USDLFX Offline Chart Creator EA

Unfortunately MT4 platform doesn't support ticks history. So, the only way is to run the chart creator EA continuously on a VPS. Another way, which is an approximation, is to use M1 data to estimate wicks. However, this requires us to have access to a deep history of M1 data which is usually not available on Broker's servers.

Thanks for your observation.
It will be the same issue if you shift to MT5?
I'm almost sure ninjatrader platform has tick data by default.
Thanks

You're welcome!

Theoretically it might be possible to copy ticks in MT5 for this purpose, but the current EA has been developed for MT4. I do not use NinjaTrader, but I've heard that the users have access to the history of ticks on that platform.

For the moment, the current EA cannot reconstruct the wicks in the absence of ticks history, unless we want to do some ugly solution to deal with this problem.

Best,
Matt

Hi Merka,

There is no need for an Expert Advisor to reproduce the LFX indices in MT5. You can create custom symbols in MT5 including wicks, like a standard chart. The formulas are written here https://www.forexfactory.com/showthr...5#post10673445

View -> Symbols -> Custom -> Create Custom Symbol

USDLFX in MT5:

Yes
It's exactly what i did using your suggestion in an early post and the results are very very good but I used the correct formula for the arithmetic mean as clearly Liteforex USDIndex is mathematically wrong and I managed to re-invert the JPY chart.

USDIdex: pow(((USDCAD*USDCHF*USDJPY)/(AUDUSD*EURUSD*GBPUSD*NZDUSD)),(1/7)) if you want the results like in Liteforex just remove NZDUSD from the formula
AUDIndex: USDIndex*AUDUSD
CADIndex: USDIndex/USDCAD
CHFIndex: USDIndex/USDCHF
EURIndex: USDIndex*EURUSD
GBPIndex: USDIndex*GBPUSD
NZDIndex: USDIndex*NZDUSD
JPYIndex: USDIndex/USDJPY (is not annoyingly inverted like in Liteforex)

Thanks

Hi Merka,

Great job that you made it work and thanks for your suggestions.

Why do you think that the formula for USDLFX is mathematically wrong. Is it because NZDUSD is not in the formula, or is it something else?

Please share the offline candle stick chart (period converter) in which I can run EA's
I already have some scripts and indicator which converts H1 chart to H2 offline, however, EA's are not working on it.

The LiteForex formula is not wrong, it is smart.

1. geometric mean:
   [code]USD-FX6 = (USDAUD * USDCAD * USDCHF * USDEUR * USDGBP * USDJPY ) ^ 1/6
   USD-FX7 = (USDAUD * USDCAD * USDCHF * USDEUR * USDGBP * USDJPY * USDNZD) ^ 1/7
   NZD-FX7 = (NZDAUD * NZDCAD * NZDCHF * NZDEUR * NZDGBP * NZDJPY * NZDUSD) ^ 1/7[/code]
2. LiteForex:
   [code]USD-LFX = (USDAUD * USDCAD * USDCHF * USDEUR * USDGBP * USDJPY * 1) ^ 1/7
   CHF-LFX = (CHFAUD * CHFCAD * CHFEUR * CHFGBP * CHFJPY * CHFUSD * 1) ^ 1/7
   or
   CHF-LFX = USD-LFX / USDCHF
   NZD-LFX = NZD-FX7[/code]

Only with the LiteForex formula (one additional factor of "1" and the 7th instead of the 6th root) it is possible to calculate all other indexes quickly from the USDLFX. Not only the known ones but also new ones. Imagine you want to add the TRYLFX index. No problem, you just need USDTRY. With the mathematically correct formula you can't do that and need much more calculation resources for calculation of all indexes (i.e. all crosses).

How does this additional factor of "1" influence the resulting chart shape? Not at all. The "shape form" is absolutely identical except that the USDLFX chart (LiteForex) is a little bit compressed compared to the USDFX6 chart shape. Chart form and therefore indicators, chart patterns, signals, relative support and resistance levels are 100% identical.

If you use 6 pairs in the arithmetic mean calculation you need to use ^1/6.
Liteforex use 6 pairs but also use ^1/7 and with that I don't agree but also why not taking into account NZDUSD?

I agree that purely mathematically it is not correct but it doesn't affect the resulting chart. Using ^1/6 vs ^1/7 is a cosmetical question but using 7 massively simplifies further calculations.

The Kiwi doesn't count as a "major currency", or said in another way the value of NZDUSD doesn't define the value of the US Dollar to a major extent. Again imagine a TRY-index. Would you argue that now the formula has to change to ^1/8? No, TRY just replaces NZD.

Hi MathTrader7,

I've been reading your threads and I'm extremely interested in learning more about your LFX charting, do you know of any brokers which trade LFX? Although the ones I have asked, don't allow trading, any ideas?

Thanks.

Damian.

[quote=MathTrader7;8237451]What is USDLFX? USDLFX is a currency index of US dollar formulated by LiteForex broker. It is an attempt to create an indicator for USD currency power.

Happy new Year and Merry Christmas to all Christians.

Hello Mathtrader7

I would really appreciate your comment on my inquiry please. You are one of very few who would comment on the difference between ADDing charts and multiplying them.


In more details:

I have been delving into "Synthetic Currency" Indices from LiteForex.

However, in tradingview, there is a feature that allows you to ADD different pairs. So for example: I can add EURUSD + EURGBP + EURAUD + EURCAD + EURJPY + EURNZD + EURCHF. and the result would be one chart with normal candlesticks to be analysed on different time frames.
(Thanks for Degolep at forexfactory for telling me about this feature)

I noticed that if I compare Euro pairs (by adding them) , the chart will be different from the EUR Index available on trading view. I can tell why there is a difference. The index is not calculated by adding pairs but rather by multiplying them.

My observation is that the addition of pairs result in a more realistic view of the currency which is better than the normal formula used for calculating a currency index by multiplying them.

What do you think of this. Have you thought of it previously? Also: do you know any MT4 indicator that would enable me to ADD pairs to produce a synthetic chart?

What I like about this feature in Trading view of (adding/subtracting/multiplying and division) is that I can control which currencies to include/exclude and which math I can use (i.e multiplying or adding, etc..).

For example I can compare three currencies against each other: (EUR/USD/GBP). So for EUR I would type the equations as follows : EURUSD+EURGBP to produce a chart of EUR index (relative to the other two currencies) or I can type EURUSD*EURGBP/2.
Of course the options are countless. If I want to give more weight to a to EUR against pound, I can type:
EURUSD+2*EURGBP, or I can multiply each pair with a 0.XY and so on to create a weight of my preference.

The whole idea for me is to "customise" my chart and create the best PA like a good breakout followed by a 123 , clear S/Rs, etc.. So instead of being limited with what is offered for me, I will create my small basket in a simple chart that has the best looking clear setups/divergences etc...

Take your time and do not rush a reply.

NB: I have attached the Canadian Index (A built in index in Tradingview)
and Attached a Summation of all Canadian pairs.
Both are weekly

I have also attached EURo Index from tradingview and another one by adding EURO pairs (both are daily).


NB: Regardless of the difference of adding vs multiplying in tradingview, I have noticed slight differences between (your indicator which is almost the same as Lite Forex charts) Versus the original tradingview available built-in indices. Can u also comment on that.

Iam sure it is all about math formulas and what Arithmetic equations are used. But which one is the best or may we should look for a confluence?


NB: I have found this link that did not really give me an edge in understanding what is going on (adding vs multiplying). It is more about spread trading which is popular in UK as there is no tax on it. https://www.tradingview.com/wiki/Spread_Charts

Thanks in advance,

[quote=emmzett;11593978]{quote} I agree that purely mathematically it is not correct but it doesn't affect the resulting chart. Using ^1/6 vs ^1/7 is a cosmetical question but using 7 massively simplifies further calculations.


Hello enmzett,

I am glad I found this thread because I feel I will finally find an answer for my long search.

If you can please have a look at my previous post to Mathtrader7 and write a comment, it will be very appreciated.

I am trying to find the value in adding pairs rather than the multiplication. And at the same time, why are indices constructed by multiplying?

I would love to know more about the math involved. I will squeeze my head to utilize my studies in math and economics 20 years ago

NB: I have found this link that did not really give me an edge in understanding what is going on (adding vs multiplying). It is more about spread trading which is popular in UK as there is no tax on it. https://www.tradingview.com/wiki/Spread_Charts

Thanks in advance

Hy, I tried to answer your questions in the other thread.
https://www.forexfactory.com/showthr...2#post11840112

Thanks emmzett


BTW,

I had to copy and edit my posts in the other thread to post them here in this thread.

Do you mind please emmzett to keep every thing posted here for the time being. I am eager to hear the comments from all professionals including you and Mathtrader7 and other valuable members.

I think this thread is more professional and has and will produce more ideas that will benefit us all.

Thanks

@Mathtrader 7


If you don't mind I would share some of posts here from another thread as Iam really looking forward for your comments when you have time.

In short there is a debate about which is better: Multiplying charts to produce an index or adding charts to produce a synthetic charts that I have found more sleeker and more realistic.

Here is an example of what I mean. I shorted CHF JPY. My logic was that CHF Index (trading view index or Lite Forex one) were showing a bullish momentum, while adding CHF pairs was showing a bearish momentum and hence I shorted the chf/jpy pair.

Any way, I kept reading online about the difference between arithmetic average vs geometric mean.

I found couple of authors more favouring geometric means. However, Arithmetic is not deemed to be quite wrong as we also use it for Simple Moving Averages. Otherwise, we would have multiplied all the data series and divide them by the total number instead of adding them. Saying this makes me ask if such Moving Average does exist. I mean multiplying all data series for N periods and diving them by N. I am sure coders can easily do this.

I however, believe that addition would generate an index that is more close to what is happening now while normal Indices provided by brokers are bit late.

---------------------------------------

Here is a trade I took on shorting CHFJPY. It made 200 plus pips and could have netted 500 pips.






With JPY, both the built in index (Lite Forex and Tradingview Index) and "Adding all Yen Pairs" (my approach), have shown a confluence with showing a bullish momentum. So I will not discuss the JPY Index.

What I want to discuss is the CHF Index.

Now notice the built in CHF Index (from either Lite Forex and tradingview), Before the 2nd of January, the daily chart was showing a clear upward momentum with a big possibility of continuation. However, a trade could examine this chart and argue there was a Resistance on smaller time frames H4 and below or any reason to say that CHF bullish upthrust was slowing down and that there was a Daily Resistance and price has almost hit the Resistance.

But, my argument would be that for me, CHF Index was bullish and I do not want to "Force" myself to believe it is bearish by applying too much analysis on LTF. This chart simply will never make me believe that CHF is bearish but it is rather bullish or at the very least it has hit a Resistance and is reacting it. But no bearish trend has been established yet.



But If we check CHF synthetic chart by adding all CHF pairs on tradingview, the story is quite different. CHF is indeed very bearish and there isn't a hint of any bullishness.
Look how smooth the PA of the daily chart is. A break of the TL and a retest and enough room for a more nosedive.

The difference between CHF built in index and adding CHF pairs is very clear in this example. CHF is indeed very bearish.



So again: Take this example as a means of me explaining what I "visually see" rather than arguing it is a correct method.

Please note, I use indices for direction bias and momentum and also to basket trade selected pairs. I can even trade one pair rather than a basket.


Edit: I just wanted to add here is that in "my observation", I found that addition gives an early bias for direction compared to multiplication which is more than often giving late entries. I might be wrong though. I have not tested it for long.

Sometimes, I find adding charts give an overall smoother look with less erratic movements . in other words, the combined chart looks cleaner.

---------------------------------------------------------------------------------------------------------------------------------------------------------------

This is another question for you please:

Generally speaking, if I choose to trade a basket of selected pairs, and I want to put the stop loss at a level where I expect price would be changing its bias. How can I calculate the amount of money that would be lost? I am not a fan of a using a certain amount of dollars as a stoploss because it does not reflect where the (Technical Analysis) stop loss should be. I want to put the Stoploss on the "Customised or synthetic chart". Let us say that I have a synthetic chart for couple of pairs. I have a breakout and a retest. If I am buying and want stoploss to be behind the last swing, How to translate this into (amount of money as a loss ) and hence I can specify my lot size to trade?


Thanks in advance

Regards,

@Mathtrader7 and all dear valuable traders

Sorry for being lengthy and adding another post but again, I appreciate your professional work and want to hear your comments please.
I have so many ideas for Renko as well but this I will post in the right thread later on.

I wanted to discuss Measuring Momentum or Acceleration of Currencies


Now what is it about?


For any trader who is analysing currency (Indices) to find bias and momentum with the aim for trading one pair. The trader wants to pick a bullish one vs a bearish one. Right?

That is good and is a good way of analyzing but in many cases it limits the number of trades.

Lets imagine for example that both EUR Index and USD Index (from the indicator we have here in this thread or from Lite Forex or any other) are both bearish. So the trader would think better not to short EURUSD. But soon enough the trader would find that EURUSD did fall significantly.

Later on, the trader would find that EUR was more bearish than USD and thinks to himself "Next time , I will not look at indices", "I should not have missed this setup. I knew EURUSD was going to fall because of my pre-analysis to the pair".


Any way, I thought of sharing this to ask you traders: How would you traders examine the "Rate of Change" or the "Momentum" or what ever you may call it to tell which currency is more bearish than the other and that it is ok to short the first against the other.

The logic here would be : Just avoid those two currencies. Bothe are bearish" . But this comes with a price : Limiting the number of setups".
I myself am between the two mindsets. Or lets say I would place all the odds in front of me and adjust my risk or lot size accordingly.


Any way, I would love to hear your comments please.


--------------------------------

Add to this that sometimes when the trader picks the right bullish Index and the right bearish one, it may be a bit late to trade the pair. It has already moved.
May be if the trader picks the right bullish one vs a "Ranging" one would be a good choice? your views will be great.

Nevertheless, there is also the risk that the trader might have done a wrong analysis for the index and established a wrong bias. (This is part of the job)

Again, would love to hear your views and comments and how have you experienced different scenarios

---------------


Another thing I have been thinking of is wether we can plot currency Indices by the same logic of using Logarithmic charts rather than Linear ones or by the logic of using PPO versus MACD Oscillator. You get the idea? We just examine cuurencies based on their Percentage Move. This will apply on Low, High, Close, Open. nd we can plot a chart based on this.

The percentage move will be our reference point and hence we can compare the strength of all currencies against each other.

What do you think?

I wish if my math was better but it has been a long time. I wish if I could write detailed equations to explain this recent idea.

BTW: I am still working my head around the arithmetic average of many instruments and try to find why they look smoother and whether they provide a fair reading or not.


@all other traders

I found an interesting article for all traders who have been reading the previous posts and wondering what is it all about.

This article explains the difference between arithmetic average, geometric mean, and the harmonic one. It is purely math and not related to forex. Read it and you will understand the logic.
https://towardsdatascience.com/on-av...is-2a703e21ea0

The author is not favouring arithmetic average and favours the other two. But there are many online reads that do not diminish arithmetic average as a means for creating Indices for economic data. (There is no talk about forex charts). It is a more of a preference from one to another. Have a look yourselves and share your views.


Thanks for every one
Regards

Look at those two charts of USDLFX and a synthetic charts adding all USD pairs.
Any way, check the 21of dec and look at previous PA.

In my view, the USDLFX may have shown some bearishness before the 21 Dec but was not convincing for me. The trend line break was not enough and there was a clear range.

The second chart however is showing a much clearer picture. PA is more spread out. am looking for visuality or in other words I want to be a chartist. I want the picture to be clear so I can make my decision.





On the other hand,

I have here another example that is showing the CAD Index to offer a better view of a bearish bias on the 12th of Nov compared to adding the CAD charts.

The index is showing a smooth clear break of the TL while the second chart does not clearly show this on the 12th. The signal came after this date..

It is a matter of visuality which ever gives a clearer picture. In this example, the built-in index wins.

Very interesting things here.

Let's say the market follows the same rules for every kind of pairs.

Then AUDUSD GBPAUD EURAUD ( AUD Pairs ) and so on , will Stop at the same Resistance or support. Right ?

Then the AUD index HAS TO stop at this Resistance level. Right.

If not, the index is wrong.

I know it may sound like unclear. lol

But the for the index to be right it need to be calculated in percentage I think.