M15 BB and MTF MA

[snailbeard]

Just an idea at the moment and I have not really looked into it...

The MTF MA flags should flip in sequence lowest TF to highest TF
There should be a couple of wickies or railroad tracks poking out from the Bollinger before MA signals

Primary entry method is if the price comes back out of the edge of the Bollinger and triggers an order when the price reenters the BB. This could happen several times.
Secondary entry method is when the price comes back to mid-Bollinger and reverses.

More details in the screenshot

Since I normally trade at a snail's pace and this requires checking every tick - (although it might work with limit orders), this would be a much accelerated version of H1 wake up and see what's changed approach, which is why I am referring to it as my hot-ice strategy.

Entries and exits could be linked/restricted to significant levels

After some more analysis I might code it up in real programming language...
Except I have the Empty4 BB code, the MTF MA code, the pivot code,
the order entry code, the basket code, the...
Doh!

I would interested to know if any veterans have tried this approach before?

[mjws00]

I did something very similar using TMA Bands vs Bolli. Admittedly it was quick and dirty, only bothered with a slope filter, a time delay between orders, and a max number of orders. I ran it as a stop and reverse ea as I couldn't be bothered to right much exit code. Hard SL. Tested Hard TP as well. Didn't do the MTF MA, but have studied them forever. MTF Heikan Ashi has some neat properties as well.

It did pretty well. But I was only testing the logic, not something to trade real money with. Primary weakness was during a long trend where price can hug the bands for a long time. The run after E would be an example. Over the course of a few years data you'll find a few market conditions where it will need intervention.

Enjoyed your work on 10.4 EA.

Mike

[snailbeard]

Can we get better entries by tracking the swing history from the H4 timeframe?

I am experimenting with functions to calculate pivots for recent weeks;
get the W1 high/low
compare them to the pivots
store the pivot level index which corresponds to
a) high point of the week and the low point of the week.
b) low point of the week
If we calculate a swing from the last low or high to a potential target base on the previous swing, we might get better trade selection.

So far the backtest trace output looks like this:

Current week:
: back: 0 : lvTargetPivIdx : 5, low pv: 1.55715, CandleLow: 1.55465, Gap : 0.00250
: back: 0 : hvTargetPivIdx : 6, high pv: 1.56250, CandleHigh: 1.56840, Gap : 0.00590
: W1 High Target Piv Idx: 6 : PIVOT
: W1 Low Target Piv Idx: 5 : MidS1

Last Week:
: back: 1 : lvTargetPivIdx : 4, low pv: 1.55331, CandleLow: 1.54904, Gap : 0.00427
: back: 1 : hvTargetPivIdx : 6, high pv: 1.56505, CandleHigh: 1.57320, Gap : 0.00815
: W1 High Target Piv Idx: 6 : PIVOT
: W1 Low Target Piv Idx: 4 : S1

Previous Week:
: back: 2 : lvTargetPivIdx : 4, low pv: 1.55046, CandleLow: 1.54582, Gap : 0.00464
: back: 2 : hvTargetPivIdx : 8, high pv: 1.57254, CandleHigh: 1.57680, Gap : 0.00426
: W1 High Target Piv Idx: 8 : R1
: W1 Low Target Piv Idx: 4 : S1

Gap tells us about the pivot overshoot: e.g. leave part of the trade going if the overshoot is significant.

So a prefered entry would have the following:
a) outside the H4 Bollinger - or recent reentry after (ZZ)
b) beyond the same previous pivot level where price previously reached ( eg. below S1, then crossing S1 up if entering buy )
c) targets and partial closure between entry and each pivots level between here and potential target where the potential target is last weeks ZZ pivot level.