AI ORASignal - First test results (we are always in profit)

Traders, hello

So, let’s once again highlight the forecast we made earlier.

One of our AI modes accumulated a series of positions on GBPUSD, and we, in turn, added some manual positions of our own.
Part of those trades has already been closed in profit, while others remain open.
Importantly, all positions continued to be placed in the direction of the bearish trend

And most of the currently open trades are still far from their profit targets — which is exactly where the main intrigue lies.

Of course, manual execution would likely be ideal here, and the fully manual account is already performing extremely well.

But our goal right now is different:
We are testing the AI mode itself — specifically f30.
Can this particular mode independently maintain the kind of stable break-even and profitable structure it has already sustained for more than half a year?

So, here is our position:
We expect the AI to be right.

If temporary “pain” appears along the way — then perhaps it simply needs to be endured.

That is our logic.
And that is exactly what we are testing

Our projection:
Risk off ⇒ USD ↑ ⇒ EURUSD, GBPUSD ↓
Especially in the case of GBPUSD ↓

Waiting and observing
Good luck to everyone.

And that’s it

The forecast has now been fully confirmed.

We talked about the possibility of a GAP — potentially right from the Monday open.

And that is exactly what happened.

We prepared for it in advance and placed higher TP levels (as we also mentioned back on Friday).
Moreover, we did not even wait for the main AI positions to fully close, since we had already secured profits from this series that exceeded the original targets several times over
(And the screenshot reflects only today’s results — while Friday had already included a large series of successful shorts.)

But the important question is this:

Was the AI actually right?
Yes.

The level was EXACTLY the same.

Price returned to it almost like a magnet.

And another important point:

Should the AI be “penalized” for those specific trades?

Of course not

Those trades would now also be in profit — had we not manually closed them earlier out of excitement and with already substantial overall gains

As they say:
Q.E.D. — quod erat demonstrandum
(“which was to be demonstrated”)

Moving forward

It is important to highlight the main point.

Why yet another forecast turned out to be correct.

At first glance, it may have seemed that everything was against the AI’s logic (and we are talking about the previous day — Friday, when many traders even turn off their automated systems).

However, the key point is that the AI’s logic is based on a unique internal metric.

This metric consistently demonstrates a very high win rate (and importantly, accounts on Forex Factory cannot be faked or manipulated).

Therefore, it is important to understand that behind the AI’s words (its forecasts and decisions) there is statistically verified decision quality.

This cannot simply be “broken” — especially considering more than 10,000 trades with a 96% win rate behind it — and it should certainly not be ignored.

Unless, of course, someone can show comparable accuracy results

In short — we feel more calm working with it
Our confidence and sense of stability are directly proportional to the level of its ongoing performance

GBPUSD — a new forecast

Let’s make another projection — and, as always, verify it in real time

So:
The AI placed short positions, which means that according to its internal logic, the 1.35580 level on GBPUSD should eventually be crossed.
There is even a feeling that we may see a stronger move — perhaps even a pump in that direction.
(Although that part would also depend on external news factors, which do not always align with such developments.)

So once again — we wait, observe, and verify

Important to note


Of course, geopolitics matters.
Of course, the USD today is still “the USD” on one hand — while also facing pressure from weak military outcomes and broader uncertainty on the other.

There are thousands of active factors, narratives, and forces currently interacting and shaping the market through constantly shifting balances and weighted probabilities.

That is all true.

BUT — this is exactly the kind of complexity that ORA AI already processes internally.
Balances, probabilities, weighted structures, conflicting forces — all of it.

Which is why we are not speculating on those themes ourselves.

This is the AI’s view.

Its forecast.

And now the question becomes:
Will its projection on #GBPUSD once again prove accurate?

We continue to observe and verify

This is our new AI mode — f70
And just during today alone, it has already generated 52 pips

Everything continues to perform very well, just as before.
(It does not publish signals to Telegram yet, but the Forex Factory account is publicly available for full transparency.)

Interestingly, today it identified both bearish and bullish trends.

Which naturally may raise a question:
How can an AI simultaneously generate signals in opposite directions?

But once again — this is exactly where statistics matter.
A win rate above 90% suggests not that the AI is “confused,” but rather that it is detecting multidirectional market behavior as it develops in real time.
At the moment:

1. the bullish side already has a noticeable buffer,
2. while the bearish side still does not.

Which aligns with what we discussed earlier — and also matches the broader logic of the other AI modes.

The expected larger trend remains bearish.

And our view is that sooner or later — but most likely quite soon — that bearish move will fully develop

Ready to place a bet on it?

Let’s wait and see

GBPUSD — the forecast was fully confirmed Again...

And in less than 24 hours.

All target levels were reached.

Honestly, even we are sometimes slightly stunned by how strange this all feels… because it actually works

In practice, what we keep observing is this:
The AI identifies a level that seems to “pull” price toward it and effectively suggests that the level will eventually be crossed.
In roughly 90% of cases, this happens almost immediately — within 20–30 minutes.
In more difficult situations, if opposing market pressure appears, it simply takes more time.

But after thousands of trades, we still have not encountered a case where the projected level ultimately remained unreached.
That is what makes this so fascinating.

It is unusual.
It is powerful.

And it suggests entirely new possibilities when combined with traditional approaches.

Very cool

Suddenly, there is now a flood of “analysis,” opinions, and market commentary everywhere about GBPUSD and EURUSD

All of it retrospective.
All of it after the move already happened.

Meanwhile, we were talking about this back on Friday.
More than that — positions were actually placed.

And even more importantly:
profits were made both on the bearish move and on the bullish reactions along the way — while consistently maintaining that the broader bearish trend would still develop.

Not after the fact.
Not with “hindsight intelligence.”

But beforehand — as a real forecast.

Honestly, this is what sometimes becomes deeply irritating about many “analysts.”
(Yes, genuinely irritating.)

Being smart after the market moves?
Seriously?

Ah yes… because they are afraid of damaging their reputation by making inaccurate forecasts.

But that itself creates a false reputation:
looking intelligent only in hindsight (?).

True reputation is built differently.

When you speak before the outcome — that is when your real value and real credibility become visible.

Personally, we find it amusing to imagine building success on something that does not actually work.
Either it works — and the results speak for themselves.

Or how else could someone keep the attention of thousands of people with nothing but empty commentary?

The main AI modes (f9 and f30) are currently active — but they are not seeing what the NEW mode is already detecting: f70

Its account is now public.

And its statistics are excellent

We think it is now time to begin publishing its signals in our (still free, for now) Telegram channel

This mode is new, and some might describe it as “more aggressive.”
We do not really like that term.
For us, there are only two important distinctions:

1. more active,
2. and whether it actually works.

This mode is definitely more active — and it definitely works.

Yes, it is CLEARLY different from the others.
Some trades do not appear as “instant” as what we are used to from the earlier modes.

But facts are facts:
at this moment, it is outperforming all the others.

And importantly, it appears to identify future levels and developing movements just as accurately.

So — we are officially starting to publish signals from this mode in our still-free Telegram channel

All good and dandy, but where can I see your long term drawdown.
That is the only metric that really really matters, first
Thank you

Thank you for your comment

The current maximum equity drawdown is 7.3% (Max Return Drawdown: -7.3%), and all trades are fully visible in the public tracker.

There are no hidden floating losses, no martingale, and no long-term averaging positions.

Regarding the idea that only this metric really matters — I completely disagree.

It matters, yes.

But only as part of a broader context.

We have an account with more than 10,000 trades and a 96% win rate — and that also matters.

Multiple currency pairs, any time of day, full operational freedom for the AI — and still such consistency in outcomes suggests it is structurally “reading” the market.

Personally, I place the most weight on profit factor.

From an entrepreneurial perspective, I only care about one thing:
is there profit or not?

Today there is profit, tomorrow there may be drawdown, then profit again, then lower drawdown — markets are inherently dynamic.
We try to move away from judging things in isolation.

All metrics matter.

And when they are aligned and consistently green together, that is when “really really matters” starts to apply.
Not individually.
Not occasionally.
But collectively.

At that point, the real question becomes how to extract more efficiency from the system — because randomness starts to play a much smaller role

The most interesting part of this entire story is that each AI mode has its own separate statistical track record.
In practice, what we currently publish is a combined account. The AI sends trades, and we also trade alongside it — on the same mode account — which technically “blends” the statistics and even distorts them a bit. In a sense, we are even “spoiling” the pure data, which is the funny part
On the individual trades, AI positions are marked with unique comments.
However, Forex Factory still displays the full picture — both AI trades and our own manual additions.
Ideally, these should be separated into different accounts:

1. one purely AI,
2. one purely manual.

And we may still do that.
For example, with f70 it was initially done separately, but we later started combining again — simply because it finds such clear and obvious setups that it feels unnatural not to add our own positions
By the way, there is also a separate account where we only copy AI logic from Telegram, just like any trader would: observe, think, and then place trades manually. That account is also publicly visible — and it performs very well.

So in summary:
the AI trades autonomously, and each mode has its own pure statistical record.
Only that isolated performance can be separated cleanly — and we publish it on our website.
Whether someone believes it or not is honestly not that important to us.
We simply document how the system behaves in reality.

We show that it does.
And sometimes… even autonomously, which is the most surprising part
What happens next?
We’ll see

Simple example from the USDJPY series (today)

How a trader usually operates
First comes visual pattern recognition (or automated screening) — basically finding an interesting market location.
For example, USDJPY might look interesting.
A trader might have a contextual bias (e.g. “USD is currently considered oversold” — based on market sentiment or macro discussions).
Then the trader studies the setup using Technical Analysis :

1. trend structure
2. support/resistance
3. patterns
4. confirmations

All of this is done in order to make a decision → whether to place a trade or not.
Each trader also builds their own performance statistics, tracking how often their “read” of the market is correct.
There are also studies suggesting that classical TA provides only a small edge over randomness — often around ~5%, sometimes 55–60% win-rate depending on method and conditions.

So in the end, the decision is:
trade or not trade
based on probability + personal historical performance

How we do it
In our case, evaluation, analysis, and pattern search are all handled by AI.
It operates 24/5 across markets.
What exactly it detects is part of a separate discussion
But the key point is:
everything described above (search, analysis, decision suggestion) is performed by the AI.
The difference is that it does not rely on classical TA tools or human-style interpretation of patterns, fractals, or indicators.
Instead, it uses our proprietary metric.

So we observe results from the “output side”, without manually deciding what to look for each time.

The system finds setups, publishes them to Telegram, and even executes trades automatically.
After that, we review performance and optionally add manual positioning if the setup looks meaningful.

The key point
The statistics of ORA AI are not 55%, 60%, or occasional 70%.
They are publicly tracked and shown:
93–95% win rate (historical tracked performance)
So with all respect to Technical Analysis — which has been studied extensively for decades — we base our view on ORA AI.
And from there, it becomes more of a game of refinement and interaction rather than basic prediction.
So far, everything remains in profit

A trader’s decision is usually based on a combination of factors — technical analysis, external opinions, “friend advice”, and various market beliefs operating in the background.
Most of these approaches also come with some form of performance tracking or a long-term estimate of “edge”. In theory, every trader has some understanding of their own statistics — at least approximately.

Right?

In our case, the decision is made by an AI system, backed by a tracked performance history of thousands of executed trades, showing a win rate above 92%.
This is not interpretation. It is recorded statistical behaviour at scale.

From our perspective, the conclusion is fairly simple:
ORA AI is simply more reliable for us personally than subjective advice or traditional TA-based decision-making.

Important update for everyone following the ORA project

So...
We have almost completed a new series of additional metrics designed to evaluate the quality of an expected trend.

Our main challenge is actually quite unusual:
we have VERY few bad lots.

And for confident statistical modelling, you normally need a meaningful separation between good and bad trades, right?
But since the number of genuinely bad lots is extremely small, we decided to approach the problem differently.

Instead of fighting bad trades, we started searching for the REALLY exceptional ones
The idea is that these specific lots may represent:

1. fast development,
2. strong continuation,
3. and potentially part of a larger multi-dimensional market structure.

In other words:
those rare situations where a sequence of trades develops into a larger trend over time, producing significantly stronger returns with relatively lower risk.

So here is what we built and how it will work.
Starting very soon, ORA signals will also include several indicators from the ORA INDy series (Y1...YN).
This means that alongside already highly accurate signals, there will now also be additional “sensors” highlighting potentially exceptional lots.
(See the second image.)

And to clarify an important point:
the red zone does NOT mean “bad trades”.
In fact, most trades will remain inside that area — this can already be seen from the upper chart.

But what enters the green zone...
Hypothetically, those may represent something very special:
very strong lots,
strong trend development,
and potentially high-quality continuation structures

So if a published signal also appears inside the green zone, it may indicate a potentially very strong market development scenario

That’s exactly what we’re going to test publicly — just as before:
free signal publication,
plus now the ORA INDy sensors as well.

Launch starts tomorrow

And one more thing regarding donations — even small ones like $5–10.

As we mentioned before:
when ORA eventually moves to a commercial launch, the subscription during the first months will remain free for everyone who supported the project with a donation.

Our reasoning is simple:
even a small contribution creates engagement, attention, and long-term interest

We are genuinely trying to build something truly useful in the huge ocean of market solutions out there.
And of course, we especially appreciate people who are genuinely interested in what we are doing

Huge thanks to everyone who has already donated!

We are trying to grow proportionally to the interest and support we receive

Since this night, the AI has been scanning 15 major pairs — and found almost nothing worth trading.
So no lots were placed.

For us, that is also a signal.

We don’t spend hours on TA, news, or manual market interpretation.

We work from the opposite direction:
if an AI system with a long-term tracked 93%+ win rate finds nothing interesting — the market is probably simply low-quality right now.
The time saved on analysis is enormous

Curious question for manual traders:
after spending hours analyzing today’s market… did you reach the same conclusion?


PS
The screenshot shows the current live state of the system we publicly display — what ORA is actively analyzing right now.
Link here
As you can see, only now has the AI started detecting interesting market locations again.
And once again, it seems to like USDJPY

Interesting case on EURJPY right now
The AI (new mode) identified a setup and effectively “states” that a certain level is likely to be broken by bullish momentum (see TG).

Just like any trader, the system detects something statistically interesting.

The difference is — we are not manually involved in the decision process.

Price keeps temporarily falling back into bearish pressure, even right now.

But the AI still maintains the same core expectation:
the level is likely to be taken by bulls.

And interestingly, the deeper price drops into resistance before the move, the more opportunity we personally see in positioning from better entries.

So we continue building the series — and repeatedly see profit appear across the structure.

At this point, the floating profit already exceeds the current open drawdown.

And for us, an open position is not the same thing as a realized loss.
If we closed the series now, it would already end in profit.

But many traders immediately focus only on drawdown statistics.

We look differently:
did the entire series produce profit or not?
Was the idea correct overall or not?
That’s simply the intuition we developed after thousands of high win-rate executions.
---
What is also interesting:
only now, looking at the chart afterwards, it becomes possible to notice a recognizable TA-style pattern that many traders may identify later.

That is exactly why we do not rely on TA itself.

Because even if a visible pattern weakens or “falls apart”, that does not automatically mean — at least for us personally — that the bullish level itself will fail as well.

That view comes from the statistics behind the system.
Without that statistical confidence, we probably would have panic-closed the position long ago

That is the key difference.
Different logic.
Different interpretation.
Built on statistical behaviour across thousands of trades.

Current expectation:
the level will likely be broken

Let’s see.At the moment, we are not closing the series — although it is worth noting that total profit has already been more than twice as high multiple times during the move

EURJPY

First of all — the projection was correct again.

The level was broken.

But the interesting part is not even the breakout itself.

It’s the level.

Price stalled and moved into consolidation exactly around that area.

Now the level is literally “drawn” on the chart by price action itself.
Earlier, this was not visually obvious.

But apparently the AI already saw it.

And yes, formally the level has already been taken, but based on previous behaviour, we still believe there is room for continuation.

Interestingly, while writing the previous post, a visible TA-style pattern was forming.

We mentioned that such patterns can weaken or completely fall apart — and that this is one of the weak points of traditional TA.

And that is exactly what started happening.

Many traders would immediately reduce exposure, close positions, or secure minimal profit.

But our logic is different.

For us, classical TA is still an extensively studied edge that statistically often remains relatively close to randomness over large samples.
We operate differently — through the AI framework and its historical statistical behaviour.

So when price dropped again:
for us, that did not automatically invalidate the bullish expectation.

Additional lots were added — and those also moved into profit.

Some will say this is risky.
But markets are always risky.

The real question is whether the long-term statistics justify the behaviour.
For us, that answer is reflected in the account history itself
We trust the AI here.
So far, it has earned that trust.

EURJPY — completed

The level was broken multiple times and now already with clear continuation beyond it.

Another successful forecast.

But the most interesting part is not even that.

When the prediction is first published, the usual reaction is:
“Impossible to know for sure.”

Then, once the move confirms:
“Now it’s obvious.”
“Probably just luck.”

Both reactions are usually based on personal trading experience and intuition.
But the most valuable part exists BETWEEN those two moments.
That is where uncertainty lives.

And with our AI, we step into that uncertainty with significantly more confidence

And yes, all lots are visible.

The statistics remain public, as always.
These are not cherry-picked “best trades only”.
If every day shows only the “best trades”...
then where exactly are the bad ones hiding?

Maybe tomorrow.

We’ll see.

New trading day restart is coming soon