Topic: Which value does PBV take in LTF settings?

Hi Popov,

I've 2 questions:

I'm not sure whether I asked this question before (forgive me if I did) and for the benefit of me and the new users, let me ask (again?).

In 1min Time Frame (platform), I use D1 indicator (LTF).

Under the opening condition, when I input D1 condition, and which "value" is the Previous Bar Value (PBV) using for this opening condition?  Is it using D1's PBV or using 1 min's PBV for D1?

I'll assume that the PBV for D1 opening condition (under 1 min time frame, LTF features) would refer to using M1's previous bar value right to calculate D1's value?  Cos that would defeat the purpose of using 1min time frame if I'm still restricted to the use PBV based on higher time frame values.

For Example:

When I look at the MT4 D1 Chart and I see a cross just happened.  For FSB, using the 1 min time frame, does the EA execute the signal immediately on the next bar (1min) opening/closing, since the PBV for D1 already happened in the previous 1 min bar or does the FSB EA execute at the end of D1's bar?

2.  "Cross over" is behaving like "higher than"

I use D1 cross over and hence technically does it happened only once under 1min time frame (at a particular bar) at the crossing? Or would it keep "happening" till the next D1 bar open?

I noticed that when I closed all the positions of the EA (with "cross over" condition), the EA still keep opening new positions even when the cross is over?  In another words, it is behaving more like "higher than", since the cross is over and now in the "higher than" condition.  But right, it shouldn't be opening anymore positions because the conditions wasn't fulfilled.  Those new entries the EA made, didn't happened at "cross over", it happened at "higher than".  I understand that D1 is a very "slow" and not so reactive like the lower time frame and hence a crossing can "stuck there for a long time till the next bar".

I need to understand better how it is define in FSB's setting so that I can programme the EA correctly.  Kindly clarify any of my misunderstanding.  Thanks

HIT

Re: Which value does PBV take in LTF settings?

Which value does PBV take in LTF settings?

The answer is simply: FSB uses the nearest fixed value.

No matter what time frame, indicator, base price, or LTF you use, FSB always uses the nearest indicator value that will not change till the end of the bar.

Two examples:

If now is 14:34 and you use indicator calculated on Bar Open on D1, FSB will use the price from 00:00 - no Previous bar value.

If now is 14:34 and you use indicator calculated on Bar Close on D1, FSB will use the price from 23:59:999999 of the previous day - Previous bar value is on.


Literally both prices are same with and without "Use previous bar value".

Even if your main time frame is M1, MA Crossover based on D1 cannot happen at 14:34. It will always happen at 00:00 or at 23:59:999999 of the previous day, which are one and the same price.

the automatic "Use previous bar value" option is designed to protect the the traders. "Use previous bar value" doesn't produce delay in the execution.


It is very easy to see it on the chart:

Example with "Use previous bar value" on.

http://s31.postimg.org/mwihzo95j/screenshot_1536.jpg

http://s31.postimg.org/4vwwc6rmf/screenshot_1537.jpg

Example with "Use previous bar value" off.

http://s31.postimg.org/5pek1x3yv/screenshot_1538.jpg

http://s31.postimg.org/tcgwc5o47/screenshot_1540.jpg

Both position open at exactly same price.

Re: Which value does PBV take in LTF settings?

Some clarification of my previous post.

"If now is 14:34 and you use indicator calculated on Bar Close on D1, FSB will use the price from 23:59:999999 of the previous day - Previous bar value is on."

Lets the day from the above example  to be Wednesday.
You want indicator based on Bar Closing on D1.

Because FSB does a backtest, it has all data available. Without "Use previous bar value", the program will take day closing at Wednesday no matter now is 14:34, because this is what you set. But you understand, that the Wednesday's Day Closing is "looking at the future". If the program uses it, it will show you only winning trades.

The program fixes that case by setting "Use previous bar value" on. Now the program will take the Day Closing price from Tuesday. This is the only correct way. We have to use Tuesday's D1 closing instead of the Wednesday's D1 closing because the Wednesday's D1 closing is in the future from the 14:34 point of view. We literary use the previous bar closing.

...
I don't know why do you think that I make the program delaying your entries. Please be sure I'm the word's most interested person in making FSB working well and bringing profit to the traders. If I remove "Use previous bar value" (as you want from long time) our strategies will not make a single profitable trade on the real market.

Re: Which value does PBV take in LTF settings?

Dear Popov,

Let's clear something first before we discuss further, so that we trust and respect each other's good intention and not take the discussion below wrongly.

1. I've utmost respect for you and I've no doubt of your utmost intention to make FSB as excellent as possible.  I don't think you intentionally want FSB to be worst off.  But we are often stuck with certain paradigm or perspective that blind us from seeing other possibilities.  We are all capable of such and hence brain storming is a way to examine and explore and test out our perspective and widen our scope.

2. I've been a strong supporter of FSB and in no way, would want to tarnish or spoil the good name of FSB.  And if I seem to be critical, please bear in mind, I've stick with FSB all these years and I'm here to make the best out of FSB and I'm here to look for ways to improve the system.

Discussion:

I'm using 1 min time frame and I use Bar open/close for 1 min (default) and hence, I'm NOT using bar open/close for D1.  Why can't we just use the current D1 value.  I know, in forex prices are changing all the time.  One moment up and one moment down and hence, we need to be skillful to utilise different LTF and different indicators and different parameter to overcome this challenge.  Using PBV to "confirm" the value is against the whole concept of trading on ticks data and letting every minute count. 

I want at any minute, when the D1 indicator cross over, the EA is able to execute the trade based on the current value of D1, not the previous value (because based on previous bar value, yesterday, the cross over didn't happen nor the next bar opening, it would be a long delay.  I know how to handle current value and don't need the PBV to "protect me" or wait for the bar to close to "confirm" cos I've other indicators and other time frame parameters to handle such a possible "false signal", that's the beauty of FSB LTF (it is such a good feature and if one know how to use it to it's fullest, it can handle without the intervention/interference of PBV.


Please reconsider making PBV optional or re-think how to work out the reliability issue with another method?

In practice, which trader users a D1 indicator and wait till the end of the day to make the execution or use the previous day's value, it is not out "dated"?  And by the time I utilized the cross over from the previous day, didn't I already missed the breakout long ago?

Please make the PBV option available for users make their own choice and turn off the PBV function.  I can't programme my EA according to what I want as long as PBV is "interfering" with how I really want the EA to behave and calculate the market base on the current values and not always be limited or restricted by the PBV.


Please what's the harm of giving FSB users such option?  Anyway, most of us test our EA in demo 1st before we trade live.  If it is a disaster (to turn off PBV), we won't even use it in live at all.   So if you are worry about trade safe, it's the users' responsibility not yours.  Furthermore, with the option to disable PBV, you ended up having 2 version instead of 1 version of FSB for users to "play" around with.

In an experiment, we always check for validity and reliability.   And validity is more more important than reliability.  Eg. a thermometer is not a valid tools to measure weight and no matter how reliable the thermometer is, it is not a valid tool for measuring weight. 

In this FSB situation, we give up validity for the sake reliability and maybe that's why the back test result isn't a reliable results when use in real trading because it is has no validity, in live trading because EA execute based on PBV values, which isn't valid anymore (imagine using what happened yesterday to make current trade, it's not valid anymore cos market changes make it invalid, unless nothing change much between yesterday and today, i.e. ranging period).

Please think of another method to increase FSB's "reliable" back testing without compromising with validity.  Using a previous bar value of D1 for cross over, is not valid cos' the data is over, not current. 

If I use a thermometer to measure mine temperature and out of 10 readings only 1 is accurate.  Would I say this thermometer is reliable?  No way, I'll dump this thermometer away.

Likewise, if I churn out so many (eg 100) EA that tells me that my EA is profitable (according to FSB back testing) and then in live trading, not even 10% of the EA is profitable...will I say FSB's back testing result is reliable?  Does PBV really improve FSB's back testing reliability?  The best way to measure whether FSB's back testing is reliable is to compare it's back testing results with real/demo trading performance.  And obviously most of us aren't getting the kind of reliability performance we expected.  What we see in FSB's back testing results aren't what we are getting in live or demo trading.  It is frustrating for users cos we don't know whether to discard the EA or not.  Those EA that didn't look good in FSB's back testing did well in live/demo and those that has profitable back testing results failed miserably in demo/live.  There come to a point, whereby users don't know what to rely on, their own judgement or FSB's results?  And for those users who don't have a strong trading theory to rely on...FSB is all that they got as their compass to point them to the profitable EA that they have been dreaming for. 

Can you imagine my frustration...I've a very good trading theory and I'm struggling to get it out into EA in the way I wanted because I have to keep working it around this PBV limitation/restriction.

Meanwhile, can you provide longer 1min historical data, and I'll try to input all my indicator/parameters into 1min time frame so that I have only 1min Bar difference to deal with...something I can tolerate with such a minute difference.  Better still, let me have the PBV option switched off, then I don't have to start my EA all over again.  You can send me a "private FSB PBV turned off version" for me to experiment.  And who knows...maybe you are right, there is no difference and I'll let stop barking you on this issue once and for all cos I've satisfied by my own experimentation and proved myself wrong and you right.  And imagine the possibility that I'm right, won't then we have at least able to pinpoint somewhere to start searching for a better way to improve FSB's back testing reliability whereby one day, what we see is what we get?  What's there to lose, for such experimentation?

Popov, you know I'm a strong supporter of FSB all these years and hence please don't take my comments wrongly.  I'm here to highlight that somewhere, something is not clicking right.

HIT

Re: Which value does PBV take in LTF settings?

I want at any minute, when the D1 indicator cross over,..

FSB does exactly that.

D1 indicators cross over always at 00:00.
If you watch a live MT chart you may see that indicators cross over at any time, BUT THIS IS FAKE.
Please try to understand that.
D1 indicators use ALWAYS prices at the end of the day. It means that they produce signals only at the end of the day. D1 indicators cannot produce intraday signals.

MT shows that indicators crosses but this is temporary. The problem is that MT (and all other platforms) shows Close price even if it is not a Close price yet.

Please see the example below very carefully.

http://s31.postimg.org/ybhc4imvr/screenshot_1542.jpg

You have to think that this is a CURRENT price or LAST price, but there is no way it to be a Close price. No one can announce a Close price before the end of the bar.

Having this in mind you have to understand that no matter what is your main time frame, if you use D1 indicators, they will rise signals at 23:59:99 of the previous day or  at 00:00:00 of the current day (literary same moment).

Re: Which value does PBV take in LTF settings?

This is the same chart as above, but 20 min later:

http://s31.postimg.org/6iy97428n/screenshot_1543.jpg

Do you see and understand that the day is the same April 20th 2016?

If you understand that, can you explain why one and the same terminal once says that the Close price for April 20th 2016 is 1.13652 and once says that the close price for the same day is 1.13593?

The only valid Close price for D1 chart is that from 23:59:99 from the previous bar. This is the only reason why FSB uses "Previous bar values" - BECAUSE THERE IS NO OTHER RELIABLE OPTION.

Re: Which value does PBV take in LTF settings?

Hi Popov,

I don't want close nor open price of D1.  I just need the current value.  Don't need a close price to confirm anything.  Every price at it's current price is valid for me but close or open price of D1 is "invalid" because it is already out dated.  The whole idea why I want PBV turned off, is that I want the current values of all the higher time frame indicators...not their open/close values.  But the current value.  No wonder it is hard to make FSB's back testing results "reliable" because it is so difficult to integrate all these different bar open/close values for LTF features are at different time markers.  Don't tell me about PBV makes FSB back testing more reliable.  The only real evident that FSB's results is reliable is when what we see in the back testing result is what we get in real/demo trading.  That's the definition of reliability.  Otherwise, what's the real measurement or yard stick use to measure back testing reliability?  When we say FSB's back testing is reliable...what does that mean?  Is it a results we can base our decision on?  Does that mean if FSB showed a EA is profitable, it is really profitable in real test?  Can we really rely on FSB's back testing results?  Is it that reliable? 


Why are we stuck with this concept that only the price at open or close is of "valid"?  Any value is valid as long as it is current and valid.  Past prices are invalid because the market is not transacting at the price anymore.

Don't need to wait till the bar close then consider the value.  I need the current price even if it changes all the time, that's what forex tick data is, it changes all the time...price changes all the time to be relevant, that's the behaviour of the market...for a EA to "survive" and do well, it must able to take in all the information and able to handle all the current data, not out dated data...it's not valid anymore, the price has changed and the conditions also changed by the time the bar close (hundreds pips would already gone by then).  The conditions are no longer valid. And since the values from the previous bar not valid anymore, it is also not reliable to base our decision on these invalid values.

Look at the market....one hour ago it is different and the next hour, it is going to be different.  How can I enter into the market based on information/conditions that were fulfilled one hour ago?  The next hour, the price may be heading the opposite way....

If I'm on a boat and I dropped something into the ocean, do I make a mark at the side of the boat where my thing dropped and let the boat continue to drift by the current and then 1 hour later, dive down at where I marked my boat thinking that the conditions/location is the same place where I 1st dropped my thing, forgetting all the while the current and the waves have already drifted my boat to a different location.  That's what FSB seem to be doing, conditions fulfilled but act 1 hour later or use the previous hour as the value.  That doesn't make any sense in real live trading.

I have limited understanding but FSB is approaching Forex from a mathematical point of view.  But if this approach is not making the decision making process of an EA relevant, we can't bend the market to behave like your mathematical calculation.  Instead, we need to make FSB as relevant to the market behaviours, every tick counts. 

In fact, I think that FSB's LTF features (without PBV) is an excellent approach to handle the constant changing prices becos in all the "randomness" and constant changing of prices....the price does "behave" in a certain predictable pattern...otherwise, how do you think I could "predict" a breakout coming and the breakout direction?  I'm ok if we use Bar value for 1 min (cos the "lag" isn't too long).  We can then use the shift features and the LTF features that incorporate the values of higher time frame help to eliminate "false signals".  I've been working with 1min time frame and hence, I have been watching and studying the behaviours to find a way to work out these challenges.  Now my real challenge is removing PBV.

As long as I'm going to be "restricted" by this PBV, I really going to have hard time developing an accurate EA/instrument cos I have to always take into consideration of all the time lag from the different LTF indicators.  How on earth are we going to develop profitable and accurate EA with "old" data.

Now I've to think "backward" like the FSB does.  So instead of using D1 cross over, which by then it would mean the breakout already happened yesterday.  I need to use divergence or falling but that would never be accurate because the market can be ranging for many days before the cross over really happened.

HIT

Re: Which value does PBV take in LTF settings?

As long as I'm going to be "restricted" by this PBV,

Hannah, I removed the UPBV especially (AND ONLY) for you.

I hope you will be happy now.

..

Link removed.

Re: Which value does PBV take in LTF settings?

Thanks Popov,

Really appreciate your understanding.

I'll keep you updated and give you my feedback.  And at the end of the day, if you are right...I'll have the guts to admit it and would be happy to accept my mistake.  But I hope I'm right, then we could find ways to improve FSB's performance.

Whatever it is...thanks for accommodating to my persistent request and I'm really very thankful to you for going the extra mile just for me.  At the end of the day...do believe that I've the best interest for FSB too.

I hope in any ways, I can really help to fine tune FSB and find ways to improve it.

Thanks once again....my dream come true.

HIT

Re: Which value does PBV take in LTF settings?

Hi Popov

One more question....with regard to the FSB Demo data download.  The historical data for all time frame is in tick data or bar value?  If I want to truly utilize the UPBV function, I need tick data right? Is the usual FSB DEMO DATA download be sufficient or do I need to look for alternative sources to download the additional tick data in order to fully utilize the UPBV function?

HIT

Re: Which value does PBV take in LTF settings?

Hannah, when we can expect your first results?

I think your current data is sufficient, tick data is about resolving ambig. bars.

Re: Which value does PBV take in LTF settings?

Hannah,

When you look at data feed in MT4 terminal you can see (in Market Watch window) that values are changing, sometimes many times during one second. Those jumping values are used to construct bars for different time frames (M1, M5, M15, ..). But there is a difference when you are downloading historical data from your broker (or FSB data). What is available there is only constructed bars, not tick data. Which means that FSB, when searching for strategies, has only standard data bars available, not the tick data. So there are no intermittent (current, instantaneous, values between 00:01 o’clock and 23:59 o’clock) D1 values available for FSB, only four standard bar values (Open, High, Low, Close) in the data file. And if trying to use current value of D1 there is also a problem with definition of D1, because if you will take current D1 value in the middle of a day (12 o’clock), it will be in reality H12 value (and so on) not comparable with previous D1 values of defined length (24 hours). I don’t know if this value will be significant in terms of telling us something valid regarding behavior of D1 curve. It will sure tell something, but I don’t know what.

(Now, what follows here is my understanding of inner workings of Forex data feed. I think it is mostly right, even if details can be wrong in some parts. Please correct me if I am wrong. You will help me to understand Forex better)

So the tick data inside D1 (or any other data frame) is not available (in downloaded data from a broker). But there is something similar to tick data inside D1 bar – all other time frames. The closest you can get to tick data is to use M1 data bars – theoretically it is from those D1 data bars are constructed (M1 is constructed from tick data, M5 from five M1 bars, M15 from three M5 bars, and so on).

As I understand the tick data can be downloaded from some sources, but… the amount of data will be enormous for even a shorter time period, not practically useful. And FSB works only on bar data, not with tick data.

Leon

Re: Which value does PBV take in LTF settings?

footon wrote:

Hannah, when we can expect your first results?

I think your current data is sufficient, tick data is about resolving ambig. bars.


Hi Footon,

I'm in a dilemma.   

1. I've currently max out my computer capacity.

2. I've to choose whether to keep my current batch EA with PBV and run a smaller group of EA without PBV.
OR

3. Totally test out as many as my old (fortunately I've stopped deleting my old EA)  n current EA without PBV cos there is a possibility that many of my EA that didn't do well with PBV may do well without PBV, since all my EA was programmed without PBV in mind cos it slipped my mind to check that in 1min time frame, the PBV uses higher time frame n not 1min.  I somehow assume that since shift feature is using 1min PBV in LTF, I assumed it was the same for PBV likewise.  Hence all these while I input my conditions with such mind frame and wonder why my EA didn't behave as it should till it dawn upon me to clarify which PBV is being used in the LTF. 

4. The final question for me to consider is whether to test my EA "as it is" (which was manually optimized to fit the data with PBV on) or to readjust one by one or as many as I could to get the optimisation manually to fit the data with PBV OFF.  This is the ideal cos that's where I can truly measure the difference or results but manually optimize the smoothing, base n signal combination is the most tiresome part of my EA development.

Once I've sort up the above issues and run the EA and test it out on Demo,  I'll keep you all updated as long as there is enough market movement to test run these EA.  If the market has been ranging like for the 1st half of April, then nothing much happening then there is nothing much I can update.

Rest assured I'll do my best to update whatever the results, whether EA performs better with or without cos I know this experiment will be of interest to others, especially if it affects the results significantly.  Whether this experiment fail or succeed, it's nothing personal.   What's more importantly is for me to learn and under how my EA behave under 2 different FSB version and learn how to get the best out of it and programme my EA accordingly to the strengths and inner workings of FSB.

HIT

Re: Which value does PBV take in LTF settings?

Hi Leon,

Appreciate your explanation and help me understand what are the current resources I've in hand.

Guess I've to make do with what I have and at the same time, go with my intuition of how my theory work.

Thanks

HIT

Re: Which value does PBV take in LTF settings?

Hi Popov,

A few questions to clarify before I start to programme new EA to suits this UPBV version so that the EA will work exactly the way I wanted it and not having any wrong assumption what it can and cannot do.

Situations:

1.  when the PBV is removed,
a) Using 1min time frame,
b) instead of Bar opening option (for 1 min), I choose enter market at MA1 (at default, i.e. 1 min)
c) In the LTF, I use D1 cross over as my 2nd opening condition.

Q1.  If at current value, D1 has a cross over.  Will the EA enter into position at this current value or the EA still need to wait till D1 bar close to confirm the closing? 

I am assuming that when PBV is removed, EA doesn't need to take into consider D1's previous bar value nor does it need to wait for D1's bar close.  EA is able to execute anytime once the current D1's value confirmed a cross (and all other opening conditions are fulfilled too), am I right in my assumption?
       

2. When PBV is removed.
a) Using 1 min time frame
b) Setting the Closing condition.  Instead of using Bar close option, I use Trailing or exit market at Frama, Fractal or MA,those options in the same grouping as Trailing selection whereby we don't need to use Bar closing as the criteria.

Q2.  For setting the Closing Condition.  Since I'm not using Bar close, am I right to assume therefore, as long as the conditions are met, eg Frama indicator (exit market at D1's MA14), so once this condition is fulfilled, i.e. the price reached that level, the EA will exit without having to wait for D1's bar closing right? 

Q3.  For setting the Closing Condition

If I choose Bar closing (for min 1) and D1 has a crossing within this 1 min's bar but not exactly at 1min's bar close, will the EA still execute the closing at the end of 1min's Bar closing? (because technically, D1's crossing happened within the 1 min bar but not exactly at bar close, ie. at 1min's bar closing, D1's value is higher than, not cross anymore).


How does the indicator's calculate the values?

Q4.  FSB's indicator calculate it's value using the current market price/value or does it only use only bar open/close values to do it's calculation?

I think once I have a better understanding what are the rules and methods of how FSB work, I am then able to put in those opening and closing conditions meaningfully into the EA...having the right expectations of what can and cannot be done.


Is FSB capable of using tick data or is it restricted to Bar values only?

Q5.  If I were to install tick data for all time frames into FSB, will the FSB able to utilise the tick data to calculate the back testing results (using 1 min time frame with LTF features)?  Or despite downloading the tick data, FSB still uses Bar values to calculate all the indicator's values? (I'm not sure whether I'm using the right jargon, basically, as a layman...I'm asking can tick data be used in FSB?)

Q6.  If tick data can't be used or not practical (I agree, cos when I have 3 yrs worth of 1min historical data, the FSB took couple of seconds gap between each changes to reflect the results), what's the best "arrangement" or solution I can opt for, to get to the closest situation similar to that of having tick data so as to make FSB's back testing results as reliable and close to market/forex situation/behaves.

Q7.  Future consideration...a matter of curiosity.

If I were to desire to achieve or modify FSB to meet my needs/expectation.  Can I install FSB in a powerful machine, with all the tick data downloaded and I access this FSB programme from this sever/powerful machine that can churn out the results instantly?  Can FSB's user use this powerful FSB in the same manner, like the way we access the EA studio (except in a more user friendly way like the FSB Pro?)

So instead of us downloading FSB software into our computer, we access FSB online because FSB has an extremely powerful sever to handle tick data calculation and users don't have to purchase expensive computer/sever to get that level of powerful calculation etc.


Thanks once again for your commitment to help FSB users find success in their EA development pursue.  I'm grateful to your kind understanding and dedication to make FSB as versatile as possible to meet the users' needs and requirement.  One of the key reasons why FSB keeps on improving and getting better and better over the years, is because you are always pushing yourself and FSB to reach it's highest potential.  You are never satisfied at it's current achievement, you always wanted to achieve more than what it is now.

If there is an award for "Best after sales Customer Service", you and FSB deserved to be nominated and win.  I'm very certain, I'm not the only FSB user, who feels and think likewise smile  That's what really set you apart from the rest in the market.




Growth comes about as a result of constant pushing/expanding of our current boundaries/paradigm.  We can't grow if our boundaries are fixed and static.  Change is the only constant thing.

HIT

Re: Which value does PBV take in LTF settings?

The program works at the absolutely same way as before.

The only difference is that the indicators will be calculated on the current bar values instead of the previous.

The effect of that is:
when trading - the signal will happen immediately when the indicator conditions are satisfied.
when testing - the indicators will use the prices of the current bar.

Tick data work as before.

I'm grateful to your kind understanding and dedication to make FSB as versatile as possible to meet the users' needs and requirement.

FSB is not more versatile with this change. The backtests will be wrong. This is not in the interest of the users. I made it only for your experiments. I've been there 15 years ago when found that UPBV is the only correct way.

I'll not commit more to this topic because it is a waist of time.  I recommend to every one to read the documentation, the forum and to watch the videos.

Re: Which value does PBV take in LTF settings?

Thanks for your clarification

HIT

Re: Which value does PBV take in LTF settings?

Here is the current open positions of my EA using UPBV version.

I'll post this week's EA analysis results over the weekend or when these open positions are closed so that their results can be reflected in the EA analytic report (because open positions are yet to be taken into account till they are closed).

My D1 EA has been better entry accuracy and more responsive to current market changes as compared to PBV version.

Overall, I'm very pleased with UPBV version and would prefer to use this version.


http://s32.postimg.org/jwd1cz9sh/April_29.jpg

HIT

Re: Which value does PBV take in LTF settings?

Below is a comparison of EA 2413011

http://s32.postimg.org/69s9on2z5/PBV_vs_UPBV.jpg

With the PBV version (EA Original, left side picture), it opened 25 trades (1 month's trading period) and have a net loss of ($460).

Whereas with UPBV version (right side picture), it opened 4 trades (in 1 trading week) and has a profit of $4338.

With PBV version, EA 2413011 entered too late and only managed to "secure" at least 150 pips/trade and when the trend turned around, it managed to close the positions because of Break Even setting (at 150).  If there is no Break Even setting used, these positions would suffered greater loss.

With UPBV version, EA 2413011 entered accurately (without having to wait for the Previous bar values of the LTF but used the current price value) and it managed to also exit in a timely manner to secure the profits and hence, from a unprofitable EA (with PBV version), it "becomes" profitable (under the UPBV version).  Now I feel that if the UPBV backing testing show a possible profitable EA, it has a better chance that these EA also is profitable in real/demo trading.  It's too early to conclude, but I do believe that my EA's calculation becomes more accurate because it relied on 1 current price value for all indicators' calculation across all LTF instead of using so many different price values from different previous bar values of LTF.  As a result, it is more accurate and hence better predictability success...in real/demo trading.

Although FSB isn't able to provide "reliable" back testing because of the UPBV function being turn on.  I feel that the ironical thing is, with the UPBV version, it allows me to programme my EA in accordance to what I intended my EA to behave (more authentic to my trading theory and enter anytime when the conditions are fulfilled via using current value and not the previous bar value).   In doing so, the UPBV function is more suited to my trading preference and enable me to translate my trading theory into the reality (in the form of an accurate EA) and have it executed accordingly based on 1 current price value. 

Although Popov warned that with UPBV, the back test isn't reliable (because UPBV uses the concept of tick data whereas, PBV version uses bar values instead), I still think that FSB can still serves as a useful guide to help me narrow down my search via manually adjusting the smoothing, base and signal settings accordingly.  It helps to point me to the general direction (profitable direction), although it may not be accurate/reliable but at least it is still better than nothing.

Hence, with the UPBV version, I have the added advantage.

1. I not only have the back testing to serve as my guide/reference

2. I can programme the EA the way I wanted, ie. using current values for my indicators calculation instead of using PBV and hence all my opening/condition inputs reflect the current situations that I wanted to fulfilled instead of having to (guess) input the conditions that I think will happened 1 bar before the break out.

3. Because now my EA can enter exactly and immediately at D1 breakout as and when it happened, my EA has higher entry/exit accuracy which imply higher profit, profit factor and win/loss ratio. 

Currently, I have about $320,000 floating profit from all my other EA that haven't closed their positions and hence no results in for me to do the EA analysis.  I'll update here once these results come in.

An observation: In the past, I racked my brain very hard to figure out a "Universal EA" and had hard time achieving it under the PBV version.  Now under the UPBV version, I am observing some of my old discarded EA (unfortunately, in the past I haven't discovered Fx Blue yet for my EA analysis tool and hence no statistical comparison) being able to behave like the way I wanted it to be, an "All time round, universal EA".  I feel very relief and very happy that with UPBV, my EA has the possibility of reaching it's highest potential.  And that I can find greater success via FSB's UPBV version.  I feel I'm in "greater control" now.  It's a very nice feeling to have smile  A sense of Empowerment (with UPBV) instead of frustration (with the PBV).

Finally, I can do what I want with my EA smile

ps: I'm not here to promote UPBV version.  Since Popov allow me to have the UPBV as an experimentation, I'm here to report my own observation/reflections with regard to this experimentation with UPBV.  Which version is better is very subjective to one's trading theory and strategy approaches.  Hence, one's man meat is another man's poison.  As Popov has given his warning, use it only if you know what you are doing.

UPBV version, in my own personal opinion is a HUGE potiental gold mine.  Huge potential if we can start using tick data for all the calculation, then FSB's UPBV version will be closer to the Forex trading environment, for us to use as a "simulator" to train and develop our EA.  At that level, I do think there will be no competitor that can come so close to FSB's capability.  We are so close yet so far (till Popov is willing to go in this direction).

HIT

Re: Which value does PBV take in LTF settings?

Reporting on my UPBV experimental findings...


The EA analysis for the 1st week trading below:
http://s32.postimg.org/g2bdkq8up/EA_Analysis_UPBV_May_3.jpg


http://s32.postimg.org/xvfjfwdn5/EA_Analysis_UPBV_May_3_p2.jpg
Above. This is taken a few hours later.  Scalper EA 10522 made another $5k within a couple of hours (from $17k to $25k)


http://s32.postimg.org/buoht0byp/EA_Analysis_UPBV_May_3_p2.jpg
Above.  And less than 30 minutes from the above 2nd screen shot.  Scalper EA 10522 made another $4k (from $25k to $29k).  Such lighting speed reaction to market changes and excellent performance would not be possible under the PBV version. (Don't think it's because of the 30 min trading results but rather, because the last EA10522 was closed 30min ago, therefore, the new results is reflected in the latest report)


1. My "Universal EA, Scalper Mode took the top 5 spots.  In the past, my "Universal EA" hardly profitable or "sleeping" all time (for those with very strict entry conditions and due to the "interference" of PBV calculation of various LTF values instead of using 1 current price value across all LTF).

2. EA under UPBV has better results than PBV version.

3. There are still some EA not closed yet for full reflection of the overall results.  There are still an estimated floating profit of $982,000 as illustrated below (Two screen shot from computer 1 and computer 2)
http://s32.postimg.org/awake7csx/Floating_point.jpg


4. EA with D1 and H4 LTF has better performance as a result of timely execution (not using PBV but use current price value and hence can pinpoint the breakout with better accuracy at the point when the breakout happened and not after the breakout happened in it's previous bar).

Computer 1
http://s32.postimg.org/727lst18x/May_3_p2_C1.jpg


Computer2
http://s32.postimg.org/42x0up09d/May_3_p2_C2.jpg


Comparison between PBV and UPBV version using EA 1060 series

Below is a comparison, the left side show EA 1060 series results under PBV version and on the right hand is the UPBV results.

EA 10602005 under PBV has profit factor 0.66 whereas under UPBV has profit factor 6.12


http://s32.postimg.org/4192lro7l/EA_1060_series.jpg

NOTE:  The above results are SUBJECTIVE to one's theoretical approach and strategies used. 

These EA were developed to target for breakout and hence it required the utilization of the actual current price value to do it's calculation (among all the LTF indicators) to pinpoint the breakout and hence this kind of strategies can't work well in PBV version (not because PBV is bad but there isn't a "fitness" between the EA approaches and the PBV "system/way of working". 

Whereas, under the UPBV version, I'm able to "programme" my EA closer and more "authentic" to my theory and the UPBV system "allows" my EA to execute in the manner that I want (using current price values instead of previous bar values).  Now there is a "fitness" between my trading approaches and the FSB UPBV's way of working.  UPBV can "carry out my instructions" accordingly to what I want my theory/EA to do.

A sound and profitable trading theoretical framework will produce profitable EA
That's my belief.  And thus despite the UPBV version giving "wrong back testing results", as Popov has warned.  The ironical part is, now I experience more of the "reliable" results that I always been hoping for...i.e. "What I see (profitable EA in FSB back testing) is what I get (in demo/live testing).

Yes, the UPBV version back testing can't give you the "actual results' because it uses bar values for back testing calculation, while your EA in MT4 platform will be using actual tick data and current values for calculation (hence in my term, I'll call it as "incompatiable" back testing comparison, which in Popov's definition, "unreliable")  but it still can point you to the general profitable direction.  And because now my EA is behaving more "authentically" to my profitable trading theory and hence I see better results and performance.

Many perspective to success.

Are we willing to expand our perspectives?
As I've said, "one man's meat is another man's poison".  I am so adamant for the use of UPBV and Popov is so adamant against it.  My meat, his poison.  Popov said he will not want to "waste" any more of his time in this discussion.  I wonder whether he means to say he is closing up his mind and refusing to expand his perspectives?  I hope not.  He is always open to new ideas and ways to improve FSB.  I hope likewise, he won't closed up his mind from willing to "see" things differently.  When I was 15 years younger, I see things very differently from now.  I hope whatever he has concluded 15 years ago, he would take some time to "re think" about PBV, now under different "light" and findings (from my experimentation).  Since he is willing to open up this UPBV for my experimentation.  Won't he be willing to open up his perspective to examine the results?

Don't limit our choices.  Not fight over preferences.
I personally feel, there is no need to fight if we are given choices.  Instead of "either or, neither nor approach", we could have just leave the choices to the users' discretion.  If one can find success with UPBV, we shouldn't deprive other's of such chances.  If PBV function can be switch on and off.  We capture 2 segments of the traders' market.  Instead of having only 1 product to offer, we can have 2 versions.  Any way, both versions are the proud products of Popov masterpiece, isn't it?

Capturing more market's share via offering many other "competing" products is part of marketing approaches.  There are some companies, recognizing that there are different preferences among customers and knowing they can't make a product that is "one size fits all", hence, they offer "competing" products to suit different types of users.  Take for example the different brands or types of shampoos and you will be surprised, some of these "competing" products exactly belong to the same company.

Which is of higher importance to a trader?

Authentic Execution or Back test Results?
I hope FSB will learn to cater to different types of trades.  There are some traders whose highest priority is to have a system that is able to "replicate" their EA accordingly to their theory and relying on back testing is their secondary priority, not the other way round.

When you fly a model RF air plane,  won't you first make sure that your RF air plane is "calibrated" correctly so that when you use the RF remote control, your air plane move accordingly to your command?  When this calibration is achieved, you know, you have greater confidence that you can get better results and performance isn't it? (because you are in control).  You won't care so much if you can't control your RF plane that way you wanted, even if I tell you that it's has excellent features, it is able to flip, etc. because you know, you cannot maneuver the plane very well if your remote control is not calibrated to follow your command exactly.

Likewise, with PBV, your EA's formula isn't "fixed".  It is variable.  Because the price values calculated are based on different PBV time gap/difference of the various LTF and these time gap changes all the time...never consistent and hence, your EA's will not have a "fixed" formula to produce any consistent pattern/curves.  And hence, the EA performance is not as consistent as those observed under UPBV.

HIT

Re: Which value does PBV take in LTF settings?

Forgive me if I have a few questions:
- How do you get to these results? Is it a FSB-Backtest, is it a MT4-Backtest, is it a Demoaccount, is it Realaccount, or what is it?
- Why do you call your special version (where you do not use the previous bar value) "UPBV"? As far as I understood this means "Use previous bar Value"
I admire your enthusiasm about your special version, but is it not too early after not quite two weeks to be so convinced?
You take a suprisingly strong oppinion against Miroslav's experience, which I do not find so constructive and spreads uncertainty.
If you think this can improve my trader-life, I would appreciate if (besides the super statistics) you would publish here at least one of your good EA's so I could verify and test myself.

Re: Which value does PBV take in LTF settings?

Hi Bru1,

1. These results are MT4 demo account testing based on current market situations.

2. UPBV is coined by Popov, (basically it mean without PBV feature/function, i.e. PBV function turned off)

3. I'm convinced by the way the EA is able to behave, ie. authentic to my theory and therefore I've great confidence that this is the path I would prefer to take (via the UPBV version).  Let me ask you, if you have two cars and you are at the steering wheel, one car is "calibrated" and the other is not calibrated to do exactly as you wanted, for eg. If you turn  your steering wheel to the left by 1 inch, sometimes the car turned a few more degrees to the left  than you expected and at times it turned a few degrees lesser than "usual".  Which cars do you have more certainty and greater confidence in? The calibrated car or the "un calibrated" car.  Do you need to drive for 2 weeks in the "un calibrated" car in order for you to make up your mind? 

UPBV would do exact as I wanted it to be, use 1 current value across all indicators (LTF) calculation to derive at the answer, whether a cross over or a breakout has happened.  Whereas the PBV version, will utilize many (not just 1 but several) price values from different sources, taken from different time reference (different LTF has different PBV, and these time gap varies all the time).  Do you think with such method of calculation, your EA can determine with accuracy whether a breakout or a cross over has really taken place at that exact timing?

4.  Which is worse? To be certain of something that is "wrong" or to be uncertain of something that is "right"?  If you have a totally different experience from someone, who do you trust more, your own experience or others?  Popov and I have "conflicting" experience.  If I can do better with the UPBV than the PBV version, what should I do?  Continue to use PBV because Popov says so? Or do I let the facts tell me that it makes better sense to use the UPBV version.  Are you saying that Popov is always right and could never ever be "wrong"?  Anyway, in this experimentation, I'm here to promote the benefits of UPBV and not to prove Popov wrong cos UPBV is also Popov's wonderful masterpiece that I would prefer to use.  I'm not here to say how terrible FSB is, but I'm here to say how Good FSB UPBV is...aren't I here to credit one of Popov's masterpiece?  I'm not here to discredit him.  In fact, I'm giving much credit to UPBV, a very neglected version/system that is worth considering.

What "uncertainty" have I raised in others?
To make people question something they think they are so sure about?  Is that wrong?  What if in the process, I make people re-think and re-discover a hidden gem in UPBV, isn't it good?  If people begin to do their own experimentation (with UPBV) and realised that UPBV can make their EA results improved by leaps and bounds, won't that be beneficial, that I've "rocked" them out of their comfort zone?  Won't they, have to thank me for shaking them out of their perimeters and do something different and get better results?  Same input = same output.  If we kept doing the same thing (thinking) over and over again, we will keep getting the same results over and over again...no improvement.  Wherefore will we ever discover anything new or better?

As I've already emphasized, the results I demonstrated is VERY SUBJECTIVE.  It depends on the person's trading theoretical framework, types of strategies etc.  There are many who do well with PBV (so did I, but with UPBV, I did better).

5.  If my "challenges" is to pushed the envelop and expand perspectives, then it can also be viewed as constructive.  However, if my "challenges" is viewed as rocking some people's boat, then what can be shaken, will be shaken.  The strength in one's perspective lies in facts.  When one is presented with the facts and still refused to change his/her perspective, then there is no point in having a meaningful discussion.  We have decided to closed our minds and refused any form of reasoning to enter into our minds...that where the term "unreasonable" means...a refusal to be reasoned.

That's why I posted my results so that people don't think I'm just stating my opinions but I backed it up with stats and facts.  If it's just about opinion and preference...there is nothing to argue.  Everyone is entitled to their own difference styles and opinions.  But when we are here to examine the effectiveness of a system...facts is what we are talking about.

6.  You quoted "If you think this can improve my trader-life, I would appreciate if (besides the super statistics) you would publish here at least one of your good EA's so I could verify and test myself"

Qns:  You want to verify the UPBV version/system or you want to verify my EA?


You can verify the UPBV version yourself without the need of my EA.
Ans: I'm here to demonstrate how the same EA can have different (and maybe better results) by comparing the two systems.  You can do this experimentation yourself.  Taking any of your own EA and test it out under these two versions and see for yourself whether there is any difference and make your own conclusion.  As I've already said several times, it all very subjective (depending on the "fitness" of the system (PBV or UPBV) to one's trading approaches) .  If someone has a lousy EA and test it out in UPBV and get terrible results, whose fault?

I'm here to "convince" Popov that the UPBV is worth the shot and to allow users to exercise their own decision/discretion,  whether to turn the PBV function on or off according to their trading approaches instead of a limited choice of only having the PBV version.


I've already proven to myself that my EA works.  I don't need to prove this to any body.
I'm not here to "convince" anyone that my EA is good.  All these years I've been testing my theoretical framework and I already have total confidence of it's profitability.  But all these years, I'm "restricted" by PBV function whereby I can't able to programme my EA exactly the way I wanted because I can't bypass this PBV function and wasn't given the choose to use the actual current value instead.  Now with UPBV, I can. 

The reasons I have to show my EA results is to:

1. Show the facts that my EA has good results under UPBV version too, so that UPBV is worthwhile to consider opening it up to users to achieve better trading results (If they can, why not?).  If Popov's customers can have far better trading results via UPBV, does it make any sense for him to "deprived" his customers of such success?  As I've said it before "If I can't, doesn't mean others couldn't...If I failed, doesn't mean others can't succeed.  It's a matter of different approaches.  I don't know what approaches Popov tried that made him concluded that PBV is better than UPBV.  I'm here to say, that the results/faults may lie in the trading approaches and not in the system (UPBV). 

If my trading results can "opened" up the possibility that UPBV can be a better system (due to one's trading approach), then I hope that will make Popov re-think, Is UPBV that "bad" for FSB users that he wanted to prevent users from having it?  If I can get better results via UPBV, then is it really that "bad" to have better results?

2. To provide comparison of the same EA under these 2 versions to "prove" that my EA do better without PBV "interference"

3. That I am a good trader who knows what I am doing, able to identify and "capture" the trend with my EA (able to develop profitable EA with both versions).  And I'm NOT someone, who is a "bad worker blame his/her tools" mentality, I am not someone who is "struggling" and clueless about how to develop profitable EA and then blame PBV for "blocking" me from doing so.  I've demonstrated several times...I'm capable to getting excellent results...so I do know what I want and what I'm aiming for.

For your own experience, test out the versions yourself and see which system works better for your trading approaches.  There bound to have some, who still prefer PBV version.  And that's fine.  That's what I'm trying to promote here, Give users the choice to choose which version is better for their EA/trading approaches.  As for me, I here to report my findings based on my own trading approaches.  I'm not here to say that PBV is bad but UPBV is better for me, as a trader...my own personal trading preference, I'm not here to impose my trading style on others (cos I'm not here to share my trade secrets with anyone, i.e. teach other trading tips = Yes.  But giving my trade secrets away = NO).

Bru1, thanks for seeking more clarification.  I hope I've answer all your queries accordingly.  And if you have better results from your UPBV version, it would nice if you also highlight your experience here as to show that mine is not an isolated case.

HIT

Re: Which value does PBV take in LTF settings?

Thank you for the answer, please keep us informed about your results, good luck.

Re: Which value does PBV take in LTF settings?

Hi Bru1, thanks for taking an interest in this experimentation.

I won't be posting any more results here cos I've already given some examples for comparison between PBV version and UPBV version.  I believe I have stated my choice and reasons why I would rather use the UPBV version very clearly. 


My intention and desire is to have this UPBV made available (not only in the form or purpose for experiment) for all users to choose.  The ability to turn this function on and off accordingly to one's trading preference (instead of having to adhere to Popov's trading preference/perspective).  We all have different strategies and approaches, my belief is that a product need to cater to different needs/traders. 

For me, the need for a system/version to execute my EA or trading theory as authentic to it' framework/formula is of highest priority and having a backtesting results is a secondary, not the other way round, whereby in order to ensure and organize the historical data to produce "reliable" results via using Previous bar values (instead of tick data), the EA is "restricted" and "force" to only act within the boundaries of PBV so as to produce the "reliable" backtest results because FSB's historical data is limited in the form of bar values and not tick data.  We "bend" our EA to fit how the historical data is organised.  In this process, we actually compromised with our EA's accuracy (calculation based on bar values and not tick data for back testing and calculation based on previous bar values instead of current price level in demo/live testing) and speed of execution (close when Bar close, and not as and when the tick data fulfill the trading conditions).


The purpose of the experiment is to test out my EA under which version has a better "fitness" with my trading theoretical framework and approaches (hence, the results and performance are subjective to one's EA approaches and doesn't apply across the board).  I have demonstrated the UPBV is a better fit for most of my EA. 

And I have no intention to provide any more results here because I believe if those who determine to close up their minds, no matter how much facts is presented, they just don't see the line of reasoning.  To those who are open to try, no one will able to stop their curiosity to search for better trading platform.  I'm glad I'm given this special privilege and opportunity to test out something I've believed very firmly all these years...UPBV is better suited for my trading approaches/framework.  I'm convinced beyond any further of doubt, now I just hope this opportunity to use it for the future upgrades is not deprived from me as a FSB customer and from others who also wishes to have alternative method of developing and executing their EA closer to their trading theory.

I've no intention to impose on any body to use UPBV, it's their prerogative, a choice one can only make for themselves whether UPBV suits their trading preference/approaches, as I've already said "one man's meat, another man's poison".  Likewise, I do not wish anyone to imposed on me to use PBV version when UPBV is a better choice for me.

I hope Popov will value customer's feedback and take their views into consideration, so as to allow his customers find greater success in their trading experience.  Isn't what this forum is about?  To provide open communication and brain storming so as to refine the system?  Is such channel of communication closed?

HIT

Re: Which value does PBV take in LTF settings?

Hi bannedUser,

Surely LTF is there to allow you to enter the market the minute (if you are set to M1 TF) the longer timeframe condition occurs.  You could achieve the same by multiplying the MA periods (assuming a MA cross strategy) by 1440 to convert your 1440 minutes to a day.

At least that's what I thought.  Your post has highlighted that the LFT refers not just to the entry conditions but also to the entry candle.  eg if you have LFT set to D1 and you are trading on a M1 TF then the BAR OPENING refers to the daily bar, and not the minute bar.

If this is the case then I can't see the point of the LTF option.  Why wouldn't you just use the D1 timeframe without the LTF option?

I can understand Popov's reasoning that bar opening with "use previous bar" and bar closing are the same thing, so bannedUser, are you saying that by selecting "Bar opening without use previous bar" you get the current price used as the entry condition.  Is Popov saying that this may be case but it make the backtest unreliable?