Daybreak

[I]***IMPORTANT CHANGES TO THE DAYBREAK APPROACH! If anyone is interested in trading breaks of previous day’s high/low, understand that the original Daybreak method has gone through several stages of evolution. It is still a “set and forget” method and I am currently trading it live, albeit small. If you would like to look into or test this method, here is how I am trading it.
1. I am still entering on a break of yesterday’s high or low. I am using the Daybreak Sprit ea to test and set orders. I manage that ea with sl’s and tp’s and set the close time to 30 so trades are not closed or canceled eod.
2. The Daybreak Currency Portfolio is limited to these pairs with the following parameters:
GBP/JPY TP 70 SL 30
EUR/JPY TP 70 SL 30
3. I am also trading gold, xau/usd, in a separate account:
XAU/USD TP 600 SL 300

Regarding results, we would have to say undetermined at this time. Proceed at your own risk. One day I may solve the mystery of why this backtests so well, but, when trading, I have yet to achieve anything close to backtest results. I will also suggest you look at a variation of this approach posted by VJ Parmar on his thread, Parmar 3P Trading System.

A trading method for those who can’t watch the market all day and, maybe, for those who can.[/I]

DAYBREAK

PIP EQUITY CURVE, JANUARY 2007 THROUGH MID-JUNE 2012,
GBP/JPY, EUR/USD, AUD/USD, AND USD/JPY.


EVOLUTION

It starts with a casual observation. Look at daily bars. Does it seem to you that when the high (low) of yesterday is broken, the close today is often also higher (lower) than the break point? Here is a shot of GBP/JPY for May 2012 where we just mark the outcome from the breakout relative to the close.


We can use a spreadsheet to ask a simple question: If all we did was go long on a break of yesterday’s high and go short on a break of yesterday’s low and we held positions until the close of today, how would we come out? We will define a breakout as a pip above the high or below the low.

Using IBFX data from January 2007 through mid-June 2012 gives us five and a half years or about 66 months of data. To replicate these results, you probably already know that you can save the data from a chart by clicking “file” “save as” and saving it to a spreadsheet. We are including the GBP/JPY for a simple reason, it moves, dramatically. For GBP/JPY, the excel formula to look at long trades is =IF(C11>C10,(E11-(C10+0.01))*100,0). That is, if our high of today (C11) is greater than our high of yesterday (C10), then take the close of today (E11) and subtract our entry point (C10+.01) and multiply the result by 100 so we can read our outcome in whole pips. And, here is your excel formula to check the short trades: =IF(D11<D10,((D10-0.01)-E11)*100,0). For pairs with four decimals, add two zeroes to your breakout spread and to the multiplier that gets you to a whole pip number. Here are the results for our portfolio just taking the breakouts and holding until close:

GBP/JPY: Long +6,714; Short +11,890; Total +18,604
EUR/USD: Long (-197); Short +5,117; Total +4,920
AUD/USD: Long +5,107; Short +2,191; Total +7,298
USD/JPY: Long +3,571; Short +3,778 ; Total +7,349

Without considering spreads, slippage, and other negatives, trading this portfolio would have yielded more than +38,000 pips, i.e., just under +600 on average per month. That looks pretty good, but I think it can get a little better and less risky. First, a couple of entry tweaks seem to be in order. 
  1. Go long at yesterday’s high plus one pip.
  2. Go short at yesterday’s low minus one pip.
  3. Exception: For Monday trades, go long at yesterday’s high plus 11 pips or go short at yesterday’s low minus 11 pips.

The reason for the Monday exception is that we are going to have a limited data base from Sunday trading. The Oanda platform is going to give us about 10 hourly bars from 14 to 23, platform time, and IBFX is only going to give us one or two, usually 22 and/or 23, platform time. See what your own datafeed provides. In any case, you will probably agree that we would feel safer asking the market to move a little before we jump in on Monday.

Some traders have a valid aversion to Monday trading and may want to skip that day entirely. By now, we should all know that the best days to trade, from the standpoint of price movement, are Tuesday through mid-day Friday. However, there are some opportunities on Monday, like the two GBP/JPY trades in April, 2012. One on the 23rd and one on the 30th for about +80 each using DAYBREAK. Or, that phenomenal Monday last October 31 where the short trade lost, but the long trade could have made +367. Your call.

STOP-LOSS

The only other matter is that of an emergency stop-loss. As indicated by the results from our spreadsheet, we want to give our positions every opportunity to go to close for a profit. On the other hand, we don’t want to look in on a trade that is down -300 or -400 pips. Emergency stops are intended to be just that, for emergencies only. Only those days when the market sucks us into a position, then turns like a rabid wolverine. Been there, right? We can determine these stops for each pair based on relative movement and the probability that the stop would occur. Until I can get in more research on this, we will use the following emergency stops:

GBP/JPY -200
EUR/USD -150
AUD/USD -125
USD/JPY -100

By adding these emergency stops, we are avoiding only the extreme moves against us and we are getting a significant improvement in our expectation for overall profit. To get an estimate of the difference these stops might make, do this. Take the original results and limit each day’s loss to the emergency stop for that pair, i.e., if my loss on a GBP/JPY trade is more than -200, show that trade at -200. Keep any other number. Spreadsheet formula: =IF(G11<-200,-200,G11). Our new projections (and where the data for the equity curve above came from) show a potential of more than +50,000, close to +750 on average monthly. These numbers have eliminated Sunday trades, but the spreadsheet will still enter Monday trades at +/- 1.

GBP/JPY: Total +24,361
EUR/USD: Total +7,395
AUD/USD: Total +9,056
USD/JPY: Total +8,599



SOME PRACTICAL CONSIDERATIONS

The spreadsheet study has a few limitations, but if someone is interested enough, this one looks simple enough to program for more accurate results. The spreadsheet will treat Mondays like any other day and enter at +1 or -1. If we trade Mondays at all, we will be holding for +11 or -11. There may also be occasions where our stop-loss is hit, but the market recovers to show a more favorable outcome. Since our stops are wide and get taken out infrequently, I don’t believe this is a major factor in our pip tally.

To hedge or not to hedge… For this system, I am always assuming my reader can’t watch the market all day and, for the most part, wants something to set and forget. On days when we get two signals, the spreadsheet will trade both, effectively putting the trade in a hedge, limiting our losses to yesterday’s range. About two-thirds of these trades come on Monday, but if we eliminate Monday trades, we lose about +10,000. Our reference or control trades on the portfolio showing us +49,000 would be including the hedge trades. Therefore, in order to have yielded the approximate results for the test period, we would have to be taking both sides of the trade on those days when two were triggered.

From a practical standpoint, most brokers do not allow hedges. We have a few options. Since the portfolio shows the highest yield with the hedges, it is preferable to hedge. In order to hedge, you will either have to be with a broker who permits those orders or have two accounts. Some brokers also have sub-accounts which work like separate accounts. You would place all your buy stops in one account and all your sell stops in the other. Should both sides of a trade be initiated, you will be hedged between the two accounts.

If you can’t or don’t want to open additional accounts and if your broker doesn’t permit hedging, you can do this. Enter both sides of the order. When one side is opened, the platform will not allow the second. Be sure your broker operates this way. Mine do, and if I attempt to enter a second order while one is open, I get a pop-up saying, “Hedge not permitted,” and the order is unfilled. Another alternative is to enter only one side of the entry stops. How do you decide which one, long or short? I ran a test choosing the entry closest to the open, seemed most likely to be triggered. Yield about +39,000.

DISCRETIONARY TRADING

I commented on this area in a response post later in the thread, but thought it was important enough to also bring up here. For the sake of those who might have to set it up and go to bed or work, for testing purposes, and for the space available on bp, I kept these rules pretty static. In real trading, if we are watching the screen, I don’t think we can or should avoid some discretion. One example would be when you see the close of the day coming at or near the high of the day. If you are already in a long position, why not just leave it open with a tight stop as we enter the new day? Or, as jj suggested, if entering a new trade, require a few additional pips movement before entering. We have seen times where the pair closes near the high/low and just doesn’t have the oomph to continue next day. Sometimes it does. For a great example of this, look at June 28 and 29, this year. We were in a short on the 28 which was successful. Had we put a tight stop on that short as it continued into the 29th, we could have saved the run up against us and even profited by closing out the short and taking the long as it broke.

Another discretionary decision might be when we have seen a good run in our favor. Then, the market enters that afternoon sleepy time where hourly ranges get small and we see a drift off our high point. Close, bank the money, go for a cup of coffee. Then, hope it doesn’t hit another big run which would have been in your favor while you’re gone. Another suggestion from jj was that we not wait until the actual closing bar, but exit the trade maybe two or three hours earlier.

Still another decision might come if we look in on a chart, no trade has been triggered, and it is getting late in the day. If we had in mind to close by end of day, why give a trade only an hour or two to develop? Cancel the orders, especially Friday. Go into the weekend flat and get it off your mind.

Discretionary decisions are hard to teach, possibly as often wrong as they are right, and just come with experience. If you are the type to panic, move stops, curse the market, then you need to keep it mechanical. If you can keep your head, make some judgment calls. In any case, maintain a trading journal, so you can look back and see what you did and what on earth you were thinking at the time.

Read more: http://forums.babypips.com/free-forex-trading-systems/45422-daybreak-7.html#ixzz1zlOmxp9H

1 Like

This looks interesting pipwoof. I am going to do some analysis tomorrow on this. I am with FXCM, will be interesting to see if the numbers look the same. Also will look at adding more pairs. GU, EJ, even NU have decent ATR’s. Nice work!

John

john, thanks for your interest. apologies to all. i should have emphasized that this is a work in progress. haven’t traded it yet. i am especially concerned about my estimates for how many times stopped out. may have missed something on that and will appreciate any help.

wow! This is cool and clean. But what if we try to know the bias and place just one order in the direction of the bias. I learnt this from my mentor (Nikitafx) on price action for dummies that if yestaday was buy today will be buy using our support or resistance for entry depending on bias. So in my own opinion I will suggest that we place a buy order on yestaday’s high if yestaday was buy and a sell order on yestaday low if yestaday was sell. What do u think pipwoof and others?

Less I forget I’m going to vote your system as the best system for the month of june its almost my style of trading

sep, do appreciate your support and input. my studies are often, at best, clumsy, but i did try to take a quick look at requiring a prior-day bias. it looks like it would reduce total trades, cutting about half the losers and half the winners. this may be helpful for those who want to emphasize risk management. thanks.

all four pairs triggered last night. aud/usd long @ 1.0089, gbp/jpy short @123.84, usd/jpy short at 79.33, eur/usd short at 1.2443. we’ll see how they hold up as the day progresses.

What about TP levels?

Cry

interesting that you bring that up, cry. i have an exit strategy i want to use for the may contest in conjunction with the daybreak entry method. different sl, starts with three positions. one is a tp, the other two…

Would be quite fantastic if this actually works.

Keep us posted of your results, and I might just pick up this one for a try in a couple of days.

Hey Pipwoof - this is a Very nice system - simple and easy to implement (I LOOOVE IT!) I’ll try and work out entry and stop loss tonight and see how it comes - defo the top system in the last year that I have seen in my research :smiley: congratz

A comment after a brief analysis on the charts: in pips u can get good results, but RR is quite bad and roi too with a 200 pips SL.

Cry

i meant for the JULY contest. heh. can’t enter may…already won.

amero, thanks for your nice comments. don’t get too excited until the bp jury has had a chance to weigh in. i may have missed any number of things.

hi pipwoof

i like the look of this method, however just a few points, firstly, if you are putting a 150 pip SL on e/u are you then looking for a profit of at least 300 pips, giving you a 1>2 risk reward?

secondly, have been looking at using wkly pivot points to help determin where to put your SL, once you get the break, high or low. the reason being it may be that the stop loss may not need to be so large, what do you think ?

Podz

Oh! I see, so does it mean one will set 2pending order till the close of the day or u cancel one when one trigger?

sep, the spreadsheet results will show to have taken both trades, long and short, if two are triggered in a day. the net of this is a hedge position where your maximum loss is the range of yesterday. in real life, i don’t think most brokers allow hedging. if we are trading one account, enter both long and short entry stops. the first one hit will be the trade for the day. it is possible that the emergency stop will take out the first trade and allow the second to proceed. not likely, though, because of the wide stops. another approach is to use two accounts or a sub-account. enter all long trades in one account and all shorts in another.

for example, my bad, i made a mistake listing the signals for today. the eur/usd actually started out long today at 1.2508. it went short later at 1.2443. so, if both trades were taken, the euro is actually in a hedge and will lose just over -50 at the end of the day. trade the hedge if you can because if we had only the long euro today, it would be losing substantially more than -50.

cry may have some good points here. thanks. looking for more info on the downside of this approach.

Does anyone with some programming knowledge wanna code an ea to back test this? Ive been looking over datat on GBP/JPY and EUR/USD. Using the original settings and formulas for excel it seems that the smaller you set the stop loss the greater the results. I think its because of the use of daily bars and know being able to tell… atleast from the sheet i set up using GBP/jpy daily data going back to 2008. Does anyone wanna share their backtesting sheets?

can we upload .xls files? because it keeps giving me errors

stefan, some of you have already spotted the limitation of using a spreadsheet with daily data. The limitation is that we are looking only at the daily high, low, and close. If I limit my loss to -50, that’s cool, but there will be some days where I would have started my trade with a -50 stop-loss and price action would have taken out that stop followed by a reversal to what would have been a better close had I stayed in the trade. For example, on March 9, 2012, we would have gone long at 129.24. After climbing to a high of 129.56, the market then plunged, definitely taking out a -50 stop. The market spent the rest of the day asking us why we’re even bothering with stops as it climbed back to a close of 129.16. Because of our spreadsheet limitations, this trade is not going to show the reality of a -50 loss. It will show only a -8. My feeling is, even allowing for these and other limitations of the spreadsheet study, we still have enough net pips to get our attention and motivate us to work this premise toward an acceptable trading method. A proficient programmer, of course, can just build a study using a shorter timeframe and get more accurate results. Since I don’t know one end of a rocket from the other, I have to use what I know how to use. In a brief study of the gbp/jpy, I used a -50 sl and compared the spreadsheet results side-by-side with what I got looking trade by trade for the first quarter of 2012. The spreadsheet showed net 1666. By hand, I got 1065.

Hi Folks,

I have been compiling the data for these 4 pairs plus 4 more. I am with FXCM and my data only goes back to sept 2008. I have come up with about a 600 pip average per month with 8 pairs, which is what pipwoof came up with using only 4 pairs. you would think with 4 more pairs that it would be considerably more than that. I am attaching the file to this post for those of you who want to look at it. One thing I have noticed is on some days that both orders are triggered, they both end up losses at the close. I think that on the days where we had an up candle the previous day, we may want to look at longs only and vise versa for down days. I think to really nail this strategy will take a bit of work and feedback from all interested.