Currency Strength Indicator

Happy New Year. We begin 2026 with the development of a new MQL4 indicator designed to acquire, structure, and reliably store a .csv file containing the operational data required for external processing. The indicator integrates a Currency Strength calculation module, which serves as the primary source for generating the strings written to the .csv file.
The entire system is organized into three main components:

1. Data Collection, Calculation, and Normalization (MQL4)
The indicator uses a Currency Strength algorithm to compute the real‑time relative strength of the major currencies. For each tick or significant update:

1. the strength values of the individual currencies are read
2. the relative strength of the current pair is calculated
3. filtering and smoothing logic is applied
4. a structured string is generated containing:
   1. timestamp
   2. symbol
   3. strength values of the two currencies
   4. strength differential
   5. logical state (reversal, continuation, neutrality)

5. the string is written to the .csv file in tabular format, ensuring compatibility with external systems

This module guarantees that the dataset is coherent, clean, and ready for advanced processing.

2. External Processing via Python Application
The .csv file is then processed by a dedicated Python application implementing an advanced analysis pipeline:

1. dataset parsing and validation
2. value normalization and missing‑data handling
3. generation of additional features (derivatives, differentials, slope, local volatility)
4. inference through a neural network trained on multi‑symbol historical datasets
5. classification of the current market state (reversal, continuation, signal strength, operational probability)

The neural network uses the currency‑strength strings generated by the MQL4 indicator as its primary input, ensuring a coherent and data‑driven workflow.

3. Automatic Telegram Notification Generation
Once processing is complete, the Python module evaluates the neural network output and, when relevant operational conditions are detected, automatically generates a professionally formatted Telegram alert containing:

1. event timestamp
2. symbol and timeframe
3. currency strength values
4. neural network classification
5. model probability/confidence
6. any additional metrics useful for operational decision‑making

The full workflow ensures robustness, traceability, and consistency across modules, forming a continuous pipeline:
MQL4 → CSV (Currency Strength) → Python → Neural Network → Telegram

In the coming days, I will publish additional updates and provide all the necessary material for anyone who wishes to test the system.

Until next time.

MRP — Market Relative Pressure is an indicator developed in MQL4 to analyze directional market pressure in the Forex market through a structured and reproducible approach.
When MQL4 is pushed beyond a certain level, the platform tends to use a significant amount of PC resources and does not always allow new strategies to be developed efficiently. After several attempts, I finally found a personal method to overcome these bottlenecks.
The name MRP refers to my previous professional background, where systems, processes, and structured analysis were an integral part of daily work. Today, as a happy retiree, I have chosen to maintain that same methodological logic and apply it to the study of financial markets. In my previous job, we used a management system that was very effective at handling the database but had a limited interface. Data processing took place overnight, and in the morning the results were exported to Excel, a more powerful and flexible environment necessary to manage the large volume of information.
The new system I am developing uses the MQL4 indicator without any graphical interface: it is lightweight, fast, and designed not to overload the PC’s memory. It adopts the same data‑export method used in the company management system, saving everything in CSV format and then processing the data with a powerful language like Python.
The goal is to create an environment that is easy to program and maintain, also leveraging the support of an AI called Copilot.
The indicator generates 31 columns, divided into 8 functional areas.
At the moment, Python processes only part of the CSV: it uses 4 out of 31 columns to produce the 3 inputs required by the neural network. The columns actually read are:

1. BAL → used as input a
2. ΔBAL → used as input b
3. bid (current price) → used as price
4. prevBid → used as ref_price

Technically, Python reads 4 columns, but the neural network uses only 3 normalized values:
CSV ColumnUsed by PythonRole
BALinput a
ΔBALinput b
bidinput price
prevBidinput ref_price
All other 27 columns✘ignored

How to Install Anaconda on Windows — Complete Guide
1. Go to the official website
Open your browser and visit:
https://www.anaconda.com/products/distribution (you can also search for anaconda.com on Bing)
Click Download.
2. Download the Windows version
Choose:
Anaconda Distribution – Windows – 64-bit
This is the standard version for all modern PCs.
3. Run the installer
Once the file has been downloaded, double‑click:
Anaconda3‑xxxx‑Windows‑x86_64.exe
(where “xxxx” is the version number)
4. Follow the installation wizard
I recommend these settings:

1. Just Me
2. Add Anaconda to my PATH (if available)
3. Register Anaconda as my default Python

These options make everything easier and help avoid conflicts between different Python versions.
5. Complete the installation
After installation, you will have:

1. Anaconda Navigator → graphical interface
2. Anaconda Prompt → terminal already configured
3. Conda → Python environment manager

Verify the installation
Open Anaconda Prompt and type:
Codice
conda --version
If a version number appears, Anaconda is installed correctly.
During the installation, some additional patches may be downloaded.

After installing Anaconda, do you need to create a folder for your Python files?
Yes, it is good practice to create a dedicated project folder. It is not required for Anaconda to work, but it is the best way to stay organized and avoid confusion.
Why create a dedicated folder?
Because it allows you to:

1. keep all your project’s Python files together
2. store your MRP CSV file in a consistent location
3. avoid path (directory) errors
4. maintain a clean and reproducible working environment
5. make it easier to use Git or perform backups

This matches perfectly the structured, methodical approach you are applying to your system.
How to create the project folder
Go to a convenient location, for example:
Codice
C:\Users\Luigi\Documents\
Create a new folder, for example:
Codice
MRP_Project
Inside this folder you will place:

1. your Python files (main.py, neuralnet.py, etc.)
2. the CSV file generated by MQL4
3. any logs or output files
4. additional scripts

How to tell Python to work inside that folder
Open Anaconda Prompt and type:
Codice
cd C:\Users\Luigi\Documents\MRP_Project
Now you are inside the project folder.
From here you can run:
Codice
python main.py
or any other script.

“I recommend installing Notepad++, a lightweight and very practical text editor for editing Python files, CSV files, and configuration scripts.”

“This will be the programming environment.”








I had to change the file extension from .py to .txt because the forum does not accept the .py extension.
Don’t be afraid to use Copilot when you start installing everything. It’s an extraordinary assistant and incredibly effective at helping you solve any issues. Let me know if you’re interested in this system so we can improve it further.

If the installation completes successfully, this is Python in operation.

Greetings. See you next time.

MRP Monitoring – Detected Signals
The system is monitoring in real time the CSV file generated by MetaTrader. Each line represents an operational signal for a currency pair, with the following parameters:
PAIR: currency pair (e.g., USDJPY)
AGE: age of the signal in hours
BAL: directional pressure (positive = long, negative = short)
SLOT: breakout or pattern type (e.g., Up618, Dw1618, box)
RATING: visual summary of signal quality (from + to +++++)
DESCRIPTION: qualitative assessment of strength, maturity, pressure, and stability
At the end, the system selects the “Prince” signal, meaning the strongest and most coherent one, ready for Telegram delivery.
The system continues to monitor and update automatically.
First test on Monday.

See you soon.

I followed the steps but it didn't work. Unfortunately, I don't have much information.
I still wanted to thank you.

python libraries are missing.

Since you are using Anaconda, generally the following command should work in that terminal
if successful you'll see it, you can additionally check the version just in case:

The code's author (AI) had to write it in requirements.txt (it's probably inversioni.txt on his screenshots) and the human middleman was supposed to upload the file here and explain how it works.

Don’t give up.

Just do something simple: copy and paste the returned text into Copilot and follow the instructions.
You’ll see that everything will be resolved.

Hi, I stumbled across your Python system, and I installed everything without any problems.
System updates every 5 minutes, OK.
I just don't understand your PowerShell screenshot ??
Did I miss a step?
Thanks anyway for your work
I'm eager to test it on Telegram.

Hi! Great to hear that you managed to install everything without any issues and that the system is updating every 5 minutes — that means the core setup is working perfectly.
About the PowerShell screenshot: no worries at all. You didn’t miss any step. That screenshot is only meant to show what the system looks like while it’s running, with the PowerShell window open and the logs scrolling. If your system is already running and updating regularly, then everything is exactly as it should be.
By the way, I’m getting ready to publish a new version with several fixes and improvements. It’s almost ready, and once released you’ll be able to update and enjoy an even more stable experience.
If you need help setting up the Telegram part, just let me know.

I compressed the file.py scripts into a single archive. The archive now contains all the main components of the MRP system: the monitor, the CSV reader, the FORCE bar builder, and all the other modules required for the complete operation of the platform.

MRP — SHORT SUMMARY OF THE MQL4 FILE ====================================
Purpose The MQL4 file is the data‑engine of the MRP system. It reads market data from MT4, calculates all MRP metrics, and writes them into a CSV file used by the Python monitor. This is a pilot version and it is not finished.
What the MQL4 file calculates • FIBO structure (box, UpXXXX, DwXXXX) • AGE (signal age in bars) • BAL (directional pressure) • CSI levels (Daily, Weekly, Monthly) • FORCE raw values (activation, consolidation, acceleration)
How it works

1. Reads price data on every tick.
2. Computes all MRP metrics through internal functions.
3. Detects meaningful changes (FIBO, FORCE, CSI, BAL).
4. Writes a new CSV row only when something relevant changes.
5. Keeps the CSV updated for the Python monitor.

Output Each CSV row contains: PAIR, TIME, AGE, BAL, CSI_D, CSI_W, CSI_M, FORCE1, FORCE2, FIBO
Role in the system The MQL4 file is the sensor. Python is the brain.
Without the MQL4 file, the monitor cannot update, detect events, or send Telegram alerts.
==================================================

See you next time.

Hi, Everything is working on my end, Telegram too!

__________________________________________
I'm testing the RISK=LOW alerts on the dashboard



__________________________________________
the monitor alerts today.


____________________________
I've been taking position all the alerts since 9am


how filter the signals sent to Telegram?

Thanks a lot for this tool, I enjoy testing new systems!
sorry for the traduction

Principe MRP
NZDJPY
FORCE: ·|□■|■□|□□
AGE=0.1 BAL=0.64
FIBO=box

Active signal — MRP Prince
NZDJPY
FORCE: ·|□■|■□|□□ AGE: 0.1 BAL: 0.64 FIBO: Box forming, no breakout detected.
Status: no operational conditions at the moment. Waiting for the box breakout.
Reason: The signal is too young (AGE 0.1), the balance is still stable (BAL 0.64), and the FORCE structure shows no directional coherence across timeframes. With the FIBO box still intact and no breakout, market pressure is insufficient to justify an entry. The system remains in observation mode until a clear structural break occurs.

1. First Signal — “ MRP Prince” (09:07)
Codice
NZDJPY
FORCE: ·|□■|■□|□□
AGE=0.1 BAL=0.64
FIBO=box
Meaning
AGE 0.1 → the signal has just appeared; it is still immature.
BAL 0.64 → the market is in near‑neutral balance, with no dominant pressure.
Mixed FORCE structure → the timeframes are not aligned; no directional coherence.
FIBO = box → price is in compression; no breakout yet.
Operational Interpretation
The system detects a potential movement, but:
no direction is confirmed,
the box is intact,
FORCE is incoherent,
market pressure is insufficient.
This is why the alert is classified as MRP Prince:
a young, early‑stage signal with no operational conditions yet.

2. Second Signal — “ MRP STRUCTURAL CHANGE” (15:03)
Codice
NZDJPY
box → Up236
FORCE: ▲|■■|■□|□□
AGE=0.3 BAL=0.68
What changed
Box → Up236
→ the price breaks out of the compression zone and chooses direction.
FORCE becomes coherent on short/medium timeframes
→ ▲ indicates rising bullish pressure.
→ ■■ confirms strength on two timeframes.
AGE 0.3 → the signal is maturing.
BAL 0.68 → balance shifts toward bullish dominance.
Operational Interpretation
The system detects a true structural shift:
the box is broken,
directional pressure aligns,
FORCE becomes coherent,
the trend structure strengthens.
This triggers the stronger alert:
MRP STRUCTURAL CHANGE
→ meaning the market structure has changed in a measurable, actionable way.

Full Sequence Summary
MRP Prince (early signal)
No direction
Box intact
FORCE incoherent
No trade conditions

MRP Structural Change (mature signal)
Box breakout
Direction defined (Up236)
FORCE coherent
Pressure increasing
Potential setup forming

Hello,

Day 1

Test I trade everything Trading risk = LOW Starting at 9:30 AM (Paris time)
It's AMAZING

Day 1 test Telegram alert


I'm letting it run to see how it behaves. Is it possible to have that "low risk" alert in Telegram?
I've looked through the files a bit (and even translated them into French!), but I haven't found what I'm looking for!

Thanks again.

Regarding your question: yes, it’s absolutely possible to include the “LOW RISK” label in the Telegram alert.
Right now the Telegram message only shows the symbol and the signal because the risk‑level variable isn’t being passed to the Telegram module yet. The indicator already knows the risk level internally — it’s simply not added to the final message string.
This risk‑mode concept was already part of the very first version of the indicator, so I’ll retrieve that piece of code and integrate it into the new version. Once added, the alert will automatically include the risk level as well.

See You next time

Thanks a lot !!!


I trade everything. Trading risk = LOW. Starting at 3:30 PM (Paris time), which is 9:30 AM US opening.
After 2 hours


I've tested several systems, but this one is very promising...

Note:
The file MRP_telegram_bot.py is not included in this release.
Please continue using your existing Telegram module, as it remains fully compatible with the current version of the system.

MRP Operational Risk Calculation

1. Purpose of the RISK Module
The RISK module of the MRP system assigns an operational risk level to each signal, based exclusively on the data contained in the CSV file.
Its purpose is to identify:
strong and reliable signals (LOW RISK)
uncertain or incomplete signals (MEDIUM RISK)
dangerous or inconsistent signals (HIGH RISK)
Risk is calculated automatically for every structural event (RE, BOX→FIBO, PRINCIPE).

2. Variables Used in the Calculation
Risk is determined by five key elements:
1) TIMONE (primary direction)
The first character of the FORCE BAR.
▲ or ▼ → clear direction
· → no direction → higher risk
2) AGE (signal age)
Measures how fresh the signal is.
low AGE → young and reliable signal
high AGE → tired or noisy signal
3) CSI_D_DELTA (currency strength variation)
Indicates whether the currency strength is increasing or decreasing.
positive → real momentum
negative → loss of strength → higher risk
4) BAL (directional balance)
Measures the real pressure between the two currencies.
high BAL → solid direction
low BAL → weak signal
5) FIBO (structural level)
Indicates the phase of the movement.
Up/Dw levels → clear direction
box → congestion → higher risk

3. Classification Rules
LOW RISK
A signal is LOW RISK only if all the following conditions are true:
TIMONE ≠ ·
AGE < 8
CSI_D_DELTA > 0
BAL > 0.20
FIBO ≠ box
Interpretation:
Fresh, directional signal with real momentum and pressure.

HIGH RISK
A signal is HIGH RISK if even one of the following conditions is true:
AGE > 20
BAL < 0.10
CSI_D_DELTA < 0
FIBO = box
TIMONE = ·
Interpretation:
Weak, tired, inconsistent signal or lacking direction.

MEDIUM RISK
Everything that is not LOW and not HIGH.
Interpretation:
Intermediate signal: not dangerous, but not ideal.

CAMBIO STRUTTURALE MRP
GBPCHF
box → Up236
FORCE: ▲|■□|■□|■□
AGE=0.4 BAL=0.40
RISK=MEDIUM RISK

MRP Update – GBPCHF
On GBPCHF, the model is detecting a developing structural move: the box reads Up236, signalling that bullish pressure is building with solid intensity.
The FORCE reading shows a consistent directional bias: ▲|■□|■□|■□ The overall push is upward, with moderate strength across all major timeframes. It’s not an explosive move, but a steady, orderly rise that’s taking shape step by step.
The AGE = 0.4 tells us the signal is still young: the structure isn’t mature yet, leaving room for further development. The BAL = 0.40 confirms a moderate imbalance in favor of the British pound, meaning the move is directional but not extreme.
The model classifies the scenario as MEDIUM RISK: an interesting setup, with multi‑timeframe alignment and a signal in its growth phase, but still requiring caution and price‑action confirmation.