Strategy: trading the daily range

[chiniebwoy]

Good afternoon all,
I had stumbled upon this strategy sometime last year on another forex site. I had another look at it this week and decided to try it yesterday. Essentially, it’s a very simple, straightforward and logical method or strategy and I would be surprised if most persons here are not familiar with it, maybe under another name. The strategy has been around from about 2008.

Briefly, you determine what is 25% of the previous day’s total movement in pips (high - low). We will call this 25% ‘X’. Your buy pending order is then set at the close + X pips and your sell pending order is set at the close - X pips. The SL is recommended to be set at 10 pips away from the opposite order.

So I traded 8 pairs yesterday/today and the results are as follows:

eurusd: + 30 pips
gbpusd: + 40 pips
usdjpy: -48 pips
usdcad: -50 pips
audusd: + 44 pips
eurjpy: + 54 pips
nzdusd: + 30 pips

Net pips: + 100 pips

The eurgbp is currently down 14 pips.

Note though that because I was trying this for the first time I was very conservative in setting my TP and in two cases I used 10% instead of 25%. If I had used 25% my TP would have been hit and more pips would have been made.

So all in all, I think the trial went well.

Of course, there is one big disadvantage or issue with this method and that is the SL. If the rules are adhered to the R:R ratio is not favorable and can easily be 2:1. For the trades above, I simply used a 1:1 ratio. Additionally, trading 8 pairs with such high SL could put one’s account at risk.

I do believe that if you apply some simple PA principles to this method you will have a fair idea of the direction that the trade will go and be able to set SL and TP with greater confidence.

This is in my real account. I will continue to trade the same pairs with this method all of next week and post my daily results.

I would love to get some general feedback, especially from those who are familiar with this method.

David.

[mastergunner99]

One day is not a trial. Tell me how you’re doing after you’re 100th day and we can talk about potential viability.

[chiniebwoy]

One day is not a trial. Tell me how you’re doing after you’re 100th day and we can talk about potential viability.

A fair comment Mastergunner99. Could have been mere coincidence or good luck!
I will nonetheless still keep my promise to post my daily results next week.
Have a wonderful weekend.

David

[RobotCapital]

I thought I’d code up this strategy and test it out. Please don’t take this as personal criticism - its simply an objective critique of your strategy and I offer it as a small contribution to a forum from which I have learned a lot. I hope it helps.

With all respect to the OP, my backtests indicate that you were probably lucky in your first day’s trading and that your strategy doesn’t appear to have any predictive power. You should probably take the money you made on day 1 and save it for when you have a better strategy.

Following is my approach and results:

Firstly, I coded it up exactly as described and then ran a backtest on all the majors from the end of 2010 to the start of 2015. The only pair it was profitable on was NZD/USD, and then only produced a very inconsistent average annual return of 3%.

Secondly, I tried optimizing some of the parameters (X, stop distance, take profit distance) within reasonable ranges. This meant that some trades ran over several days, so I also forced all trades to exit after a maximum of seven days in order to return enough trades over the test period to be able to have confidence in the results. I also optimized the exit time.

After optimizing these parameters separately for all the majors, I found that it was only profitable on EUR/USD, but that it wasn’t a strategy that you would want to trade with real money. Here’s the results from trading 0.01 lots per trade, a starting balance of $100 and leverage of 500:1:

Also note that this equity curve was obtained from the in-sample data (that is, the data used in the test was the same data used to optimize the strategy). In order to estimate how the strategy would perform in real time, you’d need to do an out of sample test. Since an out of sample test almost always produces worse results than the in sample test, I’d recommend abandoning this particular strategy and investing your time in finding a better one.

[Truth]

As Mastergunner said this is no where close to being a trial that is statistically significant.

There should be your homework, how many completed trades are needed to have confidence in the testing of a trading system?

So with 8 trades of 5 wins and 3 losses you have an 62.5% winning average.
The pip count you gave gives a win/loss ratio of 1.06 (39.6/37.3).

This means you have a razor thin positive expectancy, but 8 trades is no where close to being a valid sample size, so we simply do not know at this time.

Another question for you, given your 62.5% winning average, what do you expect would be a long streak of losses that will occur at some point using this method?

[Arbitrager_on_Acid]

How do you take profit?

-Adrian

[Arbitrager_on_Acid]

If you set a tp at 1:1, it is 75% of the previous day’s range above the previous day’s close minus ten pips.

That should over time produce a break even system that will lose costs.

The system and its opposite will have the same expectancy.

-Adrian

[chiniebwoy]

@RobotCapital

I thought I’d code up this strategy and test it out. Please don’t take this as personal criticism - its simply an objective critique of your strategy and I offer it as a small contribution to a forum from which I have learned a lot. I hope it helps.

With all respect to the OP, my backtests indicate that you were probably lucky in your first day’s trading and that your strategy doesn’t appear to have any predictive power. You should probably take the money you made on day 1 and save it for when you have a better strategy.

Following is my approach and results:

Firstly, I coded it up exactly as described and then ran a backtest on all the majors from the end of 2010 to the start of 2015. The only pair it was profitable on was NZD/USD, and then only produced a very inconsistent average annual return of 3%.

Secondly, I tried optimizing some of the parameters (X, stop distance, take profit distance) within reasonable ranges. This meant that some trades ran over several days, so I also forced all trades to exit after a maximum of seven days in order to return enough trades over the test period to be able to have confidence in the results. I also optimized the exit time.

After optimizing these parameters separately for all the majors, I found that it was only profitable on EUR/USD, but that it wasn’t a strategy that you would want to trade with real money. Here’s the results from trading 0.01 lots per trade, a starting balance of $100 and leverage of 500:1:

Also note that this equity curve was obtained from the in-sample data (that is, the data used in the test was the same data used to optimize the strategy). In order to estimate how the strategy would perform in real time, you’d need to do an out of sample test. Since an out of sample test almost always produces worse results than the in sample test, I’d recommend abandoning this particular strategy and investing your time in finding a better one.

RobotCapital,
absolutely no offense taken. I’m humbled that you took the time out to do the test.
thanks for such a comprehensive response and I appreciate the work that you have done on the backtest. The language that you used has gone way over my head, but I did get the message at the end of it all: it’s not a good strategy!

Call me bullheaded, but I’m going to see where 5 days of it will take me. I wont necessarily trade all the pairs, especially if the candlesticks look dubious. I will just pick my trades carefully.

I’ve set my trades for today and will post results tomorrow.

[chiniebwoy]

As Mastergunner said this is no where close to being a trial that is statistically significant.

There should be your homework, how many completed trades are needed to have confidence in the testing of a trading system?

So with 8 trades of 5 wins and 3 losses you have an 62.5% winning average.
The pip count you gave gives a win/loss ratio of 1.06 (39.6/37.3).

This means you have a razor thin positive expectancy, but 8 trades is no where close to being a valid sample size, so we simply do not know at this time.

Another question for you, given your 62.5% winning average, what do you expect would be a long streak of losses that will occur at some point using this method?

I agree…can’t give a verdict after one day.
The logic behind this method is that if a pair moves say 180 pips in a day, if it moves at least half (90) of that number (uniformly), then the trade will go in your favor.

David

[chiniebwoy]

If you set a tp at 1:1, it is 75% of the previous day’s range above the previous day’s close minus ten pips.

That should over time produce a break even system that will lose costs.

The system and its opposite will have the same expectancy.

-Adrian

Hi Adrian,
not sure if I agree with you. A 1:1 tp will be 50% above previous day’s close. Agreed?
The strategy is based on a minimum 50% move in either direction in order to gain pips.

David

[chiniebwoy]

How do you take profit?

-Adrian

Adrian,
profit is calculated as 25% of previous day’s range. If previous day’s range is 200pips, 25% is 50. You go long at 25%/50 pips from the close and tp is 40 pips from that. Do the opposite for short trades.
For SL, you set it at 10 pips from the opposite order (+ 10 for short, -10 for long). Risky stuff.

David

[Arbitrager_on_Acid]

Hi Adrian,
not sure if I agree with you. A 1:1 tp will be 50% above previous day’s close. Agreed?
The strategy is based on a minimum 50% move in either direction in order to gain pips.

David

I must be interpreting it wrong. Suppose the previous day data was:

High: 1.0971
Low: 1.0768
Close: 1.0947

That is a 203 pip range. 25% of 203 is 50 pips (rounded down to nearest pip).

So you would buy at 1.0947 + 50 pips which is 1.0997. You would sell at 1.0947 - 50 pips which is 1.0897.

The SL is recommended to be set at 10 pips away from the opposite order.

Your stop for the long trade would be ten pips above your short entry: 1.0897 + 10 pips which is 1.0907. So your long entry is 90 pips from your stop (1.0997 - 1.0907). If you set your tp at 90 pips above your entry to get 1:1, it is 1.0997 + 90 pips = 1.1087. 1.1087 is 140 pips more than the previous close of 1.0947. 75% of the previous range of 203 is 152.25. 152.25 minus ten pips is 142.25. So you need the price to move 75% of the previous day range less ten pips to hit that tp.

Am I misunderstanding your rules?

-Adrian

[chiniebwoy]

I must be interpreting it wrong. Suppose the previous day data was:

High: 1.0971
Low: 1.0768
Close: 1.0947

That is a 203 pip range. 25% of 203 is 50 pips (rounded down to nearest pip).

So you would buy at 1.0947 + 50 pips which is 1.0997. You would sell at 1.0947 - 50 pips which is 1.0897.

Your stop for the long trade would be ten pips above your short entry: 1.0897 + 10 pips which is 1.0907. So your long entry is 90 pips from your stop (1.0997 - 1.0907). If you set your tp at 90 pips above your entry to get 1:1, it is 1.0997 + 90 pips = 1.1087. 1.1087 is 140 pips more than the previous close of 1.0947. 75% of the previous range of 203 is 152.25. 152.25 minus ten pips is 142.25. So you need the price to move 75% of the previous day range less ten pips to hit that tp.

Am I misunderstanding your rules?

-Adrian

Morning Adrian,
you are right up to the point where you speak about tp.

Your tp would not be 90 pips above your entry, it would be the same 25% or 50 pips above your entry. That’s 1.0997 + 50 pis = 1.1047. Now 1.1047 is 100 pips away from the previous close of 1.0947. So you are really banking or hoping for a 50% move in pips of the previous day’s range. Makes sense?

The placement of the SL is as you have outlined.

Yesterday’s trades have all hit their TPs/SLs. I was a bit more judicious in the currencies I traded. Additionally, I used an indicator to give me an idea of the direction of movement so I only traded in one direction.

eurusd: +38
usdcad: +24
audusd: -40
eurgbp: +38
eurjpy: -33
nzdusd: -30
usdchf: +44

Total wins: 144 pips; total loss: 103 pips.
Net: +41 pips.

I will go back to my original way of placing both a long and short trade on each pair.

David.

[Arbitrager_on_Acid]

Your tp would not be 90 pips above your entry, it would be the same 25% or 50 pips above your entry. That’s 1.0997 + 50 pis = 1.1047. Now 1.1047 is 100 pips away from the previous close of 1.0947. So you are really banking or hoping for a 50% move in pips of the previous day’s range. Makes sense?

It does, but it is not an R:R of 1:1.

If my:
Entry is 1.0997
SL is 1.0907
TP is 1.1047

I have a 90 pip stop loss: (1.0997 - 1.0907 = 90 pips)
I have a 50 pip profit target: (1.1047 - 1.0997 = 50 pips)

That, of course, is not a 1:1 risk to reward ratio (90:50).

I realize you said from the very beginning:

Of course, there is one big disadvantage or issue with this method and that is the SL. If the rules are adhered to the R:R ratio is not favorable and can easily be 2:1. For the trades above, I simply used a 1:1 ratio. Additionally, trading 8 pairs with such high SL could put one’s account at risk.

But in order to get to 1:1 you either have to tighten the stop or lengthen the TP. Which did you do and how?

-Adrian

[chiniebwoy]

It does, but it is not an R:R of 1:1.

If my:
Entry is 1.0997
SL is 1.0907
TP is 1.1047

I have a 90 pip stop loss: (1.0997 - 1.0907 = 90 pips)
I have a 50 pip profit target: (1.1047 - 1.0997 = 50 pips)

That, of course, is not a 1:1 risk to reward ratio (90:50).

I realize you said from the very beginning:

But in order to get to 1:1 you either have to tighten the stop or lengthen the TP. Which did you do and how?

-Adrian

Hi Adrian,
I understand your question now.

For me, I do not follow that SL rule because of that same 1.8:1 R:R ratio.
All I simply do is this: for long trades I minus 10 pips from the close and use that as the SL.
For short trades I add 10 pips to the close and use that as the SL. I know it’s not very scientific, but I’m more comfortable with that.

So in the above example my SL for a long trade would be 1.0947 - 10 = 1.0937.

So to answer your question: I tighten the stop. Does this cause you to get stopped out early then probably see the trade revers and move on to the TP? I’m sure it does. Maybe this is why this strategy will not find favor with many lol.

Results for today:
eurnzd: +30
eurchf: +7
eurjpy: +27
gbpusd: +14
eurusd: +35

Total pips: +113

I 'll soon figure out how to post a shot of the account page.

David

[chiniebwoy]

It does, but it is not an R:R of 1:1.

If my:
Entry is 1.0997
SL is 1.0907
TP is 1.1047

I have a 90 pip stop loss: (1.0997 - 1.0907 = 90 pips)
I have a 50 pip profit target: (1.1047 - 1.0997 = 50 pips)

That, of course, is not a 1:1 risk to reward ratio (90:50).

I realize you said from the very beginning:

But in order to get to 1:1 you either have to tighten the stop or lengthen the TP. Which did you do and how?

-Adrian

I should also add that if the daily range is less than 100 pips, I do not trade the pair.

David

[Arbitrager_on_Acid]

Ah OK. Got it. So another question: have you thought about or tried filtering for direction? Suppose you took only longs above a certain moving average (like a 200 day) and only shorts below. Something like that?

-Adrian

[chiniebwoy]

Ah OK. Got it. So another question: have you thought about or tried filtering for direction? Suppose you took only longs above a certain moving average (like a 200 day) and only shorts below. Something like that?

-Adrian

Adrian,

The person who wrote about the strategy did look at a number of indicators to do just that. One of them was the COG (I use MT4 and it’s there) applied to the 1hr chart. He would then set a pending trade in the general direction that the central blue line was heading. Not two orders, just one. This indicator and a few others that he tried, he found to be inconsistent. So after a lot of tweaking he pretty much went back to the original way of setting his trades. This was the most profitable way for him.

With respect to the 200 sma applied to the daily, I kinda view its use different from how most persons do. Call me a gambler; or call me stupid. For example, if you apply it to the eurusd I would only be taking short trades. Trades with the trend. Nothing wrong with that; after all, it’s a fundamental rule. But then you miss the intraday ranging movements, the retracements etc. So using it would not take full advantage of this method. Just my $0.02.

David

[Arbitrager_on_Acid]

Taking a step back, it seems the general hypothesis behind this sort of implementation is that prices will generally cover much of the same range as the previous day. Would you say that is right? Or is there something more?

-Adrian

[Forexmike]

hi guys,

i also tried this strategy some time ago…the original version didn’t work out really well…but then there was a variation and that one work much better at the time, don’t know if it would work still today…here i give you the links…

original version
Advanced system #14 (Simple trading with Daily range) | Forex Strategies & Systems Revealed

variation
Pendolan’s Journal