Most Famous Blessing 3.9.6 EA and Setfiles | Page 166 | Forex Factory

Options

Go to
Permalinks
Print Thread

Related Instruments

EUR/USD, GBP/USD, Gold/USD

Similar Threads

Can anyone convert the Blessing 3 v3.9.6.09 into MT5? 10 replies

Blessing Multipair EA 453 replies

Blessing EA by ananthhh 13 replies

'Blessing' EA Modification Wanted 3 replies

Forward testing results of my ea modified from Blessing 3.9.6 18 replies

Attachments: Most Famous Blessing 3.9.6 EA and Setfiles
Exit Attachments

Most Famous Blessing 3.9.6 EA and Setfiles

  • #3,301
  • Oct 19, 2022 6:57pm Oct 19, 2022 6:57pm
Quoting Akohamio
Good afternoon together! One question: I will update MT5 EA 3.9.6.23 to MT5 EA 3.9.6.26, but my settings hasn't a good perfomance with the new version. With the older version performance is good in backtest, but not with the newest version. What can i do?
Re-optimize.

I'm having very good results lately with M1, XAGUSD and just Bollinger bands and Stochastics. M1 seems to work best, though a little slower to optimize. Since most of the indicator timeframes are adjustable there's no need to set the overall timeframe higher. I'm using M1 for everything now. I let Strategy Tester find the best combinations.

I try hard not to make incompatible changes. But I'm making fixes and additions that work. And I expect to optimize somewhat frequently.

I would be surprised if you couldn't make the new versions work at least as well as the older ones with just a little tuneup.

If you find a bug please report it.

Good luck!
  • #3,302
  • Oct 20, 2022 1:35am Oct 20, 2022 1:35am
But how it is possible? I use fixed parameters with MACD, Stochastic and MA. Why they perform differnt with both version in the same settings?
  • #3,303
  • Oct 21, 2022 9:12am Oct 21, 2022 9:12am
Having a simular problem.
2 computers having metatrader 5 installed, same version, same timeframe, same staring and stopping dates, same version of Blessing, same setfile and still very different results.
I even get different resaults if i run a visual backtest and a normal backtest on the same computer.
Guess running the setfile on a demo account is the only safest way to determent if its OK....
  • #3,304
  • Oct 22, 2022 7:05pm Oct 22, 2022 7:05pm
Quoting Akohamio
But how it is possible? I use fixed parameters with MACD, Stochastic and MA. Why they perform differnt with both version in the same settings?
Saying they are DIFFERENT means nothing. Is it a fly or an elephant?

Your comments aren't helpful. Provide more data, please.

But first, examine everything and make sure you don't have other problems/issues at play. Restart, reboot, etc.

Thank you.
  • #3,305
  • Oct 24, 2022 4:34am Oct 24, 2022 4:34am
Quoting richard96816
Ideas from the thinkers out there would be most appreciated. What is Blessing missing or most in need of being fixed, changed or embellished? How can we increase her profitability most effectively for today's market? All ideas welcome. Thank you!
I have only recently returned to fx trading after pursuing other business projects over the last 12 years. Back in 2010-2011 I had reviewed 1600+ EAs and found that none of them was able to address the regime change problem (the switch from ranging to trending) adequately. J Talon's products had not been on my radar back then.

Four weeks ago I took Blessing3 out of storage where it had been sitting for the last 4 years, and I started to look at the code. To explore, I added two more entry modes, StdDev and RSI, for a total of 16 permutations (regular=1, and reverse=2).
I ran 12 fx pairs through back tests for all these 16 permutations (single entry mode) over the last two weeks, on a dual-processor laptop and on MT 4, more or less non-stop, using standard settings for the entry modes. In fact, I had to run StdDev and RSI twice, since I had made a coding error, with (entryMode + FMC) permutations set as [1+0, 2+1, 1+0, 2+1] whereas the correct permutations are [1+1, 2+0, 1+0, 2+1].
Incidentally, the first (wrong) set of permutations produced almost always the better results, both in terms of maxDD and profit. But that is another thread that can be investigated separately.
I also ran Blessing3 (MT4 version) on a demo account for 2 weeks (6 Oct to 21 Oct).

Attached is my analysis tool, and it gives an interesting result. I am not sure whether this is already on everyone's radar, since this is not being discussed or I missed it by not reading all forum contributions yet, please correct me if I am wrong, or share your thoughts. See Excel spreadsheet, columns AA-AU, vs columns EF-EZ.
AA-AU gives the contributors in USD.
EF-EZ gives the contributors in USD/lot.

It seems to tell me that some fx pairs only appear to be constructive contributors, whereas in fact they are negative capital utilisers, for instance USDCAD or USDCHF or XAGUSD.
My test period may not be representative, as I started just one day before NFP announcement 7 Oct, and another equity drawdown followed the next Friday, 14 Oct, and I ran it only for 2 weeks.
I assume that everyone here knows how to import the detailed statement into the Excel spreadsheet, again please correct me if I am wrong.
Any thoughts?
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Picture 1.png Size: 3 KB
Click to Enlarge Name: Picture 2.png Size: 4 KB
Attached File(s)
File Type: zip Blessing 4 test.mq4.zip   27 KB | 428 downloads
File Type: zip Blessing_3_v3.9.6.26v2.mq5.zip   63 KB | 679 downloads
File Type: zip 221022 EA live data.xlsx.zip   768 KB | 394 downloads
  • #3,306
  • Oct 24, 2022 7:13am Oct 24, 2022 7:13am
Quoting roa
{quote} I have only recently returned to fx trading after pursuing other business projects over the last 12 years. Back in 2010-2011 I had reviewed 1600+ EAs and found that none of them was able to address the regime change problem (the switch from ranging to trending) adequately. J Talon's products had not been on my radar back then. Four weeks ago I took Blessing3 out of storage where it had been sitting for the last 4 years, and I started to look at the code. To explore, I added two more entry modes, StdDev and RSI, for a total of 16 permutations...
Thanks for your contribution.

I've been thinking about adding RSI to the list of entries.

I'll take a closer look shortly. Thanks very much.
  • #3,307
  • Oct 24, 2022 7:32am Oct 24, 2022 7:32am
Quoting richard96816
{quote} Thanks for your contribution. I've been thinking about adding RSI to the list of entries. I'll take a closer look shortly. Thanks very much.
Thank you, much appreciated.
I forgot to include the account statement in htm+gif and pdf, here they are, in case someone wants to verify the Excel data entries.
Attached File(s)
File Type: pdf Statement_221021_1115_final.pdf   6.1 MB | 396 downloads
File Type: zip Statement_221021_1115_final.zip   47 KB | 268 downloads
  • #3,308
  • Oct 24, 2022 1:52pm Oct 24, 2022 1:52pm
Anyone seen this before?
Ended up with a lot of trades in a very short time using Blessing v3 3.9.6.26 for Meta trader 5
Is it Blessing or my broker (IC Markets)?
Lucky I could close them manually.

Attached Image (click to enlarge)
Click to Enlarge Name: blessing1.jpg Size: 535 KB
Attached Image (click to enlarge)
Click to Enlarge Name: blessing2.jpg Size: 293 KB
Attached Image (click to enlarge)
Click to Enlarge Name: blessing3.jpg Size: 45 KB
Attached File(s)
File Type: txt EURGBP 5 min utan indikator 1159 356 2210209.set.txt   13 KB | 252 downloads
  • #3,309
  • Edited 11:38pm Oct 24, 2022 9:55pm | Edited 11:38pm
Quoting richard96816
Ideas from the thinkers out there would be most appreciated. What is Blessing missing or most in need of being fixed, changed or embellished? How can we increase her profitability most effectively for today's market? All ideas welcome. Thank you!
My apologies if this idea is a little ambitious, as I don't think this can be incorporated into Blessing3 easily, but I assume that most Blessing3 users have a bunch of different fx pairs running concurrently.
I am not sure how many of you track the portfolio volatility of their open positions as a measure of their exposure to the market.
I explored the correlation data offered by mataf (see https/www.mataf.net) and tried to calculate it on 7 Oct (NFP announcement day) for the portfolio presented in my previous post.
For that I put together all permutations of my 12 fx pair portfolio, the rest is done automatically by the VLOOKUP function.
I generated the weight vector from the buy(positive sign) and sell(negative sign) positions.
If you find any errors do let me know. Pre- or post-multiplying the weight vector in the last step does make a difference.
I followed Copeland/Weston's most excellent book "Financial Theory and Corporate Policy" (3rd edition), page 174.
There are probably ways to automate this if the data import from mataf URL can be coded into Blessing3 and equipped with an on/off button, but I assume this is low on the priority list of users, please correct me if I am wrong. Any comments are welcome.
Attached File(s)
File Type: zip 221025 mataf correlation.xlsx.zip   114 KB | 269 downloads
  • #3,310
  • Oct 25, 2022 1:05am Oct 25, 2022 1:05am
i'm wondering y my ea never open any trade?
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 29 KB
  • #3,311
  • Oct 25, 2022 6:12am Oct 25, 2022 6:12am
Quoting thehut
i'm wondering y my ea never open any trade? {image}
For XAUUSD (Gold) increase MaxSpread to something like 25 and see if this works.
  • #3,312
  • Oct 25, 2022 5:21pm Oct 25, 2022 5:21pm
RSI and StdDev are working pretty well in my testing. They seem to dovetail well with the other indicators.

So far I'm optimizing for CCI and Bollinger Bands together, first. Then I optimize each of the other indicators on top of that, adding to the performance.

I'll try a few other orders, but BBands often seem to be the best starting point.

Now I definitely need to build a faster computer for optimizing. And possibly put the time in to automating optimization.

Good luck!
  • #3,313
  • Oct 25, 2022 7:19pm Oct 25, 2022 7:19pm
Quoting richard96816
RSI and StdDev are working pretty well in my testing. They seem to dovetail well with the other indicators. So far I'm optimizing for CCI and Bollinger Bands together, first. Then I optimize each of the other indicators on top of that, adding to the performance. I'll try a few other orders, but BBands often seem to be the best starting point. Now I definitely need to build a faster computer for optimizing. And possibly put the time in to automating optimization. Good luck!
I am glad to hear that. If they prove useful the credit must go entirely to Keith Fitschen who inspired me. His analysis of entry indicators in chapter 4 of his book show that RSI and Standard Deviation are good performers. In chapter 6 he suggests Standard Deviation to act as a low volatility or high volatility filter. I am still thinking how to make Standard Deviation a non-directional filter, as opposed to the directional indicator it is now. Just a thought. That is why I mentioned that the first (wrong) implementation performed better and I am still wondering whether this was a random outcome or whether there is a sound reason behind it.

Just this quote from Fitschen's book:
>>>
I always check performance with two forms of volatility filter. Your entry setup can occur in a sleepy, almost sideways market, or in a very volatile up and down market. In the first case, your profitability potential is reduced from a normal setup.You are taking the risk of entering the trade, but your profit potential isn’t that good unless the market heats up quickly, and in the direction of the trade. To filter out those lesser‐profit potential trades, the first filter only accepts trades that meet a certain threshold volatility; that’s the low‐volatility filter. When volatility is way above normal, you have an increased probability of getting stopped out due to higher up and down market noise. The high‐volatility filter bypasses trade entry when volatility is too high. There are a number of ways to measure volatility, like average range, average true range, or some measure of standard deviation. We’ll test both range and standard deviation.
<<<

Just on this occasion may I express my deep gratitude to you for all the outstanding work you have done and keep doing that has allowed Blessing to be what it is today. Thank you!
  • #3,314
  • Oct 25, 2022 9:28pm Oct 25, 2022 9:28pm
Quoting roa
{quote} For XAUUSD (Gold) increase MaxSpread to something like 25 and see if this works.
its work! thanks for help
  • #3,315
  • Edited 7:55am Oct 26, 2022 2:43am | Edited 7:55am
Quoting roa
{quote} I am glad to hear that. If they prove useful the credit must go entirely to Keith Fitschen who inspired me.
Just to follow up on this idea of a non-directional filter. It could look like this:
declaration
extern int LengthVol = 8; // number of candles used to calculate the volatility indicator values
extern int TF_Volat = 0; // timeframe Volatility: 1-М1, 2-М5, 3-М15, 4-M30, 5-H1, 6-H4, 7-D1, 8-W1, 9-1MONTH
extern double Level_Vol = 12; // volatility cutoff: M5: 5.0; M15: 5.0-12.0; H1: 12.0 (I have not explored yet how this can be made timeframe invariant.)
int mper[10]={0,1,5,15,30,60,240,1440,10080,43200};

double Vol = NormalizeDouble(iCustom(NULL,mper[TF_Volat],"Volatility",LengthVolat,0,1), 1); // using a standard volatility indicator

Buy and Sell order conditions:
if(Ask > [MA or CCI or BB condition] && Vol > Level_Vol) BuyOp=true;
if(Bid < [MA or CCI or BB condition] && Vol > Level_Vol) SellOp=true;

Here is a first iteration implementation. I have not tested it yet myself. The timeframe in the MQL5 version is fixed to "current".
In "1" mode it is supposed to act as Fitschen's low volatility filter.
In "2" mode it is supposed to act as Fitschen's high volatility filter.

As fas as I am concerned I don't see much added value in the 'Standard Deviation' indicator, it is yet another directional indicator of which Blessing already has several that work well. The overlap and similarity with BB is quite high, and Standard Deviation could be tossed out again to avoid bloating the code and to keep it as simple as possible and fast.
What Blessing is lacking are non-directional filters except for the day-of-the-week filter recently added by richard96816.
Feel free to test and evaluate to see whether this volatility filter works as intended. Just a suggestion. Any comments are welcome.
Attached File(s)
File Type: zip Blessing 4 vol.mq4.zip   27 KB | 428 downloads
File Type: zip Blessing_3_v3.9.6.26vol.mq5.zip   63 KB | 502 downloads
  • #3,316
  • Oct 26, 2022 4:53pm Oct 26, 2022 4:53pm
Thanks for the kind words, roa. There have been many contributions to Blessing, including yours. Thank you.

I'll post my most recent mashup shortly. I've included your StdDev and RSI with a little clean up.

Please be aware that the most recent versions of Blessing posted to this list are both MT4 and MT5 compatible. You simply have to change the filename suffix and recompile.

RSI is looking better and better as I test it. It stands alone quite well.

Your non-directional filter sounds interesting. I'm not sure what level of impact it will have.

Thanks.
  • #3,317
  • Oct 26, 2022 11:12pm Oct 26, 2022 11:12pm
Quoting roa
{quote} My apologies if this idea is a little ambitious, as I don't think this can be incorporated into Blessing3 easily, but I assume that most Blessing3 users have a bunch of different fx pairs running concurrently. I am not sure how many of you track the portfolio volatility of their open positions as a measure of their exposure to the market. {file}
Following up on the automation of a correlation matrix calculation, see post above:
Writing an instruction to import from an URL seems to be the easier part, see mq4 code, rows 466-500.
That is where the trouble starts. Mataf does not include XAGUSD and XAUUSD in their standard setting, for a start.
In general, customising their data download and formatting it looks like too much headache.

It is probably easier to leverage on the correlation function the original authors of Blessing implemented for the hedging function. That would keep Blessing independent of external URL. It then depends how many fx pairs the trader has in the basket, and also whether the various Blessing implementations in the account actually know of each other. In my case they don't. Also, with 10 pairs we are already looking at 100 correlation factors. It would be computationally expensive and should be done only once a day during a quiet period.

Another aspect is applying the appropriate calculation. I quoted Copeland/Weston, but they incorporated neither the confidence interval (in Value-at-Risk, VaR, it is usually 99%) nor the event horizon (in our case it would make sense to use 1 day). A better reference is therefore for instance Carol Alexander's book "Value-At-Risk".
Any comments are welcome.
Attached File(s)
File Type: zip Blessing 4 test.WebRequest.mq4.zip   28 KB | 298 downloads
  • #3,318
  • Oct 27, 2022 3:32am Oct 27, 2022 3:32am
Bug report Blessing 3.9.6.26.ex5

I was running a back test and made a mistake not marking any indicator “enabled” (in my case BollingerBands)
The test is on USDJPY 15 min
I want to trade only Long in this chart

Settings:
ForceMarketCont to Long only
UseAnyEntry is set to false
All indicators are set to Disabled
With this setting no trade should be opened but it trades as soon as it starts.
If I set ForceMarketCond to Automatic or Range no trades are opened.
If I enable BollingerEntry (or any other indicator) it works fine.

setfile attached
Attached File(s)
File Type: txt USDJPY 15min BUY 39626.set.txt   13 KB | 247 downloads
  • #3,319
  • Oct 27, 2022 4:47am Oct 27, 2022 4:47am
Quoting stiffe
Bug report Blessing 3.9.6.26.ex5 I was running a back test and made a mistake not marking any indicator “enabled” (in my case BollingerBands) The test is on USDJPY 15 min I want to trade only Long in this chart Settings: ForceMarketCont to Long only UseAnyEntry is set to false All indicators are set to Disabled With this setting no trade should be opened but it trades as soon as it starts. If I set ForceMarketCond to Automatic or Range no trades are opened. If I enable BollingerEntry (or any other indicator) it works fine. setfile attached {file}...
From my understanding, if you set UseAnyEntry = false, it allows ALL entries to open orders, and therefore it is correct if it does.
  • #3,320
  • Oct 27, 2022 12:47pm Oct 27, 2022 12:47pm
Quoting roa
{quote} From my understanding, if you set UseAnyEntry = false, it allows ALL entries to open orders, and therefore it is correct if it does.
Not sure it’s supposed to do that.
If you set FMC to Range or Automatic it doesn’t open any trades just because it doesn’t get any signal from the indicators.
The way it should be in my opinion.
Never mind, it was a mistake from my side not to enable Bollinger Bands and I was surprised it traded anyway.
0 traders viewing now