The orignal concept used 1 Hour bars using different SMA but for ease of use and increased profit ratio some of us looked at the 4 hour model to which i will describe here. The benift of 4 hours is you get a good nights sleep, can go out during the day for many hours and so basically are not tied to the PC all day. 4h also provides a balanced blend between the short term intraday 15m 20m 1h time and longer daily and weekly cycles so we are still capturing a good portion of the market swings.
Its uses a 55 SMA as its core. This is not a random picked number its a fibo number sequence and has been found to follow on almost any time frame the general movement of the market.
Overlayed on the same chart are fibo number sequence lines where we take profit. In case you didnt know for newbies reading this the sequence is 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, and higher.
However it has been noted that in the case of euro it will only deviate from the 55 SMA out 233 pips away before it has to retrace. There are some rare spillage exceptions but this is a rule of thumb. Other pairs like cable and some volatile crosses will go to 377 pips away from the 55 SMA and retrace. So the basis of the system is we buy as soon a we are above the 55 SMA and Sell below it.
Now this is the clever part that makes nearly ALL moving average systems fail. Fibo numbers are well known retracement points and they are also excellent supp resistance areas so by taking a partial profit at 89, 144 and 233 use on euro we capture the main places a price is most likley to reach before it returns back to the 55 SMA line.
For pound and others the 3 fibo numbers used are 144, 233 and 377. OK so you might ask well how do i know what set to use on each pair? Easy plot these lines based on the 55 SMA then SHIFT and draw these fibo lines 144 pips 233 pips and 233 pips higher and lower off the SMA so the chart looks like wavy rail tracks both above and below the 55 SMA then look to see for a particular pair it will fit like a glove 98% of the time to the furthest price you can see on the charts over many months.
The age old problem is should we use filters to make sure the price goes over the 55 SMA to take a long or visa versa for a short. Vegas suggest that a 12 EMA is overlayed which needs to cross over the 55 SMA to take a long posi. I found 13 (fibo) JMA Jurik much better then EMA. If you dont know what Jurik JMA is then google it. I seen a thousand other filters can be used BUT anything you use will result is a delay and can miss major moves. It matters not if you CCI, EMA, mommetum RSI etc etc it will result in some moves being completly lost.
Just recently on MT they have a Moving Average Expert to play with as standard on new MT 4. That system just appears to wait for a complete bar close above or below the 55 sma and thats your signal. I was amazed how well this works. it needs no stops, no filters nothing. If it closes above we go long, a close below we go short.
Now we left the tunnel going long with say 4 lots. A the 1st fibo i take 1 lot profit at 89 fib, another lot at 144 fib and another lot at 233 fib. Because the price wont go futher then this 98% of the time we have locked in the entire range but the last lot is left open. This is becuase if the last lot is left open if we return to the 55 SMA its so small it wont really harm our profit but there is a chance the 55 SMA can begin to slope up at a steep angle and the price and the 55 SMA will run for hundreds of pips. So the last lot left open begins a free trade. This is the opposite of most conceptions where you think you should be adding to trades but we are reducing our risk by taking profits in fibo stages. In practice this is MUCH safer.
OK you hopefully have the basis of this system. Its a swing trader with partial profits but instead of saying TP at "normal" 30 50 or 100 pips we are using something of real mathmatiical substance of where the market is likely to stall at ie fibo points.
In practice a risk map needs to be defined the 55 SMA does get whipped. Most of the danger is in the crossing becuase we are fully loaded. An expert script on mt4 is really needed to find the initial risk value of the system eg i have 100k to start risk %= 30% will this blow the account??
Next most important i described taking 1 lot at each fibo but ideally this should be a % of the initial order eg 30% This is so we can scale in any size of account and compound growth.
So 100k cap start with 30% = 30k or 30 lots
fibo % is set to 30% TP so
30 lots - 30% = 9 lots TP on 1st fibo = 21 left
second fibo hit now 21 - 30% = 6.3 so round this to 6 leaves 15 lots
Third fibo hit 15 lots left - 30% = 4.5 round up to 5 = 10 lots.
These last lots are left runnning incase we trend hard but its likely also to return back to the 55 SMA but we havent gone back empty handed as profits have been taken yet we have left some open in case we catch a nice trend which happens about 10 times a year.
What i need to know is the optimal % of the TP order value that should be used. If any one can script this on MT4 the backtester has an automatic optimiser to find the max profit of any user varaible. My manual testing has shown that 30% may be fine and certainly a starting point.
Also one can see the greatist portion of the order is taken at the 1st fib. This is because the market ranges more at the 89 fib then it does trending so we lock in the greatest protion of the profit early rather then later. Incidently profit is only taken at each fib level once. If the market ranges between the 1st and 3rd fib level we only take profit only once the first time it was hit. Only a break across the 55 SMA clean on bar close reverses the trades and opens a fresh full lots order based on the initial capital % risk value.
by Bolt1
In practice a risk map needs to be defined the 55 SMA does get whipped. Most of the danger is in the crossing becuase we are fully loaded. An expert script on mt4 is really needed to find the initial risk value of the system eg i have 100k to start risk %= 30% will this blow the account??
Next most important i described taking 1 lot at each fibo but ideally this should be a % of the initial order eg 30% This is so we can scale in any size of account and compound growth.
So 100k cap start with 30% = 30k or 30 lots
fibo % is set to 30% TP so
30 lots - 30% = 9 lots TP on 1st fibo = 21 left
second fibo hit now 21 - 30% = 6.3 so round this to 6 leaves 15 lots
Third fibo hit 15 lots left - 30% = 4.5 round up to 5 = 10 lots.
These last lots are left runnning incase we trend hard but its likely also to return back to the 55 SMA but we havent gone back empty handed as profits have been taken yet we have left some open in case we catch a nice trend which happens about 10 times a year.
What i need to know is the optimal % of the TP order value that should be used. If any one can script this on MT4 the backtester has an automatic optimiser to find the max profit of any user varaible. My manual testing has shown that 30% may be fine and certainly a starting point.
Also one can see the greatist portion of the order is taken at the 1st fib. This is because the market ranges more at the 89 fib then it does trending so we lock in the greatest protion of the profit early rather then later. Incidently profit is only taken at each fib level once. If the market ranges between the 1st and 3rd fib level we only take profit only once the first time it was hit. Only a break across the 55 SMA clean on bar close reverses the trades and opens a fresh full lots order based on the initial capital % risk value.
No comments:
Post a Comment