Weekly Range Method

I wanted to design a method so simple that my tend year old son could trade it and didn't take much time. I have visually backtested it and it seems to work most weeks or else doesn't produce a trade.I'm wondering if it's just too mechanical to work long-term. I would love some feedback and any suggestions for improvement would be greatly appreciated
<?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" /><st1lace><st1laceName>Weekly</st1laceName> <st1laceType>Range</st1laceType></st1lace> Method – Version 1

Tested on GBP/USD.

After the market closes for the week, mark the past 2 weekly ranges(high/low) and the halfway point of the most recent weekly range(which I’ll refer to as the weekly median).

Your directional bias is determined by comparing these weekly ranges

FADE: If the most recent week’s range is greater than the previous week’s range, then look to fade the most recent weekly direction. Trade it in this way:

If last week was bearish, Buy at Friday’s median, place your limit at the weekly median and place your stop at the weekly high.

If last week was bullish, Sell at Friday’s median, place your limit at the weekly median and place your stop at the weekly low

TREND: If the most recent week’s range is smaller than the previous week's range, then look to trade in last week’s direction. Trade it in this way:

If last week was bullish, buy at the weekly median, place your limit at the weekly high and place your stop at the weekly low.

If last week was bearish, sell at the weekly median, place your limit at the weekly low and place your stop at the weekly high. Do not chase the price. If it doesn't pullback to the median, just let it go and don't trade. The idea when trading with the trend is to enter on a pullback and avoid whipsaws.

The whole method in one sentence: if last week was an expansion week, trade in from the outside and if last week was a contraction week, trade out from the inside.
I hope this makes sense. I'll follow with examples if anyone's interested.



Ken
<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o> </o>

Well, a system this simple and mechanical cries out for an EA, so I'll whip one up.

Should work well for backtesting too.

EDIT: I'd like to see some examples so i can be sure I interpret your rules correctly.

It sounds simple enough. I'll check into it also as a long term system.

-The Drizzle

Thanks very much, guys. These examples are all for GBP
Example1: Most recent week was 3-8 Sept 06. High= 1.9082 Low= 1.8629 so its range was 453 pips and the weekly median is 1.8855(Low + 50% of the range).
It opened(1.9055) higher than it closed(1.8653) so it was bearish.
The previous week was 27 Aug - 1 Sept 06. High = 1.9093 Low= 1.8862 so its range was 231 pips.
Because the most recent week had the larger range our directional Bias is to FADE. Since the most recent week was bearish, our bias must therefore be bullish.
The friday high was 1.8753, its low was 1.8629. The Friday median was therefore 1.8691 and this will be our entry price for the trade. If the price doesn't come back up, we don't trade. It does come back up and triggers the order at 1.8691. The profit target is the weekly median: 1.8855. The stop loss is placed at the weekly low:1.8629. We are stopped out on sept 12 for a 62 pip loss. It then rallies to a high of 1.8920, far beyond the target of weekly median - 1.8855. An entry at friday's high(instead of the median) would have completely avoided the whipsaw but I'm getting into curve fitting here. Every method is going to have losses and I'd rather lose 62 than 124 which could have been the case if the whipsaw had've taken out the high instead of the median.
Another example to follow

I think your 10 year old must be a genius then...

Example 2: Again GBP
Most recent week(27aug to 1sept) opens at 1.8870 closes at 1.9055 so it is a bullish week. Its High is 1.9093, its low is 1.8862 so its range is 231 it's median is . The previous weeks high was 1.8997 and its low was 1.8806 so its range was 191. So, as with previous example, last week's range was bigger than the previous week's range so our bias is to FADE.
The MRW was bullish so the FADE bias means we are looking to go short at Friday's median: 1.9013. The profit target is the weekly median:1.8977. The stop is at the weekly high:1.9093. The trade is triggered on 5sept and closed a few hours later for 43 pips profit. It did continue to fall to a low of 1.8628.
Food for thought there as well in terms of optimisation. We take 43 pips profit in a week where the price moves another 385 pips. We risk 84 pips to acheive this result. Hmmm.
Example 3: one week further back.
MRW(20aug- 25aug) opens at 1.8814 closes at 1.8870 so it is a bullish week. Its High is 1.8997, its low is 1.8806 so its range is 191 and its median is 1.8901. The previous weeks high was 1.9025 and its low was 1.8776 so its range was 249. So, unlike the previous examples, last week's range was smaller than the previous week's range so our bias is to TREND.
The MRW was bullish so the TREND bias means we are looking to go long at weekly median:1.8901 . The profit target is the weekly high:1.8997. The stop is at the weekly low:1.8806. The trade is triggered soon after the open and closes within a few hoursfor 96 pip profit. We miss out on another 100 pips by closing at the weekly high but at least the RR ratio is down to 1:1 on this one.
There is obviously room for improvement with this concept but I'd like to know if the prototype works longterm before trying develope it into a more efficient model.


Ken

I like your descriptions. They are great. But, do you think you could post a chart? To do so, just keep it 600 x 600 and use the attach to post button at the bottom of the screen when replying. If you want to post a larger image, just zip it first... thanks Scott

Ken,
Thanks for sharing.

Too many traders get tunnel vision on the intraday charts and often learn to late how much better the daily and weekly charts can be in identifying sets-ups, putting on trades, and implementing effective money management techniques.

I hope a lot of the newbies take note of this thread like Dialist's "Observe" thread in the Beginner's Forum:

http://www.forexfactory.com/forexfor...ead.php?t=8058

Good trading,
MJ

I have created an excel spreadsheet to make this a little more visual. Please check me and see what I have done is correct but I think it is. Hopefully this will help people to test this theory. I'm gonna try it on demo next week and see what happens.

jhathawa, I reckon who've got it right. Thanks very much. Forward testing in real-time is the true test for any method. We'll see how it goes.



Ken

I'm sure this could be made into an EA but would take some work. Would be nice for backtesting to see if forward testing is really worth it. If it bombs over a 2+ year period forward testing really wouldn't help. Any coders who can take this on ?

....is gonna take years.

I'm not a programmer....more like a lawyer. But I've been mucking around with custom indicators for a little while. I made this indicator for the "weekly fade" method. It only works properly in 1H charts at the moment. It's not an EA so it won't backtest the method but it shows (at least on my screen) what the trading conditions under this method would be.

Colors are customizable. At the moment, the Aqua line shows the entry point for the week in which the line appears. The description, also in Aqua, tells you if we're going long or short and whether it's a Fade trade or a Trend trade. The yellow dash-line is the take-profit. the Red dash-line is the stoploss.

An additional variable that can be entered when you set up the indicator is WinLoseRatio. If set at zero, all the weeks where the method's conditions are met will display entry, exit and stoploss lines. If you set it at one, we only enter a trade if expected profit (spread not included) is equal or greater than our potential loss (if stoploss is hit). If set at 2, we're looking for a profit of twice the risk...and so on.

Vertical red lines appear just before each Monday to divide our weeks.

Obviously I can't guarantee the accuracy of this indicator (only did it on the spot).

Cheers
L

p.s. One thing I've noticed is that some weeks generate more than one potential entry, ie, price crosses our entry point, hits the stoploss and then comes back to generate another possible entry. This of course is not limited to this method; it occurs in most methods.

Comments are welcome.

Found a major bug in the indicator in the way it calculates entry and exit points (uses Open and Close instead of High and Low for median calcs)...

fixing....

Here it is

Thanks you for this contribution. I Like the idea of simple indicator thats work and require less time in front of the monitor. I think long time analysis like this one poses less risk of losing. Keep up the good work and lets improve on this.

No problems. I'll improve it to work in other time frames too. Although for weekly trades, 1H is probably small enough anyway. In the meantime, I now attach one further fix.....Just the name of the entry line was wrong (it said "daily low")....that's because I based the script on an earlier indicator I had done, which calculated weekly pivot points for backtesting purposes.

First result in....unfortunately it was a loss, although small. These are based on Interbank feeds. Others may have slightly different numbers with other brokers.

SELL GBP @ 1.9029

STOP @ 1.9062
LIMIT @ 1.8899

Stopped out for 33 pip loss. Until next week.....

Any chance of posting the source code, rather then the compiled version of this indicator?