only using SMA 21

Dear All

i always to try find the simplest system to play FOREX. i believe Holy Grail is our MONEYMANAGEMENT and DISCIPLINE.. so i found an idea, and i think it will be good with improvement and advise for all of you

okay here we go

1st. open GBPUSD 1H
2nd put SMA 21 on the chart

so the rule is... Buy when new candle open above 21 SMA
Sell when new candle open below 21 SMA

how about ranging market...? i suggest not using stop lost but we can have good idea from martiangle system... offcourse with good moneymanagement


example.. when we got BUY signal we entry with BUY 0,1 lot, but soon in next 3 candle it open below SMA 21, so we give SELL 0,15 lot.. and when it back again we can put BUY 0,2 lot...

in my observation, the maximum price ranging at sma 21 is 5 times

so we got 0,15 sell and 0,3 buy... we can put our target as $$$.. so if the target is reached, we close all, and start again

OR.. we can use higher time frame 4H as filter

so come on guys, i think it will be great.... especially when we have EA for this system..

thx

hei i found your idea is great

at TSD they talk about it and try to make EA based on it..

check this

http://www.forex-tsd.com/expert-advi...ale-sma21.html

i hope the EA will be finished soon..

IMHO.. this strategy is good for gbpusd, eurojpy

Great Job Bill!

I hope that this is very profitible for you and thanks for sharing!

Whether this will work or not I do not know. What I don't understand is how a method such as this is considered Martingale. I feel it is a wrong label but I hear it all the time.

With Martingale, the first bet is a dead horse. You have given your money to the house. As stated above, If you place another trade without touching the first one, whether in the same direction or not, this is not Martingale because your first trade is still in the game on a reversal. This is hardley Martingale.

i did backtest manually at january.. i think this system will be great at 4H TF.

my target profit is equal 200 pips at mini account


orange circle for SELL

blue circle for BUY

I backtested GBP/USD@H1 for January from 02.01 to 26.01, taking all the signals given without any other filter, and considering a 50 pips stoploss, in two versions: watching and closing manually at next SMA cross (or discretionary), and with 50 pips trailing stops.

27 trades, 18 profitable (66,67%) 9 stopped at S/L or negative.

Total pips 998 w/manual closing, 763 with trailings.

Trades criteria:
- GBP/USD H1 chart
- Long when price opens above SMA 21, close when price opens below SMA 21
- Short when price opens below SMA 21, close when price opens above SMA 21
- Ignore when price opens exactly at SMA 21 (no position open or close)
- Stoploss is at -50, no trailing stop
- Bid/ask spread is 4 pips (quite general for retail brokers)
- 24 hrs a day chart watch, but only around every hour change (hence EA is needed, we're not superman)

Backtest trades taken since the beginning of Jan 2007 till now:
(Columns: < Position >< Open price >< Close price >< Profit/loss at close >)

Short 9722 9430 +292
Long 9430 9406 -24
Short 9406 9331 +75
Long 9331 9286 -45
Short 9286 9336 -50
Long 9342 9402 +60
Short 9402 9396 +6
Long 9396 9377 -19
Short 9377 9388 -11
Long 9388 9377 -11
Short 9377 9349 +28
Long 9349 9626 +277
Short 9626 9649 -23
Long 9649 9612 -37
Short 9612 9635 -23
Long 9635 9683 +48
Short 9683 9733 -50
Long 9722 9709 -13
Short 9709 9722 -13
Long 9722 9718 -4
Short 9718 9747 -29
Long 9747 9734 -13
Short 9734 9761 -27
Long 9761 9741 -20
Short 9741 9746 -5
Long 9746 9720 -26
Short 9720 9755 -35
Long 9755 9741 -14
Short 9741 9764 -23
Long 9764 9823 +59
Short 9823 9695 +128
Long 9695 9682 -13
Short 9682 Ongoing, SMA 21 @1.9619 at close

Total trades: 32, Winners: 9, Losers: 22, Ongoing: 1
Win ratio = 29.0%
Total profits in winners = +973
Total losses in losers = -528
Net profit = +445

Note that you'll always close previous trade and open next trade at same price, except when -50 stoploss is hit.

PS: If every trade's risk is 5% (-50 pips) initial capital, +445 equals +44.5% capital

I have a small but very good filter for this system in order to avoid opening wrong position.

Enter long just when price passes 2 pips above the shadow of first candle closed obove SMA21

Enter short just when price passes 2 pips below the shadow of first candle closed below SMA21

This suggestion is based on fractal geometry

Nice filter, on my backtesting it filters out 50% (4 out of 8) of the losing or not so much performing trades listed (I couldn't edit my post, the ratio is 19 to 8 so should read 70,37% pp. instead of 66,67%).

Thanks so much for the tip

(Columns: < Position >< Open price >< Close price >< Profit/loss at close >)

Short 9719 9335 +384
Long 9335 9285 -50
Short 9283 9333 -50
Long 9350 9389 +39
Short 9389 9413 -24
Long 9413 9363 -50
Short 9375 9353 +22
Long 9353 9609 +256
Short 9609 9642 -33
Long 9642 9675 +33
Short 9675 9725 -50
Long 9727 9709 -18
Short 9709 9759 -50
Long 9762 9736 -26
Short 9736 9751 -15
Long 9751 9737 -14
Short 9737 9767 -30
Long 9767 9819 +52
Short 9819 9703 +116
Long 9703 9675 -28
Short 9675 Ongoing, SMA 21 @1.9619 at close

Total trades: 21, Winners: 7, Losers: 13, Ongoing: 1
Win ratio = 35%
Total profits in winners = +902
Total losses in losers = -438
Net profit = +464

Including Fun Trader's filter, and 1 more criteria: only 1 position opened at anytime. It's better than last one. Thanks to Fun Trader.

I believe January was a decent month for this system, but try it on November 06 or May 06 and it will have problems. Did anyone use the martingale aspect at all in their backtests?

Backtesting for January 2007 and then for all of 2006

Rules:
enteries as described in post on 1 and 8. As that was the best results.
exits as described in post 1

This has potential and with some good Money management and optimization, could be good.. It is still the consolidation areas that do it damage...

I think it would trade better in real market action, because the tick generated data is live, not simulated...

I quickly tested this over time 1 year.

14,000 in profit not including spread with 420something trades made it a losing system. However, a couple of things I did to it made it profitable.

1 enter buy on close above average of high as opposed to avg of close.
2 enter sell on close below average of low as opposed to avg of close.

3 setstoploss to 22 period averagetrurange * 2

over the past52 weeks this has brought this system to profitable.

if someone can get official figures i am very curious.

My results although not great look somewhat promising. Over 5 years produced about 20K total but with max drawdown of 20K as well - it would seem as if a ghost trade type variation would work well with this. Perhaps ghost trading breakouts on close during certain hours. Danger however also lies in entering at open of the hour as it may coincide with news events that wont give efficient entries, so for this, it may be of use to start the hour on an increment other than the round minute type.
Finally if the system performs within the realms of normal statistics, we can plot its performance with standard deviation to determine optimal times for ghost trading and optimal times to go strong. before doing this however, the equity curve will need to be smoothen out.

Here it is with no Trailing stop at all just a 50 pips stop

Edit, Here is a copy of my EA that I used to test it.

try at 4H.. no problem at all

ON a 4h it works nicely. for the last year. A little heavy on Draw Down, but profitable none the less.

Could you explain with example of your filter?

Is this question only directed to Fun Trader, or may I answer with my interpitation?

anybody would do..no harm i guess...

Here is my interpretation of the filter.

I am not endorsing this idea, only, this is how I interpreted it when I put together the EA for testing.