Boulder's Grind

[tomele]

Hi.

Just started test on 28 pairs 24/5:
Timeframe = M5
Wilder's DMI = M5
TradeLevel 1 = 2.4
TradeLevel 2 = 2.8
TradeLevel 3 = 3.2
BasisTakeProfitPips = 10
TakeProfitIncrementPips = 2
TradeCandleLife = 3
StopLossPips = 30
Account = Demo 10k
Risk= .01/1000

[tomele]

Hi,

does TradeCandleLife work as expected on your side?

[Boulder]

Heya Tomele,

It seems indeed that the trades aren't closed according to their preconfigured TradeCandleLife. Take this one for example (instance at the right side of the screenshot),

BG.JPG

Other than that the EA is doing a great job at the moment

Just saw it in live action. BG set 6 capped over 1 percent just now. If anyone should like to play around and have a better grasp of the power of the compound effect, have a look here. If BG can yield 1-2 % per day on average, I'd be a very happy man Early days though. Let's see what the nearby future brings.

Results.JPG

Also another thing I noticed is this,

BG 3.JPG

We enter a short there exactly as it is supposed to, then the market suddenly reverses on us. This unfortunately does happen. Now a neat property of this method is that it will capture that counter momentum too by trading the other way. By doing so recovering some of that failed short. In this case the buy didn't trigger at the open of that doji candle though. Is this because the EA doesn't allow more than 1 active open trade at a time Steve?

In essence this EA is based on momentum. By defining Diff as a proxy for momentum, we can quickly hop in and snatch away a few pips. By making distinctions in the perceived levels of momentum, i.e. the TradeLevels, we can enhance profit by aiming for a slightly higher amount of pips.

[tomele]

Hi Boulder,

Got mixed results until now.

AUDCADM5.png

There is something I dont understand with the stoplosses. I have set them to 30, but they are placed at positions like 80, 90, 150, 180. Ist there some sort of atr calculation involved or what did I wrong?

Thomas

[Boulder]

Hi Boulder,

Got mixed results until now.



There is something I dont understand with the stoplosses. I have set them to 30, but they are placed at positions like 80, 90, 150, 180. Ist there some sort of atr calculation involved or what did I wrong?

Thomas

Hi Tomele,

I'm not really sure what happened there with regards to those stop losses. Mine are always perfectly placed. For all 6 sets I decided to put 40 pip stop-losses instead of 35 pip ones. Also I'd recommend using BG on relatively volatile, low-spread pairs. The configurations of the EA will undeniably be pair-specifc. Actually, come to think of it, might be worth looking at gold, silver and perhaps the exotic pairs to find feasible configurations. I'll keep you posted!

[tomele]

Hi Boulder,

after closing all positions and restarting Empty4 the stops are placed correctly now. I have no idea what caused this.

Thomas

[Boulder]

Hey guys, quick update.

Results.JPG

BG doing its thing steadily at the moment. BG set 4 on top and bagging 1.4% throughout the day. The percentage would've been higher if it closed out properly within 15 mins (avoiding those nasty plummets). The configurations w.r.t. the TradeLevels seem to hold up too. Perhaps lowering the lower-bound to 2.1 - 2.2 may be more optimal as we'd engage in more trades (those of you who'd like to test that out, be my guest , I've already got a lot of things running and don't have more capacity to spare).

I'm really beginning to get a thorough feel of the DMI (M5) and the extent of its power. Went to dig a little deeper on the matter I described in a previous post. What I found is that the same pattern occurs over and over again. Have a look at GBPJPY today:

Pattern.JPG

The pattern consists as I reported of 3 DMI histogram bars falling for shorts and rising for longs. Since price has wave-like features, we could ride the little waves as it were and grab tiny amounts of pips there too (say 5 - 6 pips). A major point to by-pass is when exactly something could be considered momentum or a peak in price with a subsequent reversal - being the tops of a sinusoidal wave one can imagine. The answer lies yet again in the Diff. Relatively high values indicate 'true' momentum with relatively high probabilities of success. I use 2.4 here as a default threshold. Medium values just under that threshold have a higher chance of reversal. Try this for yourself on a chart if you're interested. You may even find more suitable and optimal values. Now the exercise I did was to indicate with a rectangle 3 rising or falling bars with sufficient yet not too large Diff's between the first and the 2nd pair of the histogram bars. Especially not too large ones on the 2nd half. Visually it would be the equivalent of capturing consistent and congruent stair-shapes. I included a 4th bar to complete the picture. Then went looking at the actual price and whether we would be able to snatch at least 5 pips. To my amazement, we actually do. Even the ones where we don't have 4 steps downwards/upwards; still would've gotten 5 pips on a spike regularly. Now of course the pattern isn't foolproof. But it works out quite a number of times.. A strict rule of no more than 2 secondary trades in the same direction seems imperative as well in order to avoid getting hit by reversals. To be even more prudent and allow just 1 secondary trade, until e.g. 20 minutes have passed, could be more optimal.

The process for these secondary trades would go something like this. I define Diff_1 and Diff_2 as the Diffs between respectively the first pair and second pair pertaining to a set of 3 bars.

Go long if:
1. 3 rising histograms
2. Both Diff_1 and Diff_2 between 0.4 and 1.8.
3. Open trade if maximum 1 trade in the same direction has been triggered in the last 10-15 minutes.
4. TP = e.g. 5 pips
5. SL = 35 to avoid disaster
6. If 4. and 5. didn't happen after 15 minutes; exit the trade.

That's it. Analogously for a short. By the way, for those that are testing BG, I'd recommend to turn off the EA for a few hours at times of high spreads e.g. around roll-over time. I'll leave it on for now just to see how it does over the course of the night.

[Boulder]

How amusing.. as I was writing that last post, the pattern actually occured again.

Pattern 2.JPG

[Boulder]

Hi Boulder,

after closing all positions and restarting Empty4 the stops are placed correctly now. I have no idea what caused this.

Thomas

Thomas I just experienced the same thing with the TP/SL randomly changing, as you can witness here:

Screen Shot 2016-06-07 at 02.38.37.png

Steve any idea as to what this might be due? I've restarted my terminal and TP/SL seem fine now. On a positive note though, the trades at the right would have capped their TP

[SteveHopwood]

Thomas I just experienced the same thing with the TP/SL randomly changing, as you can witness here:


Steve any idea as to what this might be due?

Nope.