Bear&BullQuantCartel–MT4/MT5 Quant Analysis Lab FX Algo&Manual

Before anything else
If you are interested in my original thread where I explain my workflow in detail:
https://www.forexfactory.com/thread/...t-real-systems

That thread is focused on:
- how I build systems
- how I test them using EA Studio and StrategyQuant
- how I move from backtest → demo → live
- the full workflow behind my trading

This thread is different.
Since the beginning of that thread, a lot has happened.
A lot of work has been done.
Over time, I reached a point where I wanted more than what EA Studio and StrategyQuant provide.
They are powerful tools.
I still use them.
They are my builders.
But I wanted something deeper.

I wanted:
- deeper analysis
- deeper stress testing
- clearer insight into real system behavior
- more structured decision-making before putting capital at risk

So I built my own environment.
An institutional-style quant analysis lab.

Over time, the project evolved into something much bigger than the original idea.
What started as a simple analysis concept slowly became a broader quant analysis environment with deeper tools, better structure, and more serious testing layers.

The analysis tools themselves will remain free.
No subscriptions.
No hidden access.
No paid features inside the tools.

I already have free guides available, and I also have a paid full EA Studio guide.
I want to be transparent about that.
That full guide is not something I can simply give away for free, because it contains years of mistakes, lessons, testing, trading psychology, workflow, and everything I have built in my head around EA Studio and system development.

That does not change the purpose of this lab.
The tools remain free.
In the future, I will also create guides around these tools:
- some may be free
- some may be paid

But there will be no subscription model, no locked tool access, and no pressure to buy anything.
The tools are there to use.
The guides are optional.
Just to be clear, this does not change what I said in my original thread.
I am not selling robots, signals, or any hidden strategy access.
The tools are free.
Any guides are separate and optional.

Why this lab exists
The goal is simple:
To answer questions that basic backtests and standard metrics do not answer clearly enough.

For example:
Is your portfolio actually stable?
Can it survive stress?
Where are the weak points?
Which robots are carrying the risk?
Which ones should be reduced, watched, or removed?
How does performance degrade under pressure?
What does your trading really look like beyond surface-level metrics?

This is not about:
“Is it profitable?”
This is about:
“Is it structurally sound?”

What this lab does
This lab is built to analyze:
- MT4 / MT5 statements
- portfolios with multiple EAs
- individual robots (symbol + magic)
- manual trading results

You can take your own history and see:
- how strong it really is
- where it breaks
- how it behaves under stress
- what the real structure looks like

Development
This did not start as something big.
It started simple.
Basic Monte Carlo.
But that was not enough.
So I kept building.
From a basic framework → to a structured system:
- better tools
- better design
- deeper analysis layers
- clearer outputs
The goal was always the same:
When someone uses it, it should feel like:
“I am actually testing this on a serious level.”
Not just running another basic report.



Current state

The full lab is live and the tools are available.

At this moment, the tools are built to work with both MT4 and MT5 statements.

Just to be clear, I am sharing the lab link because the tools are free and directly connected to this thread.

This is not an affiliate link, not a sales page, and not a promotion for robots, signals, or subscriptions.

You can access the lab here:
https://www.thebearandbullquantcartel.com

However, I want to be very clear about the current development level.
The EA Analyzer / Monte Carlo page is currently the most complete and most deeply tested part of the lab.

That tool has already been pushed much further:
- MT4 and MT5 parser compatibility tested
- full statement and filtered statement behavior checked
- Monte Carlo settings tested
- preset behavior tested
- confidence table output checked
- drawdown and survival-related calculations improved
- institutional drawdown overview added
- export layer added for PDF, JSON, and separated CSV outputs
- bugs and edge cases fixed during validation
- output structure improved for review, archiving, and comparison

So right now, the EA Analyzer / Monte Carlo page is the part of the lab that has been brought closest to the institutional level I had in mind.
The other tools are available and working, but they will continue to receive deeper upgrades, testing, and refinement over time.

The goal is not to pretend everything is finished forever.
The goal is to keep improving the full lab step by step, while keeping the tools available for everyone to use.


Thread note

This thread is not meant to become a support desk or a place for endless basic questions.
I will keep improving the tools and I will explain the main updates here, but the tools themselves are built with explanations inside the platform as much as possible.

If someone has a serious question, a useful observation, or a real point about analysis, testing, robustness, MT4/MT5 statements, Monte Carlo, or portfolio structure, discussion is welcome.
What I want to avoid is noise, ego posting, vague criticism, or people turning the thread into arguments without actually testing or reading properly.

If you have ideas, improvement suggestions, or a bug you want to report, you can contact me directly at:
[email protected]

The goal is to keep this thread clean, structured, and useful

The Bear & Bull Quant Cartel – Institutional Quant Lab Overview
Below is a first look at the environment.

Main Interface

Attached Image (click to enlarge)

EA Analyzer + Monte Carlo (Combined Workflow)

Attached Image (click to enlarge)

Strategy Intelligence Report – Deeper Breakdown

Attached Image (click to enlarge)

EA Breakdown / Robot Scanner + Drawdown Survival Simulator

Attached Image (click to enlarge)

Advanced Analysis Lab – Equity DNA Analyzer + Portfolio Builder

Attached Image (click to enlarge)

More detailed breakdowns and examples will follow.

Development Progress – Part 1: From Basic Monte Carlo to Structural Analysis
I want to give some context on how this lab actually evolved.
This was not built in one straight line and it did not start as a full institutional framework.

It started much smaller.

The first idea was simple:
take real MT4 statement data and run a deeper Monte Carlo analysis on closed trades.

At the beginning, the focus was mainly on the basics:
- parsing real statement data
- rebuilding equity behavior
- running Monte Carlo simulations
- reading robustness from actual trade history

But very quickly I noticed something.
Basic Monte Carlo was not enough.

A simple reshuffle or resample can be useful, but it does not answer enough questions.
It can show how results change when trade order changes,
but it does not fully show how a strategy behaves under deeper structural pressure.

So the first major step was improving the Monte Carlo engine itself.
The sequence side was expanded with:
- reshuffle
- resample
- block bootstrap
- combined sequence mode

That made it possible to test different path-generation styles instead of looking at only one type of simulation.

After that, I separated the engine into two different layers:
Sequence layer
Stress layer

That separation is important.
Changing trade order is not the same as adding real execution pressure.
So the sequence layer looks at the trade path itself.
The stress layer adds pressure such as:
- skipped trades
- slippage
- spread cost

That moved the tool away from simple “what if the trade order changes?” testing,
and closer to:
what happens when the path changes and execution pressure is added on top?

After that, the confidence table and interpretation logic were expanded.
I added better reading for:
- profit retention
- drawdown expansion
- Return / DD degradation
- streak behavior
- tail-risk outcomes
- worst 5% and 10% behavior

The goal was not only to show more numbers.
The goal was to make degradation easier to read.

A strategy can still look good on the surface, but if profit collapses too fast, drawdown expands too aggressively, or Return / DD breaks down under stress, then the structure is not as strong as it looks.
That is where the framework started moving from basic simulation into structural interpretation.


The next step was preset logic and verdict modes.
I added:
- Soft Mode
- Hard Mode
- Survival Mode

The preset modes control the Monte Carlo test profile itself.
Soft Mode is a lighter structural scan.
Hard Mode is stricter.
Survival Mode is the harshest stress profile.
Those same mode ideas also help the verdict layer read the output more clearly.
So the tool no longer only runs simulations.
It starts judging the structure under different levels of pressure.
That was a major turning point.

Development Progress – Part 2: From Structural Analysis to Decision Framework
Then I started building beyond Monte Carlo only.

Inside the Advanced Analysis Lab, I added:
Equity DNA Analyzer
Portfolio Builder

The Equity DNA Analyzer focuses on the actual behavior of the equity path.
Not only:
- profit factor
- winrate
- net profit

But questions like:
- is the curve smooth or choppy?
- does recovery behavior look healthy?
- is the result too dependent on a few bursts?
- is the path psychologically ugly even if the numbers look fine?
- does the equity curve show hidden pressure?

The Portfolio Builder was added because single-system analysis is not enough.
A robot can look average alone but improve a portfolio.
And a good-looking robot can still damage the portfolio if it adds too much correlation or similar risk.

So the Portfolio Builder looks more at:
- combined equity behavior
- portfolio drawdown effect
- return / drawdown
- correlation between robots
- better and worse combinations
- combo optimization

That pushed the lab further into portfolio thinking instead of single-system vanity metrics.
After that, the Monte Carlo engine was pushed again.

This is where the more institutional-style layers were added.
I wanted the framework to test pressure in a more realistic way.
Because real stress is not clean.
Pressure tends to cluster.
Bad periods come together.
Correlation can rise when you need diversification most.

Systems that look independent can weaken together.
Some systems only break in specific market conditions.

So the engine was expanded with deeper structural layers:
- bad month clustering
- regime-based analysis
- portfolio component stress
- correlation pressure
- survival / death detection
- tail-risk summaries

That changed the purpose of the tool.
It was no longer only asking:
what happens if trades are randomized?
It started asking:
where does the system break?
how does it break?
under which conditions does it weaken?
does the portfolio still survive when components interact under stress?

Then the verdict system was upgraded.
Earlier versions gave data.
Later versions started interpreting that data through layers.
The tool now looks at:
- structural robustness
- survival behavior
- deployment quality

That became the three-layer verdict framework.

Instead of one simple verdict, the system reads the strategy through different lenses:
Structural
Survival
Deployment

That matters because a system can look structurally acceptable,
but still fail when survival pressure becomes too high.
After that, the decision layer was added.
This is where the tool moved from analysis into decision support.

The final decision block can classify output into:
- DEPLOY
- DEMO FIRST
- HIGH RISK
- REJECT

That does not mean the tool replaces human judgment.
It means the framework helps turn many stress layers into a clearer next step.

That is the point.
Not just:
what happened?
But:
what should I do with this structure now?

After that, the scenario stress engine was added.
This was another big step.
Because market pressure does not usually arrive as one isolated variable.
It can come as:
- crash shock
- volatility expansion
- liquidity drought
- broker degradation
- combined scenarios

The scenario engine was built to test hostile pressure in a more event-based way.
Not only random stress,
but more realistic combined stress.

On top of that, dynamic sizing behavior was added.
That includes things like:
- drawdown-based reduction
- streak-based reduction
- volatility-scaled behavior
- equity curve feedback
- capped compounding

This matters because exposure itself changes the structure.
Some systems survive fixed exposure,
but become fragile once compounding or adaptive sizing is added.

Then came the allocation layer.
Because even if a portfolio survives,
there is still another important question:
how should capital actually be distributed?
So the allocation engine was added with:
- Conservative mode
- Balanced mode
- Aggressive mode

And then the allocation output was expanded with:
- allocation input summary
- allocation comparison
- allocation verdict
- allocation-weighted Monte Carlo curves
- baseline vs allocation comparison

This made the framework more practical for portfolio decisions.
Not just:
is this robot good?
But:
does this allocation make the full structure better or worse?

The latest major stage was validation, MT5 compatibility, export, and institutional drawdown.
At that point, the framework was already strong inside the interface.
But once you left the page, the structure was harder to review.

So I added the export layer:
- structured PDF reports
- full JSON export
- separated CSV exports

That makes it possible to store runs, review results later, compare outputs, and inspect different layers outside the page.

Then I added the Institutional Drawdown Overview.
Because max drawdown alone is not enough.
Two systems can have the same max drawdown,
but one recovers quickly while the other stays underwater for a long time.
That difference matters.

So the drawdown review was expanded with:
- Ulcer Index
- Time Underwater %
- Recovery Factor
- MAR Ratio
- Median DD %
- 75th % DD
- 90th % DD
- 95th % DD

That gives a better view of:
- how painful the equity path is
- how often the system stays underwater
- how efficiently it recovers
- whether drawdown pressure is rare or structurally present

At the same time, MT5 support was extended across the platform.
So the tools are now built to work with both MT4 and MT5 statements.

The framework was validated across multiple use cases, including individual systems, portfolios, and different stress scenarios.
So the development story is simple.

This started as a basic Monte Carlo idea.
Then it became:
- a linked EA Analyzer / Monte Carlo workflow
- a structural stress-testing engine
- a verdict and decision framework
- a portfolio and allocation review layer
- an exportable analysis environment
- an MT4 / MT5 compatible quant lab

That is the direction.
Less surface-level reporting.
More structural analysis.
The goal is not to make results look better.
The goal is to expose what is really happening underneath the result.