ash.invest wrote:sualou - sounds like you are using 5pips buffer from the S/R Level. Some of the trades I have done have been better with 10pip buffer. Yes I have given up some pips as PA would come as close as 1 pip to SR.
What has been your experience ?
Hi Ash
yes, 5p buffer atm, giving 10p SR "zones" range.
I set my SL to the SR zone range, so 10p.
Rather small, i agree, which means smaller losses, but hit more often.
But because the "next SR level" is usually 50-90p away, i can afford several SL hits when PA decides to hang around...
And the R:R remains quite high, >= 5:1 (a little less due to spread).
The win/loss ratio should of course remain over 1/5 to be profitable
Wider zones will give potentially less loosers, but also less R:R.
It is not easy to decide what buffer to apply, this will depend on
- the pairs due to their specific volatility patterns: use ATR(14,H1) to evaluate the volatility;
- time of the day : again ATR(14,H1) will give you the typical volatility pattern in terms of extension AND time distribution;
- a major factor is the "strength" of the S/R level itself: strong S/R levels will most likely act as bareers, with PA BOUNCING off these strong levels very accurately, sometimes even to-the-pip, rather than continue in the same direction.
These strong levels can be traded with a very small buffer, typically 5p max.
The thing is, because these levels act as bareers, it means PA will NOT care about the general trend when it approaches them.
Therefore Bob's 10.1 logic will not work here unless PA is approaching the SR in a counter-trend move.
In that case the SR will act as a trend resuming trigger, giving a low risk / high reward trade.
It is therefore vital to be able to estimate the "strength" of an SR level.
Not an easy task...
On low volatility periods (for EU : typically from NY-close to Frankfurt-open), PA will very often "go to sleep" and very likely BOUNCE between its last swing around NY close, and some 20-35p away from that, and you can expect a few LOW-RISK BOUNCE TRADES using a simple Stochastic set to 2 hours (stoch(24) on M5).
I am going a bit off topic here... although the "last swing" that gives the anchor for the "night camp" becomes a very strong SR level, giving a high-probability / low-risk bounce trade opportunity, and PA will very likely bounce MULTIPLE times off that level area.
This happens on "quiet" nights, where nothing special happens during the Asian session.
An ATR(14,H1) gives a good view of the volatility pattern of a pair.
As ATR is calculated using an SMA, it has a lag of 14/2 = 7 bars.
You can show a quite time-accurate past volatility line by drawing the ATR(14) on an H1 chart, and apply to that ATR line an SMA(period=1, shift= -7): the SMA-of-ATR line will be synchronized with the PA itself.
You can then see very clearly the periods of low volatility.
I understand all this is quite evasive and is not helping the EA at this very moment... 
But this gives several research paths to improve the strategy and performance of the EA, eventually even leading to separate derived sub-systems...
Food for thought.
SQ |
