5 Min Index Trading

Hi Jack - with reference to you c down, im looking at an alt here

dl.dropbox.com/u/40497882/ftse19thfeb.JPG

Actually, yesterday (and today) I got drawn into trading waves as they seemed so clear so I was expecting a bounce off 6302 this am and bought that. Sold it just inside the 61.8% Fib and gap at 6330 and then went short again since I had that as a wave 2 up of either C or 3 down.

Anyway, going back to systems, I was searching updata's system library and while they have a simple MA crossover system, I couldn't get it to short and cover short as well as sell and go long so the results didn't tally. In fact, of about 20 systems or so that I tried with systems tests on the FTSE 5 min (and some on GBPUSD too), most of them failed to produce a profit, even when trade costs were reduced to zero.

One of the few which showed a profit over both the period I've been trading the MAs (since 11 Feb) and also over the last 2000 periods was a similar system but using MAs on RSI rather than price. Here's a chart showing the price chart, RSI chart and an equity line (which began at 10000 on 11 Feb):



dl.dropbox.com/u/20815047/FTSERSIMAsystem.gif

And here's a screenshot showing the detail of the trades:



dl.dropbox.com/u/20815047/FTSERSIMAsystemdetail.png

I've looked through the trades and it seems to be a much smoother system with fewer false signals and fewer trades in general. As you can see, yesterday's buy signal was not phased by today's drop first thing and remained valid (as it still does) thereby keeping you in the trade. When I checked some other stats on this system I think (from memory) that it showed an annualised return of c. 337% which ain't too shabby if correct!

What is very clear, as has already been noted, is that few systems work during sideways moves. Apparently, the only type of system that does work then are what are termed "trend fading systems". There are a few of these in the updata system library but not one of them showed a profit during the period tested.

I'll continue to search systems and test them (there's also an optimisation facility) to look for something better but, for now, the RSI based MA crossover system is the best I've found. There was one other with similar results (called the Cliche System!) but that produced over 250 trades during the same 4 day period!

Thanks for your input. It's much appreciated [/quote]

your welcome Jackozy - and i'll check it out thanks..

i had a sell all day on ftse yesterday, yes it went up but every bounce got sold into.... macd remained below 0, nothing fancy in it, was all abt price and using in conjunction with the MA's and Macd. indicators imo lag price, today price was important - the indicators just confirmed my buy.

today Ftse has remained in Buy and i too am long. incidentally i'm looking for final parabolic thrust now 6380 - 6420 before a deeper correction set's in. we shall see.

all the best mate

Thanks wavesurfer.

Actually, yesterday (and today) I got drawn into trading waves as they seemed so clear so I was expecting a bounce off 6302 this am and bought that. Sold it just inside the 61.8% Fib and gap at 6330 and then went short again since I had that as a wave 2 up of either C or 3 down.

Anyway, going back to systems, I was searching updata's system library and while they have a simple MA crossover system, I couldn't get it to short and cover short as well as sell and go long so the results didn't tally. In fact, of about 20 systems or so that I tried with systems tests on the FTSE 5 min (and some on GBPUSD too), most of them failed to produce a profit, even when trade costs were reduced to zero.

One of the few which showed a profit over both the period I've been trading the MAs (since 11 Feb) and also over the last 2000 periods was a similar system but using MAs on RSI rather than price. Here's a chart showing the price chart, RSI chart and an equity line (which began at 10000 on 11 Feb):



dl.dropbox.com/u/20815047/FTSERSIMAsystem.gif

And here's a screenshot showing the detail of the trades:



dl.dropbox.com/u/20815047/FTSERSIMAsystemdetail.png

I've looked through the trades and it seems to be a much smoother system with fewer false signals and fewer trades in general. As you can see, yesterday's buy signal was not phased by today's drop first thing and remained valid (as it still does) thereby keeping you in the trade. When I checked some other stats on this system I think (from memory) that it showed an annualised return of c. 337% which ain't too shabby if correct!

What is very clear, as has already been noted, is that few systems work during sideways moves. Apparently, the only type of system that does work then are what are termed "trend fading systems". There are a few of these in the updata system library but not one of them showed a profit during the period tested.

I'll continue to search systems and test them (there's also an optimisation facility) to look for something better but, for now, the RSI based MA crossover system is the best I've found. There was one other with similar results (called the Cliche System!) but that produced over 250 trades during the same 4 day period!

Thanks for your input. It's much appreciated

Jackozy - I've used this system and still do, however i have found this system will keep chopping you in and out of trades, buy, sell, buy ,sell etc...

Try using this in conjunction with the MACD. crossover on your MA's confirmed by Macd X and going up abv 0 to confirm buy signal and remains in a buy until: This i have found to maximise the trades.

a Sell, cross on your MA's followed by MacD X, confirmed sell by going -0

a good example of this morning @ 8:05 strong Buy signal & 9:00 a weak sell signal due no X on Macd and remaining abv 0.

Sorry for the delayed report on this. Due to tiredness yesterday morning I lost my discipline and broke the rules, started playing catch up and lost the plot completely so these results are not my trading results, just those from the chart and triggers:

10 trades, 2 winners for + 16pts, 8 losers for - 43pts.

Again, this was down to the largely sideways day. It did give me chance to run a few systems tests using updata and I'll report my findings on those later on.

OK, I've done a bit of a review and some testing on this system. I still think there's a lot of room for improvement, but I've only toyed around with mixing up some of the MA pairs and also using a combo of MAs and EMAs.

The previous results, based on trading every signal by going long/cover short and selling/going short at each one, were (results for first two days are amalgamated)

12th, 13th Feb: +47.5pts
14th Feb: +46.2pts
15th Feb: -21.2pts

This gives a total gain of +72.5pts for the period.

Now, using MA5 as the fast oscillator and MA 21 as the slow one and also using the worst case points of the triggering candles (eg if a signal to go long occurs, I've used the highest price of that candle to close the previous short and open the long), the results for the same period would have been:

12th, 13th Feb: +56pts
14th Feb: +31pts
15th Feb: -7pts

This would have given a gain of +80pts for the same period. Remember, that was assuming that the trades where opened and closed at the worst possible price on every occasion, unlike the live version above where few of the trades were at the worst price.

I will therefore be running with the 5 and 21 MAs from open tomorrow. This trial is also only being run during market hours.

I'm pretty sure that this can be improved upon, perhaps by using bull and bear divergences to ignore counter-trend signals. I'll let you know how it goes.