I trade with Parmar-1P system
What are your trade results?
So far not very good,but two weeks doesn´t tell reliable results.As Carninos back test showed,there could be even 12 consecutive losses,and anyway in a long run, it was second best after Daybreak TrailâŚ
Off course I could have doubts ,that political situation reflect to gbp/usd,but maybe its only amateur´s doubts:)âŚAnyway Im going to continue with this strategy,lets see how it will goesâŚ
Good for youâŚkeep us posted.
I willâŚLast evening here ,before NY close,I was little too lazy to wait our mid night,and set orders before.
So,some 23:35,sell order triggered,and made soon 5 pips lossâŚAnd for my miss,soon after mid night, price got enough move to give a profitâŚMaybe I am next time little more high.spirited,without EA it is only option:) .
Donât worry! I will be releasing the EA on Monday or Tuesday, here on this thread.
Very good news,thanks Carnino!
Hi Carnino.
Looking forward to the release of your EA.
I have been following this thread for awhile and I am hoping your EA will eliminate some of the frustrations I encounter at the moment.
1)With my day-job I am often unable to access my computer to set the pending orders at NY close. I live in New Zealand and NY close is either 9.00am or 10.00am my time (depending on daylight saving). This has meant my trading is inconsistent and I haven't been able to get a good continuous flow of trade results.
2) I haven't been able to workout how to get my MT4 platform to have a stepped trailing stop. I seem to recall reading in one of your posts that this may only be available with an EA.
I am fairly new to Forex Trading and not computer savvy either so I hope you can provide some information or links on how to download your EA onto MT4 and get it running.
Sorry for asking a lot, but I hope to be able to pay some back with contributing soon.
Thanks
Kiwiman:)
Although this is a response to [I]Kiwimanâs[/I] post, [B]it really is aimed at all traders who have no experience with EAâs[/B].
If this is your case, then you should not be running EAâs in the first place. As traders, it is your obligation to understand the environment in which you trade and that includes your trading software. If you still do not fully know how to use it, then how do you expect to succeed as a trader?
So, before trading with an EA, you should first do some research into the matter and experiment with some simple and freely available EAâs on a demo account. Here are some introductory places to visit:
[ul]
[li]BabyPipsâ School of Pipsology - Undergraduate, Junior Year: [B]How to Use MetaTrader 4[/B]
[/li]This includes a section on how to install an EA.
[li]Investopedia: [B]Beginnerâs Guide To MetaTrader 4[/B].
[/li][li]MetaTrader 4 MQL Website
[/li]Includes Book, Documentation, Articles, Codebase of free EAâs, User Forum and much more.
[li]Last but not least, Hit the [B]F1 key, or click on âHelpâ->âHelp Topicsâ[/B] in the [I]MetaTrader 4[/I] application menu for access to the Userâs Guide.
[/li][/ul]
When I release my EA, I will assume that you know how to install and use one. The only instruction guide I will be providing will be with regard to the EA functionality, inputs and the external parameters.
My apologies if this seems rather harsh or unfriendly, but sometimes in order to help someone develop and reach their full potential, it is best not to [I]spoon-feed[/I] them but to allow them to learn how to [I]forage for food[/I] by themselves.
Hi Carnino
Thanks for your honest and helpful reply. I see it as constructive and agree with your last statement.
I have done the Babypipsâ course but need to reread it. The others I will look into as well.
Thanks again for your honest opinion
Kiwiman
Last night there was again a huge gap,when the markets openedâŚAnd Carnino,I`d like to bother you again for one thingâŚI had sell order over weekend,1.62719, sl 1.62769 and tp 15 pips 1.62569 .And after midnight,price jumped straight to 1.61710(for my sense,there is already quite enough move to get profit,but seems it doesnt work that way),and the first hour bar high was 1.61859âŚSo my trade triggered in 1.61713 and immediatly sl triggered in 1.61741,not in the full 5pips,but game was over anywayâŚ
Somehow I can understand ,if the price would have reached my sl,but it was far away from itâŚIt really seems,that it is better to avoid pending orders over weekend.
Sorry,my posting,but I dont know anybody else to ask thisâŚ
Thanks in advance 
Vesa
According to the rules of all the variants of this strategy, you are only supposed to place pending orders [B]AFTER the New York Close and NOT BEFORE[/B]. That means, that you would only be placing the orders Sunday Evening or Monday Morning after the markets opened and not on Friday/Saturday.
So, during the opening you would have seen the price gap and according to another of the rules, that states that if there are [B]opening gaps[/B], then you should [B]NOT place pending orders in that direction[/B].
So, you lost Money because you did not follow the rules of the strategy. Had you followed the rules, you would have been safe. So correct the problem by not making the same mistake twice.
That is how you trade, you learn by the good and the bad trades so as to do better next time!
[I]PS! I actually made money on this one. I sometimes like to fade gaps, especially big ones like this one and made 31.1 pips. Usually gaps tend to fade/reverse and close on themselves, at least for a while, and this one was no exception and faded up for a total of 60+ pips before turning down again.[/I]
I am sure I don`t make this mistake again after this,thanks again for you!
Hi to all,
yesterday I have done some backtesting and the results are very poor :
PERIOD : 2014-02-01 00:00:00 - 2014-09-01 00:00:00
CANDLE TYPE : DAILY BAR (GMT 00 and no NY CLOSE) Sunday and Monday No trade
Symbol : GBP/USD
STRATEGY : DayBreak.Trail
STOP LOSS : 5 pips
TRAILING STOP : 5 pips
EXPIRATION PENDING ORDERS : 24H (daily close)
TAKE PROFIT : 100 pips ()
[B]GAIN (IN PIPS) : 86.90 [/B]
Iâm not interested in the $ gain so I fixed lot size 0.001 lots
here are the detail of the backtesting on DuckasCopy GMT 0 DAILY BAR.
backetesting.rar - Speedy Share - upload your files here
Choose Slow Download is Free
Thank you.
Bye
IN SHORT normalizing the results data we obtain 111 Trades (100%) divided into:
Lose Trades (~ -5pips) : 42 (38%)
Lose Trades (>-5 pips) : 2 (2%) <â I must investigate whyâŚmaybe volatility of the currency
BreakEven Trades (~ 0 pips) : 33 (30%)
[B]Win Trades (~ +5 pips) : 20 (18%)
Win Trades (~ +10 pips) : 6 (5%)
Win Trades (~ +15 pips) : 5 (4%)
Win Trades (>15 pips) : 3 (3%)[/B]
Hello Life100,
It is good that you are doing back-tests on your own so that you can test and see things for yourself. And yes, if you include the summer months of 2014, the strategy did indeed do badly for the period in question.
[B]However, your results were actually much better than you think.[/B] They just seem bad because you used the wrong amount for volume. A lot size of 0.001 is completely unrealistic. Donât just discredit the volume, because you think it unimportant. It is in fact VERY IMPORTANT. You also used the incorrect High/Lows but I discuss that further on.
Lets look at the metrics based on the Pips, since the money gains is completely unrealistic.
Your [B]Profit Factor is 1.32[/B] (good). Your [B]Win Rate is 42.30%[/B] which is on par with the strategy. Your [B]Reward:Risk ratio is 179.91%[/B] (good) and your Kelly is 10.23% (reasonable). Based on a 1% Risk, your overall [B]Gain is 17.38% for a period of 7 months[/B] which is actually not bad at all, considering the bad summer months.
[B]Overall, it did rather well considering![/B] It looked bad for you because of looking at the wrong metrics and not applying the rules of the strategy correctly. If you want to compare results however, you have to use the same rules as the rest of us. Also, consider making the back-test longer, as the time period you used is too short.
[ul]
[li]Firstly, the [B]High and Low Breakout prices have to be based on the NY Close[/B]. That means that you would have to use a Time Zone of UTC+2 or UTC+3 depending on US DST or adjust your code accordingly. However, you are using GMT which is not the same thing and will cause results to be much worse.
[/li]
[li]Secondly, unless you are using market orders triggered by a breakout on the Bid price, the [B]High Offset[/B] for Pending Orders for the DayBreakTrail should to be the value of the average [B]Spread[/B]. However, you do not mention what your spread was. In order to compare results, both [I]Vijay[/I] and I have used a reference spread of 1.5pips for GBP/USD and 1.0 EUR/USD. Obviously in real life that spread is most probably not what you will get, but in order to compare results, one has to use a reference spread and stick to it for all tests.
[/li]
[li]Thirdly, is the [B]lot size[/B]. We have been using a [B]1% Risk (with compounding) [/B]and a $10000 opening balance. I suggest that you use something similar so that it is easier to see the results and can truly see what the percentage gains or losses are. Using a [B]fixed lot size 0.001 is completely unrealistic[/B] and will be impossible to calculate realistic gains. I do not even know of any broker offering such small lot sizes and the smallest I have seen is 0.01 or micro accounts.
[/li]
[li]Also, please also share the [I]MetaTrader Strategy Report [/I]files and not just the web analysis, so that we can get down to the nuts and bolts of the test.
[/li]
[li]I also suggest that you use [I]MyFXBook[/I] for strategy analysis instead of [I]FX Backtesting[/I] as it offers much more information and statistical analysis and is completly free.
[/li][/ul]
[B]Keep up your interest and your R&D - It is good for you and all other traders too![/B]
I was hoping for a reply like this soâŚThank you Carnino.
I want to ask some question:
- I not use MetaTrader but I use JForex (duckasCopy platform) so I canât share MetaTrader Strategy Report⌠I must find a method to hook JForex to MyFXBook⌠I donât know.
2)How you have calculated the metrics (excuse the ignorance) :
*)Profit Factor : 1.32
*)Reward: Risk ratio is 179.91%
*)Kelly : 10.23%
And why you calculate kelly?? You use the Kelly Criterion to decide how much risk? Donât you use 1%?
3)I use a Swiss Broker so the Daily Candle close is GMT 00:00 ⌠but I can use the method that you show in past threads (1Hr chart) for the NY Close ⌠the problem is the exchange standard time and daylight saving time that I donât know how to manage
4)High OffsetâŚmhmhmh⌠example : if the high is 1.55550 and the spread is 1.0 pips I must open long order at 1.55560⌠itâs right? You tell me this?
5)This evening I update Backtesting with 1% on the equity and show you results tomorrow.
Thanks for all the good informations that you give to me.
Bye bye.
Hi people, I did some backtests also but I think the best way to really test this strategy is in demo or live account.
For example you can test it live for 1 month and then backtest it for that month to see if the results are the same you got live.
I will start to do it next week(will post myfxbook) in IC Markets because they have pretty good spread and that is important for this strategy.
Regards
You are welcome [I]Life100[/I]! Here are my answers to your queries:
[ol]
[li]If using [I]JForex[/I], then unfortunately, [I]MyFXBook[/I] will not work for you as they only accept [I]MetaTrader[/I] reports.
[/li]
[li]Metrics (in your case)
[/li]Avg. Profit = 7.61 pips, Avg. Loss = 4.23 pips,
Good Trades = 47, Bad Trades = 64, Total Trades = 111
[LIST=a]
[li]Profit factor = Gross Profit / Gross Loss
[/li]Profit factor = (Avg. Profit x Good Trades) / (Avg. Loss x Bad Trades)
In your case: Profit Factor = 357.67 pips / 270.72 pips = 1.321181
[li]Reward:Risk Ratio = Avg. Profit / Avg. Loss
[/li]In your case: R:R Ratio = 7.61 pips / 4.23 pips = 1.799054 = 179.9054%
[li]Kelly Criterion = Win Rate - (1 - Win Rate) / Risk Reward Ratio
[/li]In your case: Kelly = (47/111) - (64/111)/(7.61/4.23) = 0.102935 = 10.2935%
The Kelly Criterion can be used to increase the Risk%, but I also use it to measure how âsafeâ a strategy is. If the Kelly is high then it means that it is safer and the [I]Risk of Ruin[/I] is lower and will be less likely to have high drawdowns.
[/ol]
[li]Yes, use a 1Hr Chart to find the correct High/Low for the New York Close. For automation, use a parameter (which you can change during run-time) to offset the hour based on the Daylight Savings adjustments for New York, which you can find at the following [I][B]TimeAndDate[/B][/I] webpage:
[/li]Current local time in U.S.A. â New York â New York
[li]Yes, that is correct. Add the Average Spread to the High because that will simulate the Ask Price that will be used when the Pending order is triggered. Alternatively, if you are automating it, then you can use a Market Order instead when the Bid Price breaks the High bid price. That is how I do it in my EA, as this is more accurate for the real world where the spread varies and is not constant. Using the Market Order also makes it easier to account for slippage than a Pending Order.
[/li][/LIST]
Regards, Carnino.
Wow thank you for the metric informationsâŚI never consider the Kelly Criterion to measure how safe is a strategyâŚbut it seems a good idea.
I have another question⌠if you look at my resultsâŚthere are a couple of trade n°31 and n°82 where the stop Loss is not respectedâŚand I donât understand whyâŚ
Trade 31 : -25.9 pips
Trade 82 : -21.7 pips
When I open an order I always set 5 pips stop Loss and I adjust it when the order is filled so are 5 pips from the open price⌠so why the stop loss is not respected by Broker?
-25.9 pips is -5.2% loss in a trade (5 pips == 1%)âŚitâs very high
there is a way to avoid this issue?
Thank you
Bye bye
That is called â[B][I]slippage[/I][/B]â. I am surprised you did not look into that before as it has been a major topic on the thread on several occasions.
Slippage can happen at any time during the open or the close of an order and can be a negative or a positive value.
Slippage Definition | Investopedia
PS! In [I]MetaTrader[/I], you can define maximum allowable slippage for an order to open and then later to close it, but I do not know how it is managed in [I]JForex[/I]. In the case of those two orders, had you set the maximum slippage allowed, they would never have been opened in the first place since they triggered the stop-loss at the same time they were opened (equivalent to a boundary violation), probably caused by a widening spread.