Is this one Spot on?

JonnyT

Senior member
Messages
2,560
Likes
22
Hi guys and dolls,

Another system for you to ponder.

This one has been tested using Spot data from 18th Sept 2003 using hourly bars for the EUR/USD Spot Market. Its rather simple:

1) Trade in the direction of the first breakout of the hours 02:00 to 04:00 US time (i.e. the first breakout from 07:00 to 10:00 UK time)

2) Close the trade at 16:00 (21:00 UK time). There are no stops.

Since 18th September (the start of my spot data) this has returned about 26 pips per day and has a remarkable percentage win rate.

66 winning trades, 29 losing trades, only two occurances of two losing trades in a row!!!

It seems too good to be true.

I haven't 100% confirmed these results as this was a quick and simple test of principle. It looks like these results need to be verified :)

JonnyT
 
Last edited:
I think you may be falling for the classic trap of over
fitting your system to past data...
 
Er I don't think so.

I've done no optimisation whatsover. I've not looked for better approaches, not carried out mathematical analysis or anything.

It's apparent from viewing charts that by 10:00am UK the trend for the day normally sets in. This system just trys to catch the trend move.

If I tried to make the system fit then there are lots of things I could do like stops, profit targets etc.

This system is pure price action and follows the KISS principle so I think its safe from the 'data fitting' argument

JonnyT
 
A friend of mine use to base his system on lunch time and the fact that trends either re-establish themselves after lunch or reverse. He looked for breakouts based on this. It actually worked pretty well for a while. He had stops and he use to take the price about 15-30mins from closing.
 
JT-

I have heard of people trading the 1 hr bar breakout finnishing at 9am UK............. they had a 60 point target with a 60 stop........

U mention 02.00 to 04.00 am ESt time which means the 7.00am - 9.00 am UK time breakout.............


Also, are there any stops ?? or are we just supposed to close the trade at 9 pm UK, regardless of what happens during the day !??


At least thanks for your efforts.................


cheers
Al
 
JT- p.s. forgot to ask which chart u are using for data, since the data for the spot mkt varies slightly between diff brokers.

take care
Al
 
JonnyT,

02:00 to 04:00 US time (i.e. the first breakout from 07:00 to 10:00 UK time)

02:00 to 04:00 US time is 07:00 to 09:00

Is this what you meant ?


Paul
 
To clear up the confusion:

It is the break out of the 3 hour bars from 07:00 UK time. (07:00 to 08:00, 08:00 to 09:00, 09:00 to 10:00)

i.e. the trading range between 07:00 and 10:00 UK time.

The data I have used is from Mytrack and covers the IMM (the biggest) Spot market. It sometimes has a few erronous bars but they are unlikely to have affected this result by any real degree.

JonnyT
 
Jonny

Volatility breakout systems by definition work well to very well in volatile markets like FX at the moment.
The key thing seems to be to a have a "filter" on your system that stops you handing it back when volatility contracts as it will do inevitably.

Heard of 2 people who have done very well for a period out of it, only to hand it back when volatility contracted and they kept trading, one in BTPs (Italian Bonds that used to be on LIFFE) and more recently the DAX - guy made £60k in about two or three months only to hand it all back.
 
Good point DT and I guess it's possible to build in some form of volatility aspect into this system?

The really interesting thing I note from JT's original post is that he's worked this from mid-September last year - over 4 months.

I assumed volatility in the FX markets was a minute-by-minute kind of beastie.

So does this mean that volatility can trend?
 
Volatility hasn't particulary changed since the inception of the Euro.

ATR for example has varied between 0.0085 and 0.0135 but obviously will be higher when the market has moved 25% for volatility to be the same.

I do not think Forex can be considered the same as Index Futures, they are worlds apart.

JonnyT
 
JonnyT

I use Mytrack data, can you tell me what symbol you are using to view this?

SallyT
 
Yesterdays trade:

On the globex future, short at 1.2422 out at .12411 for +11 pips

FXCM short at 1.2436 close at .1244 for + 14 pips (need to take off 4 for spread so really +10)

Al, no stops.

JonnyT
 
Verification?

JT (or anyone!)

Did you or has anyone done any more investigation into this system?

What was the biggest drawdown from it over your 4 months data?

And I didn't quite understand your maths on the FXCM example ...
"FXCM short at 1.2436 close at .1244 for + 14 pips (need to take off 4 for spread so really +10)" ... was there a typo on the closing figure?

Would it be worth putting at least some stop in there, 'cos it feels like if you don't watch it, you could end up with a runaway disaster if a trend reverses and stays that way all day?
 
No, stops have reduced the returns significantly. Which is what they do with most trading systems...

JonnyT
 
Testing Testing!

Thought I'd test this out with CapitalSpreads.com on their simulation platform and post the results to see how it goes, also as a way of seeing if I'm doing it correctly!

Am quite aware that it may have made lots of good pips since Sept last year as most of that time there was a very strong upward trend, so it'd be interesting to separate out the results for a strongly trending time period compared to when it's bobbing about like most of October and mid Jan to early Feb this year.

JT, man of many systems, please point out if I'd doing this correctly if you like for the benefit of all!

Of course, I know you use automated trading signals and buys and sells with APIs directly with ... so I'm trying this with a spreadbetting company to see if us newbie minnows can use this type of system as well, with all the added baggage that a SB company dumps on you (bigger spreads, spreads always biased against trends, etc).

I have to say, JonnyT, that I'm well impressed with your generosity with these simple systems, because they a) give us a head start instead of trying to reinvent the wheel ourselves and b) give us a means of beginning to make money instead of losing it!! Respect to you man - there don't seem to be many out there who realise that there's enough money to be made for us all!!!

OK - adulation over :!:
Back to the testing ...
Between 7am and 10am UK time today using the 1st 3 hourly bars, the low to high I registered from EUR/USD graph from dailyfx.com was 1.2512 - 1.2571.

So I put in auto orders to buy at 1.2580 or to sell at 1.2505. 1st question - were these too far away from the max and min - should I have them set just a pip or 2 away?

Stops were left at the automatically CS-placed points which I can't recall now, but are about 150 pips away by default. I didn't change them. Question 2 - should the stops be further in for damage limitation or should they be further out to allow for large swings, or are they just right?

At 12:34pm, it pushed up and triggered the Buy at 1.2580.

At 9pm UK time (well, 9:14 according to CS, as I was out and left it to end of day), it closed out at 1.2555, making a loss of 25 pips.

JT: Am I doing it correctly? How far away are the numbers from your automated triggers? Would you have made anything on it today?

At least if I'd triggered at 1.2572 I'd have pulled back 8 of those precious 25 pips! Lesson 1 learnt I think!

Better luck tomorrow for the system anyway!

Cheers
3rddawn
 
Between 7am and 10am UK time today using the 1st 3 hourly bars, the low to high I registered from EUR/USD graph from dailyfx.com was 1.2526 - 1.2584.

So I put in auto orders to buy at 1.2584 or to sell at 1.2525. Decided to just go 1 pip out.

Filled at 1.2584, ended the day at 1.2683, a gain of 99 pips.

I did see it rise to about 118 pips profit at one point.

Wondered about something to lock in some profit at some point, rather than risk an early day hike to see it all eroded later on. Only lost potentially 19 pips today, but have seen swings within a day that would lose the lot. Is it worth having some sort of limit once it reaches a certain gain?, say put in a stop at breakeven when it's gained 50 pips or something along those lines?

Does anyone have any thoughts? Have I started trading this correctly?
 
Top