back 17 hours.
10-17 is start hour of 08GMT (London open) looking back 17 hours.

North Finance is 2 hours ahead of the IBFX accounts as far as time goes so each are set for 2 hours more than the IBFX demos.

This seems like a good test and I decided to leave everything for 17 hour look backs because that is enough to just slightly overlap a previous market open and is not a full day. My reasoning is to see which pairs do best under which time frames and see if IBFX does better than NF or vice versus.

I restarted my testing because last week was a mess with the holiday and I noticed that NF was running during Thanksgiving? IBFX shut down for a good portion of the day which just messed things up. Also, IBFX doesn't start until 6pm EST whereas my NF demos all started ticking well before 5pm EST. And then there is the issue with IBFX EA opening double orders on Sundays when the market opens. I think it best to start the IBFX demos just after 6pm EST on Sundays and run all week long.

I am also running live on the 3 pairs Mikelath is running on my IBFX account. The situation was the same opening double pending orders on my live account which was scary. I only have $72 left in that account and am trading 0.01 fixed lots. I almost closed it a few weeks back but left a minimal amount just in case. Once I found this thread I was inspired to think I might have a chance again. It will be a good test to see what can happen starting with $72 dollars.

Last nights live account results...
USD/JPY +15 (buy at 108.61)
EUR/JPY +15 (buy at 161.13)
EUR/USD not triggered yet (buy at 1.4943 / sell at 1.4808)

This is after manually closing all the double orders the EA opened 6 pending orders for all 3 pairs again.
I noticed on the NF 15-17 that the NZD/USD only placed buy stops.

The NF 10-17 only placed sell stops on the AUD/USD and only buy stops on the USD/CAD.

I also noticed that the prices are off between brokers which can be expected but some are a good distance from each other.

IBFX

AUD/CAD buy 0.8722 / sell 0.8649

NF

AUD/CAD buy 0.8724 / sell 0.8654


IBFX

AUD/USD buy 0.8877 / sell 0.8740

NF

AUD/USD buy 0.8876 / sell 0.8761

Quoting mtaboneweb
Last nights live account results...
USD/JPY +15 (buy at 108.61)
EUR/JPY +15 (buy at 161.13)
EUR/USD not triggered yet (buy at 1.4943 / sell at 1.4808)

This is after manually closing all the double orders the EA opened 6 pending orders for all 3 pairs again.
For USDJPY I see how you made 15 pips one one of the trades.
But at approximately 2100 EST you would have been stopped out
on the other two lots at 108.31. Your net would be -15, not +15,
or you manually intervened?
Quoting Longboard
For USDJPY I see how you made 15 pips one one of the trades.
But at approximately 2100 EST you would have been stopped out
on the other two lots at 108.31. Your net would be -15, not +15,
or you manually intervened?
The EA moves SL to break even once 15 pips is hit on the first order. The 30TP and 50TP orders were stopped out at break even. I only manually interfered to close all the initial pending orders which were doubled (6 buys and 6 sells for each pair). The EA opened pending orders (3 buy stops and 3 sell stops for each pair) afterwards and I let the EA handle everything else.

My USD/JPY took its profit at 108.76 and my EUR/JPY took profit at 161.28.

As I'm typing this my EUR/USD orders just expired. So I am up 30 pips today on my live account. That's 0.28 cents since both pairs are not worth $1 a pip.
OK Thanks.
Here are the results for Monday on the 22-17 demos. Each of them have EUR/GBP trades running negative right now and I have a feeling this pair will lag since it doesn't move much. So far it looks like the IBFX demo is doing better but only time will tell. Also, the IBFX demo achieved the same 15 pips on the USD/JPY and EUR/JPY as my live account. The EUR/USD was not triggered for any of the accounts.
Attached Image (click to enlarge)
Click to Enlarge Name: Daily 95 (2007-11-26).jpg Size: 174 KB
I noticed that the NZD/JPY on the NF demo is -45 which should not happen. The statement shows +15, -30, -30 for a total of -45. I can only assume this happened because the SL was not moved up to break even when +15 was achieved. If you look at my recent posts you can see other problems with the NF demos.

When setting up the NF demos they also took forever to create demo accounts whereas IBFX is usually very fast. I am going to go out on a limb and say NF is a little sluggish.
Attached is my latest chart.

If 0 then trade wasn't triggered or hasn't been triggered yet.

Darkest green is the most profitable on a row.
Darkest red is the least profitable on a row.

Take the AUD/JPY row as an example. It has pips for all demos meaning all were triggered. It did the best on the IBFX demo with 22 start hour and 17 look back hours. It did the worst on the first 3 columns/demo accounts. 13-17 (NF) and 22-17 (NF) are a lighter green and were positive but not as good as 22-17 (IBFX).

Column J (Pair Totals) colors are the same idea. This column is a total of each row/pair. The GBP/CHF is the best and darkest green for the column.

Row 22 (Account Totals) show the total for each column/account and graded with color like all the others.

Column H (22-17 (Live IBFX)) is my real account. I am running this with 0.01 fixed lots, 22 start hour, 17 look back hours, Update to break even at 15, all the rest are defaults. I am only running this on the EUR/JPY, EUR/USD, USD/JPY pairs.

These results are not complete for the first day and can not be considered hard evidence for anything yet. As time passes I am sure things will start to look different. The 08-17 and 13-17 demos still have many pending orders which will make things look different when completed. The main purpose is to put this out there for any feedback, suggestions, etc.

Mikelath has done a great job here and I value is experience. This testing is to suit my own curiosity and see what would happen in different time frames and different brokers.
Attached Image (click to enlarge)
Click to Enlarge Name: Daily 95 (2007-11-26) New.jpg Size: 266 KB
Monday's results

GBP/CHF bought @2.2764
+105pips
bought 2 positions at the 15pip target.
Seek and ye shall find, knock and the door will be opened.
Hello. It seems there are quite a few trading this successfully. Reading posts #1 to present, I believe it's o.k. to conclude that this a strategy based on mathematics (expansion of TP levels and daily HL) which is what has peaked my interest. However analyzing the formula, trading on paper, and with various backtesting, there are a few things that don't quite boost my confidence in trading this strategy. Many have mentioned that ranging markets would work best. Could someone please explain this, as I'm just not getting it. In a ranging market it is very likely to get losses on both sides. Would this not work better when price trends and is moving unidirectional? Even when price does move in one direction, it seems the likelihood of taking on losses is still very high. I think this system needs an improved method of taking back one's losses.

Thanks
The market is like the matrix - it is you that must learn to bend
Quoting GridTrader
Hello. It seems there are quite a few trading this successfully. Reading posts #1 to present, I believe it's o.k. to conclude that this a strategy based on mathematics (expansion of TP levels and daily HL) which is what has peaked my interest. However analyzing the formula, trading on paper, and with various backtesting, there are a few things that don't quite boost my confidence in trading this strategy. Many have mentioned that ranging markets would work best. Could someone please explain this, as I'm just not getting it. In a ranging market it is very likely to get losses on both sides. Would this not work better when price trends and is moving unidirectional? Even when price does move in one direction, it seems the likelihood of taking on losses is still very high. I think this system needs an improved method of taking back one's losses.

Thanks
Hi GridTrader.

To a degree this strategy reduces our risk against the effect of a ranging market by having a relatively small first profit target whereby if achieved the stoploss on the 2 remaining orders is adjusted to breakeven. We do see that price will sometimes trigger the orders and then immediately retrace and stopout all 3 orders for a -90pips result.
Bear in mind that the EA has a OCO feature that cancels the long orders when the short orders are triggered and visa versa, eliminating the possibility of getting losses on both sides.
Earlier in the thread there was considerable discussion on the idea of the markets ranging more than trending, thus suggesting the odds being against the success of this strategy. I for one am always open to any positive alternative and a ranging system based on using ATR was touched upon, however not developed sufficiently for greater acceptance. For a while I simultaneously traded a basic ranging strategy along with the D95P system to compare results. There was an idea that one could hedge against the other. My realisation was that not only did the ranging strategy require a lot more development, both strategies tended to cancel out the results of the other. Thus it made more sense to rely on one or the other. Because my focus has been testing the D95P system which relies on a breakout from the previous days high or low I've maintained that focus in order to present clear conclusive results which form the real assessment of it's worth.
So despite the apparent odds being against this strategy based on it being a trending system rather than a ranging one, ultimately if it's providing good results then all other considerations are rendered less significant.

You're suggestions for improving this system are most welcome.

rgs
mike
Mike,

Am in awe of all of yours and others contribution to this concept. Wish my head would stop spinning tho!

For uniformity sake, when you posts your results, are you still at "UpdatetoBE" of 30, No Trailstop, LookBack of 17, OrderExpire at 23?

Thanks so much EVERYONE!

Keith
Quoting mikelath
Hi GridTrader.

To a degree this strategy reduces our risk against the effect of a ranging market by having a relatively small first profit target whereby if achieved the stoploss on the 2 remaining orders is adjusted to breakeven. We do see that price will sometimes trigger the orders and then immediately retrace and stopout all 3 orders for a -90pips result.
Bear in mind that the EA has a OCO feature that cancels the long orders when the short orders are triggered and visa versa, eliminating the possibility of getting losses on both sides.
Earlier in the thread there was considerable discussion on the idea of the markets ranging more than trending, thus suggesting the odds being against the success of this strategy. I for one am always open to any positive alternative and a ranging system based on using ATR was touched upon, however not developed sufficiently for greater acceptance. For a while I simultaneously traded a basic ranging strategy along with the D95P system to compare results. There was an idea that one could hedge against the other. My realisation was that not only did the ranging strategy require a lot more development, both strategies tended to cancel out the results of the other. Thus it made more sense to rely on one or the other. Because my focus has been testing the D95P system which relies on a breakout from the previous days high or low I've maintained that focus in order to present clear conclusive results which form the real assessment of it's worth.
So despite the apparent odds being against this strategy based on it being a trending system rather than a ranging one, ultimately if it's providing good results then all other considerations are rendered less significant.

You're suggestions for improving this system are most welcome.

rgs
mike
OK, But this system seems to have changed considerably from what was proposed at the outset. If you are manually setting trades, an OCO or trailing stop feature would not be possible. Now it seems the only way this becomes a profitable strategy is to either watch the manual trade and adjust the stops accordingly or to trade the EA. And in my case I do not use an MT4 broker. I was attracted to the strategy because of its simplicity and "set and forget" feature.

Tonight using the original strategy,you'd be down 90 on USDJPY and up 15 on EURJPY. I will continue to test, however. (Assuming using EURUSD, USDJPY and EURJPY)

I would also be interested to test the posted profitability versus trending indicators. When the market breaks out and continues in a direction, it works. However this has not been the case for a few weeks.
Quoting KeithMN
Mike,

Am in awe of all of yours and others contribution to this concept. Wish my head would stop spinning tho!

For uniformity sake, when you posts your results, are you still at "UpdatetoBE" of 30, No Trailstop, LookBack of 17, OrderExpire at 23?

Thanks so much EVERYONE!

Keith
Hi Keith.

I'm using the following input settings:
UpdateToBE:15
UseTrailingStopOnLastOrder: false
LookBackHours:17
OrderExpirationHours:18

rgs
mike
Quoting Longboard
OK, But this system seems to have changed considerably from what was proposed at the outset. If you are manually setting trades, an OCO or trailing stop feature would not be possible. Now it seems the only way this becomes a profitable strategy is to either watch the manual trade and adjust the stops accordingly or to trade the EA. And in my case I do not use an MT4 broker. I was attracted to the strategy because of its simplicity and "set and forget" feature.

Tonight using the original strategy,you'd be down 90 on USDJPY and up 15 on EURJPY. I will continue to test, however. (Assuming using EURUSD, USDJPY and EURJPY)

I would also be interested to test the posted profitability versus trending indicators. When the market breaks out and continues in a direction, it works. However this has not been the case for a few weeks.
Correct. The 'set and forget' objective is being achieved by the use of the Expert Advisor which is making the implementation of this strategy even simpler than originally expected. Many thanks to pr1mer !
Quoting mikelath
Correct. The 'set and forget' objective is being achieved by the use of the Expert Advisor which is making the implementation of this strategy even simpler than originally expected. Many thanks to pr1mer !
Yes, unless you do not use an MT4 broker. Let's just say that I know a great deal aboui the inner workings of MT4, dealing desks and some of the brokers that use them. Actually the EA makes the strategy more complex, not less.

What I may do is convert your EA into EasyLanguage for Trade Station. I am more comfortable with that product.
results, so far, for today's EA trades:

EUR/JPY -30, -30, -30
USD/JPY -30, -30, -30
EUR/USD +15, -30 -30

OUCH!!!!


Are these the same results others have? Thanks!!
Quoting bjk2801
results, so far, for today's EA trades:

EUR/JPY -30, -30, -30
USD/JPY -30, -30, -30
EUR/USD +15, -30 -30

OUCH!!!!


Are these the same results others have? Thanks!!
I only trade the EUR/USD using this system, but I chose not to set trades last night on this system because of consolidation, and either way it went, ran into S & R . It may take a day or 2 before it resumes some sort of trend, as this works better when it's trending, IMO.
Quoting bjk2801
results, so far, for today's EA trades:

EUR/JPY -30, -30, -30
USD/JPY -30, -30, -30
EUR/USD +15, -30 -30

OUCH!!!!


Are these the same results others have? Thanks!!
I had the same results on my live account and both demos (IBFX & NF). The EA should have set the SL on the EUR/USD once it hit +15. It should not have gone back and hit a -30 SL on the other 2 positions. I checked my settings and they are all set to 15 on update to break even.
Quoting mtaboneweb
I had the same results on my live account and both demos (IBFX & NF). The EA should have set the SL on the EUR/USD once it hit +15. It should not have gone back and hit a -30 SL on the other 2 positions. I checked my settings and they are all set to 15 on update to break even.
Same result for me. I do not know why the other 2 trades continued on for -30 pips each. The stop losses were not moved up once the +15 pips hit. Why does the EA sometimes fail us?
0 traders viewing now