MNB 10.4 Streamlined by Snailbeard

[snailbeard]

I've been meaning to do this analysis as well. Haven't had time yet. I figured I'd have to dump or capture data somewhere to get css to play nicely in a back test. I have done lots of visual back testing, with the CSS bailouts and entries. Frankly I don't think the bailouts are any better than a psar, or shifted MA. But would love to see hard data. I think a lot of the added value is visual. Are we converging, diverging, parrallel, flat, etc.

I'd love to see some proper stats. Not to mention, is there a huge difference using a closed bar vs. the live bar? It doesn't seem to whipsaw much, but the price is paid for in lag. 15M and 30M are actually quite nice as well. Still very few false signals, and much more responsive for entries.

Nice work snailbeard. I'd love to be playing and coding with this more.

Mike

I agree with your supposition that other traditional indicators give similar signals, so perhaps a bit of a wild goose chase, but hopefully I can reuse some of the techniques and utility code in another project.

However, if anyone wants to experiment with it here is the code.
Note that it is currently limited to H4 and H1.

[snailbeard]

But then again, having another look at H1 CSS - provided the signals are qualified by other criteria, there are possibilities - interestingly, its like a SMA20 price cross detector but without noise.

I have updated the indicator with a few more CPU saver features and added an option - saveCPU which draws only the crossover and break threshold lines.

This light weight version should be very light on CPU compared to the full featured original version by Baluda - thanks Baluda - I would not have known where to start without it. My version is not a substitute, if you want to use other time frames other than H1 and H4 then you need Baluda's full featured version.

Also if anyone wants to use the logic in an EA it should be almost plug-and-play - although I haven't confirmed that yet!

[snailbeard]

EA modifications:

After reworking the indicator code slightly I have it working in the EA.

I've added an option to place direct trades when H1 CSS crosses-over in the same direction as the H4 MA240, and an option to disable pending orders, so I can look at just the CSS trades.
The exits use the same logic as before (not CSS) - one step at a time.

A short range back test produced interesting results, although some errors occur:

GBPUSD.G,H1: OrderModifyReliable v4.1: Attempted modify of #1 price:0.00000000 sl:1.54983202 tp:1.56969667

Some bug fixing and further experimentation and analysis is required to find if this is going to be useful.

[snailbeard]

After improving/bug fixing the entry logic so as not to clash with the pending order logic, the results are more consitent and there are no warnings/errors in the log file.

This short back test looks quite promising:

[snailbeard]

Can we use leading indicators to determine better entries and exits?

[snailbeard]

These results represent near optimum operation, so they can be refered back to for checking future 'improvements':

Code: Select all

1	2013.05.02 09:00	buy	1	0.20	1.55719	0.00000	0.00000	0.00	10000.00
2	2013.05.02 09:00	modify	1	0.20	1.55719	1.54719	1.56970	0.00	10000.00
3	2013.05.02 10:00	close	1	0.20	1.55720	1.54719	1.56970	-1.20	9998.80
4	2013.05.03 07:00	buy	2	0.19	1.55449	0.00000	0.00000	0.00	9998.80
5	2013.05.03 07:00	modify	2	0.19	1.55449	1.54449	1.56970	0.00	9998.80
6	2013.05.03 08:00	close	2	0.19	1.55323	1.54449	1.56970	-25.27	9973.53
7	2013.05.07 01:00	buy	3	0.19	1.55480	0.00000	0.00000	0.00	9973.53
8	2013.05.07 01:00	modify	3	0.19	1.55480	1.54480	1.56883	0.00	9973.53
9	2013.05.07 02:00	close	3	0.19	1.55453	1.54480	1.56883	-6.46	9967.07
10	2013.05.08 09:00	buy	4	0.19	1.54858	0.00000	0.00000	0.00	9967.07
11	2013.05.08 09:00	modify	4	0.19	1.54858	1.53858	1.56883	0.00	9967.07
12	2013.05.08 10:00	close	4	0.19	1.54932	1.53858	1.56883	12.73	9979.80
13	2013.05.09 07:00	buy	5	0.19	1.55445	0.00000	0.00000	0.00	9979.80
14	2013.05.09 07:00	modify	5	0.19	1.55445	1.54445	1.56883	0.00	9979.80
15	2013.05.09 08:00	close	5	0.19	1.55531	1.54445	1.56883	15.01	9994.81
16	2013.06.04 05:00	buy	6	0.19	1.53202	0.00000	0.00000	0.00	9994.81
17	2013.06.04 05:00	modify	6	0.19	1.53202	1.52202	1.53837	0.00	9994.81
18	2013.06.05 03:00	buy	7	0.19	1.53128	0.00000	0.00000	0.00	9994.81
19	2013.06.05 03:00	modify	7	0.19	1.53128	1.52128	1.53837	0.00	9994.81
20	2013.06.05 13:00	t/p	6	0.19	1.53837	1.52202	1.53837	119.40	10114.21
21	2013.06.05 13:00	t/p	7	0.19	1.53837	1.52128	1.53837	133.38	10247.59
22	2013.06.10 01:00	buy	8	0.20	1.55548	0.00000	0.00000	0.00	10247.59
23	2013.06.10 01:00	modify	8	0.20	1.55548	1.54548	1.59561	0.00	10247.59
24	2013.06.10 11:00	buy	9	0.20	1.55401	0.00000	0.00000	0.00	10247.59
25	2013.06.10 11:00	modify	9	0.20	1.55401	1.54401	1.59561	0.00	10247.59
26	2013.06.11 16:00	close	9	0.20	1.55875	1.54401	1.59561	93.48	10341.07
27	2013.06.11 16:00	close	8	0.20	1.55875	1.54548	1.59561	64.08	10405.15
28	2013.06.12 13:00	buy	10	0.20	1.56660	0.00000	0.00000	0.00	10405.15
29	2013.06.12 13:00	modify	10	0.20	1.56660	1.55660	1.59561	0.00	10405.15
30	2013.06.13 14:00	buy	11	0.20	1.56778	0.00000	0.00000	0.00	10405.15
31	2013.06.13 14:00	modify	11	0.20	1.56778	1.55778	1.59561	0.00	10405.15
32	2013.06.17 04:00	close	11	0.20	1.57008	1.55778	1.59561	44.76	10449.91
33	2013.06.17 04:00	close	10	0.20	1.57008	1.55660	1.59561	68.60	10518.51
34	2013.06.18 18:00	buy	12	0.21	1.56533	0.00000	0.00000	0.00	10518.51
35	2013.06.18 18:00	modify	12	0.21	1.56533	1.55533	1.58973	0.00	10518.51
36	2013.06.18 19:00	close	12	0.21	1.56499	1.55533	1.58973	-8.61	10509.90

[snailbeard]

Although, I am not currently working on counter-trend trades, I noticed an interesting set-up which looks like it could be straight forward to code, although the trend logic might throw a wobbler!

If we were playing poker we might describe this set-up as a full house...

[sponn]

Great work mate. Really. When I have some time I will try to help in BT (I have no idea about coding )

[snailbeard]

Following simple crossover logic in the following chart we see three entries. However, visually, we notice exhaustion candles before the third entry. In addition, some of the those candles appear to be bumping against an invisible line, i.e. there is a channel boundary lurking in the vicinity of the current price.

The third entry if taken results in a retrace to the stop loss.

[theforexedge]

Brian

Thanks for your insightful analysis...it's really helpful for me also with my live trading, and my attempt at trying to learn MQL albeit rather slowly. I have Streamlined2 on test and it's averaging around 0.5 - 0.87% per day, but last week due to my VPS it didn't trade. Is v2.0 the latest incarnation?

One indicator I'm playing with right now is Baluda's Global Market Trend, so not sure if it's worth some further analysis for trading..(see attached)

Jason