One trade per week (a work-in-progress trading idea)

[dukeaugustus]

Background: I have noticed that specific currencies, within a single week, after price moved away from the week's open, will likely to either

(a) "revert back to the week's open price" at least once;

(b) or "run-away even further from the week's open price."

 

So here is my work-in-progress trading idea:

(a) 1st group applied to two currenncy pairs who like to "revert", and

(b) 2nd group are two currency pairs who like to "run-away":

 

"Reversion" Trade Pairs: USDCAD, EURCHF

"Run-away" Trade Pairs: GBPJPY, EURJPY

 

"Reversion" Trade Rules:

(1) At start of week, place pending sell order (sell limit order) 90 pips above week's open price

(2) at the same time as (1), place pending buy order (buy limit order) 90 pips below the week's open price

(3) Set take profit at 90 pips

(4) Set stop loss at 90 pips

(5) On each currency, if one pending order is triggered, then cancel the opposite one (or you can use OCO order if your broker offers it)

(6) Before end of week, close all active orders, and cancel all pending orders

 

"Run-away" Trade Rules:

(1) At start of week, place pending buy order (buy stop order) 90 pips above week's open price

(2) at the same time as (1), place pending sell order (sell stop order) 90 pips below the week's open price

(3) Set take profit at 90 pips

(4) Set stop loss at 90 pips

(5) On each currency, if one pending order is triggered, then cancel the opposite one (or you can use OCO order if your broker offers it)

(6) Before end of week, close all active orders, and cancel all pending orders

 

Weekly start time and end time:

Although we probably do not need precise weekly start time or end time. But for the purpose of consistency, I have been using:

(1) Sunday 9 pm New York time price as the week's open price, and

(2) Friday 3 pm New York time for end of week.

 

Advantages of this trading idea:

(1) Very low time requirement

(2) No indicator or chart reading required

(3) Purely mechanical

(4) Can be mostly set-and-forget

(5) Somewhat balanced risk exposure to individual currency: Of the currencies involved (USD, CAD, EUR, CHF, GBP, JPY), only EUR and JPY have double risk exposure

(6) Somewhat balanced risk exposure to overall market behavior at the time (for example, if overall market favors strong trend, then at least the "Run-away" trades will shine. Likewise if market is mostly ranging, then the "reversion" trades will shine.)

(7) In my back test, each individual currency pair is profitable trading on its own. The "synergy" of trading all four pairs together seem to smooth out the equity curve somewhat.

 

Disadvantages of this trading idea:

(1) Very slow equity line growth (NOT a grow-rich-quick method)

(2) Probably more suitable as a supplementary method, and less suitable as a main trading method.

(3) Can have long losing streak, especially during the summer months for the "Run-away" trading pairs.

 

Final notes:

The reason I call this an "idea" (as opposed to "method" or "system"), and the reason I call this "work-in-progress", is I believe this still has much room for improvements. There are so many variables that can be further fine-tuned, like currency pairs used, level of target profit, level of stop loss, method of taking profit (use trailing stop? partial profit at 50%?), etc. So any suggestions for improvements are welcomed from friends here at II.

 

Note on Money Management:

Although I am trading this live; if anyone here is going to put real money in this, I would ALWAYS recommend (a) do your own due diligence first to see if this is truly your style of trading, and (b) maintain good money management. For example, anyone using Martingale for the JPY pairs will likely see their account wipe out. I know I am probably preaching to the choir, but just in case any new trader is reading this. ALWAYS have caution where money management is concerned.

Edited January 9, 2012 by dukeaugustus

[dukeaugustus]

Here are the results for the previous 13 months (60 weeks) in pips:

 

UC = USDCAD (for reversal to week's open price)

EC = EURCHF (for reversal to week's open price)

GJ = GBPJPY (for run-away from week's open price)

EJ = EURJPY (for run-away from week's open price)

 

Week of __UC _EC _GJ _EJ

08/11/10 +00 n/a -90 +90

15/11/10 -90 n/a -90 -90

22/11/10 +90 n/a +90 +90

29/11/10 -90 n/a -90 +90

06/12/10 +30 n/a +90 -90

13/12/10 +90 n/a +90 -90

20/12/10 +00 n/a +90 +90

27/12/10 -70 n/a +90 -90

03/01/11 +90 n/a +90 +90

10/01/11 +00 n/a +90 +90

17/01/11 +30 n/a -90 -90

24/01/11 +00 n/a +90 -90

31/01/11 -90 n/a +90 -90

07/02/11 +70 n/a +10 +30

14/02/11 +00 n/a +90 +30

21/02/11 +90 n/a +90 -90

28/02/11 +30 n/a +90 +90

07/03/11 +00 n/a +90 +90

14/03/11 -90 n/a +90 +90

21/03/11 +80 n/a -90 -90

28/03/11 -90 n/a +90 +90

04/04/11 +20 n/a +90 +90

11/04/11 +60 n/a +90 +90

18/04/11 +90 n/a +90 +90

25/04/11 +00 n/a -90 -90

02/05/11 -90 n/a +90 -90

09/05/11 +90 n/a +90 -90

16/05/11 +00 n/a +90 +90

23/05/11 +00 n/a -90 -90

30/05/11 +90 n/a +90 +90

06/06/11 +00 n/a -90 -90

13/06/11 +90 n/a -90 -90

20/06/11 +90 n/a -90 -90

27/06/11 -90 n/a -90 +90

04/07/11 +90 n/a -90 +90

11/07/11 +90 n/a +90 +90

18/07/11 +10 n/a -20 +90

25/07/11 +90 n/a +90 +90

01/08/11 -90 n/a +90 +90

08/08/11 -90 n/a +90 +90

15/08/11 +90 n/a -90 -90

22/08/11 +90 n/a -90 -90

29/08/11 +00 n/a -90 +90

05/09/11 +90 +90 -90 -90

12/09/11 -90 +00 +90 -90

19/09/11 -90 -90 +90 -90

26/09/11 +90 +30 +90 +90

03/10/11 +90 -90 +90 +90

10/10/11 -90 +90 +90 +90

17/10/11 +90 +90 +90 +90

24/10/11 +90 +90 -90 -90

31/10/11 +90 +00 -90 -90

07/11/11 +90 +90 +90 +90

14/11/11 -40 +00 +90 +90

21/11/11 -90 +14 +15 +15

28/11/11 -90 +90 +90 +90

05/12/11 +90 +00 +00 +03

12/12/11 -90 -67 -10 +90

19/12/11 -90 +00 +90 -53

26/12/11 +00 +14 +90 +90

02/01/12 +90 +00 -17 +80

 

60 wks Total 760 351 1705 915

Net all Total 3651 pips

 

60 wks average for USDCAD = 13 pips per week

13 wks average for EURCHF = 21 pips per week

60 wks average for GBPJPY = 28 pips per week

60 wks average for EURJPY = 14 pips per week

 

general average for combined 4 pairs = 76 pips per week

general average for any single pair = 19 pips per week

 

longest losing streak for USDCAD = 3 weeks

longest losing streak for EURCHF = 1 week (not enough data)

longest losing streak for GBPJPY = 5 weeks

longest losing streak for EURJPY = 3 weeks

 

longest winning streak for USDCAD = 4 weeks

longest winning streak for EURCHF = 5 weeks (not enough data)

longest winning streak for GBPJPY = 8 weeks

longest winning streak for EURJPY = 7 weeks

 

Please note that EURCHF only has 17 weeks of data here, I started counting EURCHF from the week that its trading behavior did a 180 degree turn after Swiss bank intervention in early September.

Edited January 21, 2012 by dukeaugustus

[dukeaugustus]

Some currency specific notes

 

USDCAD specific notes:

(1) Although not the strongest performer, but it is the most consistent performer.

(2) So far the longest single losing streak has been 3 weeks, and longest winning streak (ignoring no-trade weeks) has been 4 weeks.

(3) Although I do not suggest using Martingale, But this is the only one in the four currency basket I may consider using such potential high risk loss recovery.

(4) At least from the past 13 months data, it can be traded throughout the year, even on summer or winter holiday months.

 

EURCHF specific notes:

(1) Up until early September 2011, this pair was in fact the most profitable "Run-away" trade currency of 2011.

(2) But after Swiss bank authority announced they were "fixing" CHF to EUR in early September , EURCHF is now changed to "Reversion" type trades.

(3) Will need to monitor this pair closer than the other pairs: if Swiss authority once again let CHF trades freely, then we may need to abandon this pair as a "Reversion" type trade.

(4) Because of this pair's short time period since Swiss bank authority's intervention, there is not enough data to say how profitable this will be in the long run.

 

GBPJPY and EURJPY specific notes:

(1) Both pairs are far stronger performers than USDCAD.

(2) But both pairs are also less consistent, with likely longer streak of winners and losers, with longest losing streak at 5 weeks, and longest winning streak at 8 weeks.

(3) Both pairs suffered worst losing streak during the summer months.

(4) I would recommend either not trade these two pairs during the summer, or to reduce lot size during those months.

Edited January 7, 2012 by dukeaugustus

[dukeaugustus]

Forward Test Results 2012

 

*** Last 2012 week updated: week of 24 December 2012 ***

*** Total Weeks in 2012 Tracked: 52 weeks ***

 

 

Weekly pips for USDCAD, EURCHF, GBPJPY, EURJPY, and Total four pairs

 

week of 02/01/12: +90 +00 -17 +80 Total = +153

week of 09/01/12: +20 +00 -07 -90 Total = -077

week of 16/01/12: -10 +00 +90 +90 Total = +170

week of 23/01/12: +90 +00 +90 +90 Total = +270

week of 30/01/12: -07 +00 -90 +90 Total = -007

week of 06/02/12: +00 +00 +90 +90 Total = +180

week of 13/02/12: +00 +00 +90 -90 Total = +000

week of 20/02/12: +05 +00 +90 +90 Total = +185

week of 27/02/12: -37 +00 +90 +90 Total = +143

week of 05/03/12: +90 +00 -90 +90 Total = +090

week of 12/03/12: +00 +00 +90 +90 Total = +180

week of 19/03/12: +14 +00 -90 -90 Total = -166

week of 26/03/12: +00 +00 +90 -90 Total = +000

week of 02/04/12: +00 +00 +90 +90 Total = +180

week of 09/04/12: +00 +00 -90 -90 Total = -180

week of 16/04/12: +18 +00 +90 +90 Total = +198

week of 23/04/12: -50 +00 +00 -30 Total = -080

week of 30/04/12: -53 +00 -90 +80 Total = -063

week of 07/05/12: +00 +00 -90 +00 Total = -090

week of 14/05/12: -90 +00 +90 +90 Total = +090

week of 21/05/12: -22 +00 -90 +90 Total = -022

week of 28/05/12: -90 +00 +90 +90 Total = +090

week of 04/06/12: -90 +00 +90 +90 Total = +090

week of 11/06/12: +70 +00 -90 +90 Total = +070

week of 18/06/12: +70 +00 -90 -90 Total = -140

week of 25/06/12: +90 -90 +90 +90 Total = +180

week of 02/07/12: +90 +00 +50 +90 Total = +230

week of 09/07/12: +00 +00 -90 -30 Total = -120

week of 16/07/12: +00 +00 +00 +50 Total = +050

week of 23/07/12: -15 +00 -90 -90 Total = -195

week of 30/07/12: +00 +00 +90 -90 Total = +000

week of 06/08/12: +00 +00 -90 +35 Total = -055

week of 13/08/12: +00 +00 +90 +90 Total = +180

week of 20/08/12: +00 +00 +00 -90 Total = -090

week of 27/08/12: +00 +00 +00 +00 Total = +000

week of 03/09/12: -02 -02 +90 +90 Total = +176

week of 10/09/12: +10 +00 +90 +90 Total = +190

week of 17/09/12: +20 +00 +90 -90 Total = +020

week of 24/09/12: +00 +00 +02 +20 Total = +022

week of 01/10/12: +40 +00 +90 +90 Total = +220

week of 08/10/12: +00 +00 +90 +90 Total = +180

week of 15/10/12: -35 +00 +90 +90 Total = +145

week of 22/10/12: +00 +00 -90 -90 Total = -180

week of 29/10/12: +00 +00 -90 -70 Total = -160

week of 05/11/12: +00 +00 +90 -90 Total = +000

week of 12/11/12: +00 +00 +90 +90 Total = +180

week of 19/11/12: +00 +00 +90 +90 Total = +180

week of 26/11/12: +00 +00 -90 -90 Total = -180

week of 03/12/12: +00 +70 +00 -10 Total = +060

week of 10/12/12: +00 +00 +90 +90 Total = +180

week of 17/12/12: +07 +00 +90 +90 Total = +187

week of 24/12/12: +00 +00 +90 +90 Total = +180

 

 

Cumulative USDCAD = +193 pips

Cumulative EURCHF = -22 pips

Cumulative GBPJPY = +1198 pips

Cumulative EURJPY = +1030 pips

Cumulative all pairs: +1475 pips

 

 

Average weekly USDCAD = +4 pips

Average weekly EURCHF = +0 pips

Average weekly GBPJPY = +23 pips

Average weekly EURJPY = +28 pips

 

Average combined per week = +55 pips

Average single pair per week = +14 pips

Edited December 28, 2012 by dukeaugustus

[sillykiddo]

duke

As always i like your idea... 153 pips is not bad at all... btw do you use go market broker weekly open or any broker specific?

[dukeaugustus]

duke

As always i like your idea... 153 pips is not bad at all... btw do you use go market broker weekly open or any broker specific?

 

Hi sillykiddo, glad you like it.

 

My live trade broker is Oanda. But I assume that any broker will work. If there is any broker differences, they will probably even themselves out in the long run.

 

In my case, I simply use New York time Sunday 9 pm as week's start, and Friday New York time 3 pm as week's end. These are NOT Oanda's week start or end time, just simply a time that is convenient for me. And in back test these times seem to work so far. Though I assume if we are off by 1 or 2 hours either way, as long as price has not moved strongly already, probably does not matter much in the long run. (And, in general, prices do tend to move slowly most of the weeks during those early hours anyway).

 

Edit: also, since we are placing pending orders, sometimes I come in a few hours to half a day late, and can still place my pending orders based on the price at week's open.

Edited January 7, 2012 by dukeaugustus

[dukeaugustus]

Month 1 (Jan 2012) Update: 500+ pips

 

I have been, and will attempt to continue to, update on a weekly basis directly into Post #4 on the front page of this thread.

 

Then on a monthly basis, I will also try to do a simple summary report of my 2012 forward test up to the end of the last full week for that month.

 

So here is my first such monthly report: for January 2012:

 

For actual detailed break-down figures, please see Post #4.

 

Overall, the method is working out as planned. The first month of 2012 started off well. With total just over 500 pips in one month (approx 516 pips). Which is well above back-test average.

 

So I will continue to test with a live account. But I do expect some losing weeks to come. The long term expected weekly average was about 76 pips, or expected 300 pips per month. So January's weekly average of 129 pips and month's total of 500+ pips are both way above expected average.

 

Although I am not complaining. But I do expect the numbers to eventually come back down to long-term average. So I would not be surprised if we see some losing weeks soon.

 

Side Note:

(1) My live account did not get triggered on the last week for the USDCAD winning trade. For some reason, my broker must had increased the spread to at least 5 pips or more when the price actually reached and surpassed the pending order price by a few pips, before retracing. So I could only sit on the sideline watching it went back to take-profit price, without my money on it. Ah, well, that's trading for me. At least no harm's done.

 

(2) I notice EURCHF has not triggered any trade for a while now. The weekly range is now far tighter than last year's. But since non-triggered pending orders do not take up any margin, so I will still keep placing the pending orders at the start of each week. But if this continues for much longer, I will either have to (a) consider lowering my TP an SL and where Pending Order is placed, or (b) drop this pair completely. We'll see.

Edited February 11, 2012 by dukeaugustus

[dukeaugustus]

Month 2 (Feb 2012) Update: +358 pips

 

This past week was the last complete trading week for February, so here is a quick monthly update of my forward test for this.

 

February single month gain: +358 pips.

 

Net January + February gain: +874 pips.

 

For details of week-to-week and currency-by-currency details, please see weekly update in post #4.

 

General thoughts:

(1) Seems to be still performing positively so far;

(2) My biggest problem is my own trading psychology (read: I still try to micro-manage this, and always to my own detriment, for example, I still find myself closing winning trades early)

(3) Looking at 2011 results, this had a great run from Jan thru April, let's see if the next few months still add more pips to my account.

[energetic]

Thanks for sharing ")

[dukeaugustus]

Month 3 (Mar 2012) Update: +247 pips

 

This week ends my 3rd month of the 2012 forward test of this, as well as end of first quarter of 2012.

 

March single month results: +247 pips

 

First quarter (Jan - Mar) results: +1121 pips

 

For details of per currency and per week results, see the weekly updated post #4.

 

General thoughts:

 

(1) March was the worse performing month so far in 2012, with the last 2 weeks having results of -166 pips and then break-even. Fortunately, the month still ended up in a net profit of +247 pips, mainly from strength of early part of March.

 

(2) For my own trading account, I actually ended the last week at over 90 pips profit (instead of breakeven): After GbpJpy took full profit, I was at the screen when I saw EurJpy made over 50 pips, then hit a long term trend line I had drawn previously for my other trading system, and bounced strongly back down. So I decided to close that trade out at a few pips above breakeven before the full loss. However, for consistency of this thread, I recorded a full loss for for EurJpy on this last week.

 

(3) After 3 full months of not a single trade, I will no longer use EurChf for my live trade. With Swiss bank authority interfering with this pair, it just no longer moves anymore in 2012. The closest to trigger was within 10 pips, once, in the entire 3 months period. Last week in March it hardly moved more than 10 or 15 pips either side of the week's open price. Just not worth the hassle of putting on trades for this.

 

(4) I have not decided what pair, if any, to replace EurChf in my basket. Any suggestion is welcomed.

 

(5) Given the slowness of this system. And the seeming lack of activities on this thread. I will probably only gives update here once every 2 to 3 months from now on. But I will try to update post #4 at least once every 2 to 3 weeks.

 

(6) Overall, I am still happy this this method, for something that requires only 10 minutes at start of week to place pending orders, and another 5 minutes max per week in managing this, I can't complain about the +1000 pips in 3 months.

[kraven]

well done duke. interesting to see how this progresses. i know there's not much activity here but use it for your own reflection

[Roels Major]

Similar to the ideas on these thread, Check it up:

 

1. http://indo-investasi.com/showthread.php/10067-First-Strike/

2. http://www.forexfactory.com/showthread.php?t=293209

[san1111]

Hi dukeaugustus,

 

nice work and thanks for sharing!

[Jay007]

Hi, duke,

thanks for a brilliant effort.

Could you please elaborate your method/way of using martingale with USDCAD, how would you try to recover the loss, and yes am not in the choir))

thanks once again

USDCAD specific notes:

(1) Although not the strongest performer, but it is the most consistent performer.

(2) So far the longest single losing streak has been 3 weeks, and longest winning streak (ignoring no-trade weeks) has been 4 weeks.

(3) Although I do not suggest using Martingale, But this is the only one in the four currency basket I may consider using such potential high risk loss recovery.

(4) At least from the past 13 months data, it can be traded throughout the year, even on summer or winter holiday months.

 

EURCHF specific notes:

(1) Up until early September 2011, this pair was in fact the most profitable "Run-away" trade currency of 2011.

(2) But after Swiss bank authority announced they were "fixing" CHF to EUR in early September , EURCHF is now changed to "Reversion" type trades.

(3) Will need to monitor this pair closer than the other pairs: if Swiss authority once again let CHF trades freely, then we may need to abandon this pair as a "Reversion" type trade.

(4) Because of this pair's short time period since Swiss bank authority's intervention, there is not enough data to say how profitable this will be in the long run.

 

GBPJPY and EURJPY specific notes:

(1) Both pairs are far stronger performers than USDCAD.

(2) But both pairs are also less consistent, with likely longer streak of winners and losers, with longest losing streak at 5 weeks, and longest winning streak at 8 weeks.

(3) Both pairs suffered worst losing streak during the summer months.

(4) I would recommend either not trade these two pairs during the summer, or to reduce lot size during those months.

[dukeaugustus]

Similar to the ideas on these thread, Check it up:

 

1. http://indo-investasi.com/showthread.php/10067-First-Strike/

2. http://www.forexfactory.com/showthread.php?t=293209

 

Thanks for pointing this out Roels. I have only read your first post so far (will continue to read rest of your thread later this weekend), and already I like your idea of adjusting the weekly entry distance dynamically. This may well solve my problem with EurChf having zero trade for three months.

 

Although for this current thread, for the sake of consistency and continuity, I will keep recording the results according to my post #1 rule unchanged, but for my personal trading, I may well test out your idea of a dynamic changing entry position. Again, thanks.

[dukeaugustus]

Hi, duke,

thanks for a brilliant effort.

Could you please elaborate your method/way of using martingale with USDCAD, how would you try to recover the loss, and yes am not in the choir))

thanks once again

 

Hi Jay, you're welcome.

 

In general, I do NOT recommend using martingale. At least not the full doubling after every loss approach.

 

For any loss recovery chasing-type method to work, obviously, we have to assume UsdCad in 2012 will behave similarly to 2011, which has very short losing streak (max in 2011 was 3 consecutive losses).

 

So I would probably do something like this (I use 1.3 as the increase factor below, but you can use anywhere from 1.1 to 1.5, I would not go much higher than that):

 

Trade 1 = 1 lot size

If Trade 1 was a loss, then Trade 2 = 1.3 lot size

If Trade 2 was also a loss, then Trade 3 = 1.3 x 1.3 = 1.7 lot size

If Trade 3 was also a loss, then Trade 4 = 1.69 x 1.3 = 2.2 lot size

If Trade 4 was also a loss, then Trade 5 = 2.2 x 1.3 = 2.9 lot size

etc.

 

Then the above sequence goes in reverse when I have a winning week. For example, if I had a win in Trade 5 = 2.9 lot, then Trade 6 will be the previous size smaller than Trade 5 of 2.9 lots, which is 2.2 lots. And I will continue until either I am happy, or until it returns back to 1 lot size.

 

But you must set a top limit of how many consecutive losses you are willing to accept and then stop chasing. For example, since last year's max was 3 consecutive losses, perhaps you will set 4 consecutive losses as the top max. Otherwise, the losses can mount up very large very quickly. For example, in the case of 5 consequent losses, flat lot size will only be a loss of total 5 lots. But at chasing with 1.3 increase each week, the same 5 weeks losing streak would result in over 9 lots total loss. (If chasing at full martingale doubling after each loss, these 5 weeks of losses would total 31 lots!!)

 

Personally, in my live trade so far, I have not applied lot size increase to chase after loss, even on UsdCad. I guess I am always worried that Murphy's Law would knock on my door. Although UsdCad in 2011 had very short losing streaks of no longer than 3 consecutive losses, there is no guarantee that it cannot start a long losing streak sometimes in the future.

 

For me, as long as the method is still profitable with flat lot size (I will increase lot size only after equity has increased to a certain ratio), I am willing to take slower equity gain in return for sleeping better at night, without worrying about possible martingale wipe-out.

 

But everyone has different risk-tolerance. What works for me may not suit others.

[Jay007]

Thanks for your time and explanation, I do use martingale but i dont close my loosing order, and not has you explained it, i just keep adding to the same position with a fixed no. of pips(its only when i am sure that the price will return back or if its just testing a different support/resistance level), if not sure, i just happy to put stop loss n lose the amount of pips, usually this works well when price has already moved its usual average daily range, and i can pick up the top or the bottom, counter trend trading for a small profit.

Hi Jay, you're welcome.

 

In general, I do NOT recommend using martingale. At least not the full doubling after every loss approach.

 

For any loss recovery chasing-type method to work, obviously, we have to assume UsdCad in 2012 will behave similarly to 2011, which has very short losing streak (max in 2011 was 3 consecutive losses).

 

So I would probably do something like this (I use 1.3 as the increase factor below, but you can use anywhere from 1.1 to 1.5, I would not go much higher than that):

 

Trade 1 = 1 lot size

If Trade 1 was a loss, then Trade 2 = 1.3 lot size

If Trade 2 was also a loss, then Trade 3 = 1.3 x 1.3 = 1.7 lot size

If Trade 3 was also a loss, then Trade 4 = 1.69 x 1.3 = 2.2 lot size

If Trade 4 was also a loss, then Trade 5 = 2.2 x 1.3 = 2.9 lot size

etc.

 

Then the above sequence goes in reverse when I have a winning week. For example, if I had a win in Trade 5 = 2.9 lot, then Trade 6 will be the previous size smaller than Trade 5 of 2.9 lots, which is 2.2 lots. And I will continue until either I am happy, or until it returns back to 1 lot size.

 

But you must set a top limit of how many consecutive losses you are willing to accept and then stop chasing. For example, since last year's max was 3 consecutive losses, perhaps you will set 4 consecutive losses as the top max. Otherwise, the losses can mount up very large very quickly. For example, in the case of 5 consequent losses, flat lot size will only be a loss of total 5 lots. But at chasing with 1.3 increase each week, the same 5 weeks losing streak would result in over 9 lots total loss. (If chasing at full martingale doubling after each loss, these 5 weeks of losses would total 31 lots!!)

 

Personally, in my live trade so far, I have not applied lot size increase to chase after loss, even on UsdCad. I guess I am always worried that Murphy's Law would knock on my door. Although UsdCad in 2011 had very short losing streaks of no longer than 3 consecutive losses, there is no guarantee that it cannot start a long losing streak sometimes in the future.

 

For me, as long as the method is still profitable with flat lot size (I will increase lot size only after equity has increased to a certain ratio), I am willing to take slower equity gain in return for sleeping better at night, without worrying about possible martingale wipe-out.

 

But everyone has different risk-tolerance. What works for me may not suit others.

[dukeaugustus]

Thanks for your time and explanation, I do use martingale but i dont close my loosing order, and not has you explained it, i just keep adding to the same position with a fixed no. of pips(its only when i am sure that the price will return back or if its just testing a different support/resistance level), if not sure, i just happy to put stop loss n lose the amount of pips, usually this works well when price has already moved its usual average daily range, and i can pick up the top or the bottom, counter trend trading for a small profit.

 

I am glad it is working out for you, Jay. Although I am probably not likely using Martingale type strategy in my own live trade, but I can see the way you describe to work especially in intra-day trade with something like a London close type of entry: For example, wait for price to move beyond its average daily range after all major news near London close, and trade a bounce back from a support/resistance level. If first attempt fail, barring unexpected news event, then it is still likely for price to return back in your direction after finding another level of support/resistance. Then if you add on another trade at the next support/resistance further out in the original direction of the first trade, you have averaged up/down the break-even price of your entire position.

 

For the system in this current thread, I have two pairs which I want to return to weekly open, and two pairs which I want to run away from the open. So, for the two currency pairs (now just 1 as I just dropped EurChf) which we trade its reversal after moving away from the week's open, looking at the weekly price dynamic, perhaps you can use average weekly range, instead of daily range. But I have not tested this out.

 

And of course, for my two Jpy pairs (EJ, GJ), I do NOT want them to come back to weekly open price, I want them to run away from the open, at least far enough to book my profit.

Edited March 31, 2012 by dukeaugustus

[Jay007]

Duke,

what about calculating the trend run or reversal setups at the end of the week, rather than following a static approach all the time, might it have some influence on the risk reward and equity curve, what i want to make of note is that, market are ever changing, analogical to humans))

If i come up with something measurable, will let you know, as for now, i am really liking the idea of one trade per week,thanks

I am glad it is working out for you, Jay. Although I am probably not likely using Martingale type strategy in my own live trade, but I can see the way you describe to work especially in intra-day trade with something like a London close type of entry: For example, wait for price to move beyond its average daily range after all major news near London close, and trade a bounce back from a support/resistance level. If first attempt fail, barring unexpected news event, then it is still likely for price to return back in your direction after finding another level of support/resistance. Then if you add on another trade at the next support/resistance further out in the original direction of the first trade, you have averaged up/down the break-even price of your entire position.

 

For the system in this current thread, I have two pairs which I want to return to weekly open, and two pairs which I want to run away from the open. So, for the two currency pairs (now just 1 as I just dropped EurChf) which we trade its reversal after moving away from the week's open, looking at the weekly price dynamic, perhaps you can use average weekly range, instead of daily range. But I have not tested this out.

 

And of course, for my two Jpy pairs (EJ, GJ), I do NOT want them to come back to weekly open price, I want them to run away from the open, at least far enough to book my profit.

[dukeaugustus]

Duke,

what about calculating the trend run or reversal setups at the end of the week, rather than following a static approach all the time, might it have some influence on the risk reward and equity curve, what i want to make of note is that, market are ever changing, analogical to humans))

If i come up with something measurable, will let you know, as for now, i am really liking the idea of one trade per week,thanks

 

Thanks Jay. And yes, I understand the draw-back of my static approach. Although it has the beauty of very low time requirement for calculating and putting on trades, my static approach will not react to the market when market dynamics change. For example, EurChf is no longer in my active trading basket starting next week, precisely because its personality in 2012 is very different from 2011.

 

I understand the logic of using some kind of average weekly range to adjust for entry and TP/SL levels as time moves on, something like what is suggested in the strategy mentioned by Roels Major in post #12 above.

 

As I mentioned in my original post #1, I believe this strategy is a work-in-progress idea, with many variables that can be adjusted. I am thankful for all the great suggestions for me to improve this method. I will probably play around different ideas a bit over the next few months in my private trades, and see how they go. Will report back if I discover anything worthy of sharing.

[absolute beginner]

Thanks for all your effort to report your findings and system description in general. It is the first time I have seen this thread. Will definitely look into it.

[dukeaugustus]

I am stopping the Forward Test for now

 

Hi friends of II. Just a post to say that I will be stopping my forward test of this method, at least for the foreseeable future.

 

My reasons are as follow:

 

(1) Ever since discarding EURCHF, and with poorly performing USDCAD for the bulk of 2012, I have lost my initial "balance" of having two reversal pairs to complement two expanding pairs. Although the two JPY pairs have given my some nice profits in 2012, but the overall method just does not seem as "balanced" as I like.

 

(2) Also, we recently just suffered three consecutive losing weeks. Last year that did not happen until at least June. As I mentioned in my first post, I always planned to stop trading the JPY pairs at least over the summer months (June, July, August). And with non-performing USDCHF and USDCAD, I thought I may as well stop this test a few weeks earlier at this point in time.

 

(3) My initial target was to double my account by end of May. But my account for this is only sitting at about +50% right now. Not complaining of course, but still failing my interim test target.

 

(4) My philosophy is that the market is always changing, it seems what used to work last year is no longer as valid this year. So it is time to get out of the party, especially when I am still up overall in profit.

 

(5) I may still come back to it later in this year. But in the mean time, I am focusing my attention on a couple of other ideas that I have been playing with. Will share them if they are profitable.

 

(6) Finally, thanks for all the encouragements and advices from my friends here at II over the last few months for this trading idea thread.

 

Wish you all many happy pips to come.

 

p.s. if I have the time, I will still update post #4 on an irregular basis, just to see how it would have done over time.

[joe_sixpack]

"My philosophy is that the market is always changing, it seems what used to work last year is no longer as valid this year."

 

Yes, the market does change all the time and yet it stays the same. The core principles and ideas do stay the same, but don't try to construct too elaborate approaches because they tend to be inflexible. I am afraid that your approach is just that: too elaborate. That was my first impression to your first posts. Try to be more flexible. You also mentioned some problems with your trading psychology. You need to fix those because otherwise no system will work for you. Often these problems are caused by undercapitalizion. I am not sure if that's your case, but it could be.

 

Anyway, just my 2 cents. I wish you good luck.

[dukeaugustus]

"My philosophy is that the market is always changing, it seems what used to work last year is no longer as valid this year."

 

Yes, the market does change all the time and yet it stays the same. The core principles and ideas do stay the same, but don't try to construct too elaborate approaches because they tend to be inflexible. I am afraid that your approach is just that: too elaborate. That was my first impression to your first posts. Try to be more flexible. You also mentioned some problems with your trading psychology. You need to fix those because otherwise no system will work for you. Often these problems are caused by undercapitalizion. I am not sure if that's your case, but it could be.

 

Anyway, just my 2 cents. I wish you good luck.

 

Thanks Joe for your feedback.

 

Fortunately, my "bread-and-butter" trading is more of trading break-out or bounce back from trend-lines and other significant levels. And that seems to be working so far for me, even across different market conditions.

 

The method mentioned in this thread was more of an experiment to see how a certain idea would pan out. I was at the time looking for a low-time-requirement, and decent return, pure mechanical method. But as you pointed out, it was not flexible. Which was its problem.

 

So I agree with you on needing to be flexible, and improving on trading psychology.

 

Out of curiosity, may I ask what kind of trading method/system do you employ in your own trading?

[maximillian]

I have found this trading system interesting and have made an indicator for it, which projects the 90 pip levels.

 

The JPY pairs do seem to have been punishing the strategy though as you indicated.