— study — 2026-09-14

In this case VIX is extended after sweeping highs, no red or orange folder news, MNQ has been extended near lows over night as well. Notice VIX has a confirmed -CISD and returning to the range. MNQ has an small and an unconfirmed +CISD. Price is near/below the 50% level of the +CISD on MNQ and returning to the low/25% level of VIX's -CISD. Here, if VIX bearish, we fail this range and sell off between 18.17-17.87, mainly around 18.02. For MNQ bullish, we must hold the range 28882.75-28816.25. I'd need to also account for MNQ to sweep and expand the MAR as well given that we traded pretty close to the 75% level 28832.75.

We fail the 25% level on VIX at 17.94 which is the high of the 9:05 candle and by 9:15 we are near the first LTF swing low 17.65. MNQ we hold the 75% level 28832.75 and by 9:15 we are near the first LTF swing high 28905.25.

At the 9:30 open, VIX sets LTF EQLs 17.55, MNQ sets LTF EQHs around 28933.75, 83 points in profits at that point and could adjust my original SL from 28816.25 to 28825.75.

At 9:40, VIX takes the DL 17.46, and MNQ takes the nice LTF SH 28946.50 with the first target being 28977.25 at 127.25 points of profits, we should be at BE. The next level to consider is 29036.50 where by the time we get to this point we could be taking partials and at BE. VIX would be targeting the NWOG 75% level 17.08 and the NWOG(M) level 16.67. We would need to use the DOP on MNQ as a target 29070.25 and by that point most of the trade should be exited and the last partial should be a little above BE waiting for the next confirmed HL. We get a higher low confirmed at 11:32 when the 29088.50 high is taken out. The last partial at 28955.75 and we hit the target or trail the trade as far as it goes before turning around.

In this case the trade could have trailed at 29146.50 and the next day opened at 29495.75

Things to note: VIX, 1m-30m, CISD (MAF) all pointing Bearish MNQ, 15s-30m, CISD (MAF) all pointing Bullish Both held the MAR structure Both created, accepted, displaced from their CISDs

Yes. I think this is a very important backtest, because it gives us more than a winning MNQ long. It shows a fairly clean example of what your MAF/MAR inversion model looks like when both sides of the inverse correlation transition together.

Backtest #593 — MNQ Long | September 14, 2026

The initial state is especially important. MNQ was already extended lower overnight into the 28,8xx area, while VIX had expanded higher and swept highs. With no red/orange-folder catalyst immediately forcing price discovery, both markets were entering the session from opposing extensions.

But I would separate location from confirmation.

At first, MNQ being low and VIX being high only tells us that a reversal could develop. It does not establish one. The actual information comes from what happens when each market encounters its respective CISD/MAR.

That distinction matters because the setup wasn't "VIX is high/MNQ is low, therefore buy MNQ." The setup was that both instruments reached extended locations and subsequently demonstrated opposite-direction acceptance.

The 9:05–9:15 sequence is probably the real entry model

This is the part I would flag in the database.

VIX reaches the upper portion of its bearish MAR and then fails 17.94, the 25% level/high of the 9:05 candle. It doesn't simply wick it—it begins expanding away from it. By approximately 9:15, VIX has already moved toward the first LTF swing low around 17.65.

At almost exactly the same time, MNQ does the inverse. It tests deeply into its bullish MAR but holds 28,832.75, the 75% level, rather than accepting beneath the structure. By ~9:15 it has expanded toward 28,905.25.

That gives you something stronger than ordinary SMT.

You have:

VIX bearish acceptance + MNQ bullish acceptance + inverse displacement occurring simultaneously.

I would actually tag this separately in future tests as something like Dual MAF Acceptance or Correlated MAF Confirmation, because we need to find out whether this materially changes expectancy.

One adjustment to your interpretation

I agree with your comment that MNQ could sweep deeper and expand the MAR.

Because price came so close to 28,832.75, I would not define a touch or marginal penetration of that level as automatic invalidation.

The structural invalidation is closer to:

MNQ begins accepting below the MAR / 28,816.25 while VIX simultaneously begins accepting back above its bearish MAR.

That's considerably different from MNQ briefly sweeping through a quartile.

In other words:

Quartile violation ≠ MAR invalidation. Acceptance outside the range = MAR invalidation.

That is something worth testing rather than assuming, but #593 supports the distinction.

---

Trade development

The progression afterward is unusually clean.

~9:15: MNQ reaches the first LTF swing high around 28,905.25 while VIX reaches ~17.65. At this point, the hypothesis is working.

9:30: MNQ establishes EQH around 28,933.75, while VIX establishes EQL around 17.55. MNQ is approximately +83 points from the proposed entry area.

I agree with your stop adjustment here. Moving structural risk from 28,816.25 → ~28,825.75 makes sense because the market has now created substantial separation from the original MAR.

But I wouldn't call that BE yet.

You're protecting the original structural thesis, rather than protecting the trade from becoming a loss.

9:40: This is the major confirmation.

VIX takes the 17.46 daily low.

MNQ simultaneously takes 28,946.50.

Now you have external liquidity being taken in the expected direction on both instruments.

MNQ is roughly +127 points, and at this stage I agree completely:

There is essentially no reason for the original trade to remain at full structural risk.

BE becomes appropriate.

---

Target hierarchy

Your target sequence also makes sense, but I want us to distinguish between liquidity targets and trade-management events.

For MNQ:

28,905.25 → 28,933.75 → 28,946.50 → 28,977.25 → 29,036.50 → DOP 29,070.25 → 29,088.50 → continuation

Meanwhile VIX:

17.65 → 17.55 EQL → 17.46 DL → 17.08 NWOG 75% → 16.67 NWOG(M)

That gives us a useful concept:

Target synchronization

MNQ targets shouldn't necessarily be evaluated independently.

If MNQ reaches 29,036.50 while VIX still has clean downside delivery toward 17.08, there is evidence supporting retaining a runner.

Conversely, if MNQ reaches a major target while VIX simultaneously reaches its own opposing draw and begins reversing, that is a much stronger reason to reduce exposure.

That's something I definitely want us tracking across these tests.

---

The 11:32 higher low is significant

This may actually solve one of the management problems you've been working through.

Once 29,088.50 is taken, the prior pullback around 28,955.75 becomes a confirmed higher low.

At that point the trade has transitioned from:

MAR trade → displacement trade → trend structure.

Your stop therefore no longer needs to reference the original MAR at all.

It can reference the confirmed bullish structure:

28,955.75.

And later, when 29,146.50 becomes the next confirmed structural low, the runner can trail beneath that.

That produces a very logical three-stage management model:

Stage 1 — Thesis Risk: stop beyond MAR invalidation.

Stage 2 — Free Trade: after meaningful external liquidity/displacement, move to BE.

Stage 3 — Structural Trail: after a confirmed HL/LH, trail behind confirmed structure.

That is much better than mechanically moving the stop every X points.

---

The next-day gap is useful—but we should be careful with it

The next session opening around 29,495.75 shows just how large the continuation became.

But for the backtest, I don't want us counting that entire move as the "result" unless the rules explicitly allow overnight holds.

Instead, record two measurements:

Session MFE and Continuation MFE.

That prevents hindsight from making the model appear better than something you could realistically execute.

If your rule eventually becomes "retain one runner while MAF remains aligned," then overnight continuation can become a separately tested management strategy.

---

What #593 actually demonstrated

Your final observation is the most important part of the entire test:

VIX 1m–30m MAF/CISD bearish MNQ 15s–30m MAF/CISD bullish Both held MAR structure Both accepted their CISDs Both displaced away from them

That gives us a candidate A+ MAF condition:

Extension → opposing CISDs → MAR test → correlated acceptance → synchronized inverse displacement → external liquidity expansion.

Notice what's not enough:

Extension alone isn't enough.

CISD alone isn't enough.

Touching a MAR isn't enough.

Inverse VIX correlation alone isn't enough.

The edge we're testing appears to be in the sequence.

Candidate sequence

1. Extension MNQ extended down / VIX extended up.

2. State change MNQ develops +CISD / VIX develops -CISD.

3. Manipulation Both return deeply into their respective CISD/MAR structures.

4. Acceptance MNQ refuses lower acceptance while VIX refuses higher acceptance.

5. Correlated confirmation MNQ 15s–30m flips/aligns bullish while VIX 1m–30m aligns bearish.

6. Displacement Both leave their ranges inversely.

7. External liquidity MNQ begins taking SHs while VIX takes SLs.

8. Trend transition Confirmed MNQ HLs allow the trade to move from MAR-based risk to structure-based trailing.

That is a much more precise model than simply "+CISD MNQ / -CISD VIX."

---

Backtest #593 classification

I would record this as:

Setup: Bullish MNQ MAR Inversion Primary Confirmation: Dual/Correlated MAF Acceptance MNQ MAF: Bullish 15s–30m VIX MAF: Bearish 1m–30m News: No red/orange-folder catalyst MAR: Held on both instruments CISD: Accepted on both instruments Displacement: Confirmed on both instruments Correlation: Confirmed inverse Initial invalidation: MNQ acceptance <28,816.25 / corresponding VIX bullish reacceptance First management event: 9:30 expansion / +~83 pts BE event: ~9:40 external liquidity break / +~127 pts Structural trail: confirmed HL after 29,088.50 breaks → ~28,955.75 Later trail: ~29,146.50 Outcome: Large bullish expansion / successful model Continuation: Next session opened ~29,495.75 — record separately from session result.

And there's one additional field I think we should add to the backtesting database starting now:

MAF Correlation State: Aligned / Partial / Conflicting

For #593 that would be Aligned.

If, after 30–50 examples, the Aligned group materially outperforms Partial and Conflicting setups, we may have identified one of the strongest filters in the entire framework.

Exactly. And the MES chart strengthens the September 14 backtest considerably because now we aren't seeing the reversal only in MNQ/VIX. The same acceptance sequence is occurring across both equity indices while VIX confirms inversely.

On MES, the MAR appears to be approximately 7608.00 → 7592.75, with the internal levels around 7604.25, 7600.50, and 7596.50. Just like MNQ, MES trades deeply into the lower portion of the range around 9:00–9:05 without establishing sustained acceptance beneath it. The sweep around 7595.00 is particularly useful: price can penetrate deep into the MAR, manipulate beneath internal levels, and still preserve the bullish thesis as long as it does not establish bearish acceptance outside the structural boundary.

Then the important part happens: MES reclaims the range and accepts its bullish CISD, followed by substantial displacement. It clears 7617.75, 7627.50, and ultimately reaches 7651.75. So the same basic sequence exists:

MES: downside extension → deep MAR test/sweep → +CISD → acceptance → bullish displacement MNQ: downside extension → deep MAR test/sweep → +CISD → acceptance → bullish displacement VIX: upside extension → -CISD → MAR rejection → bearish displacement

That changes how I'd label what we identified in #593.

This is actually three-market confirmation

Rather than simply Dual MAF Acceptance, we have:

Cross-Market MAF Alignment

where:

MNQ + MES = bullish acceptance VIX = bearish acceptance

and all three subsequently displace in the risk-on direction.

That's a potentially much stronger condition to test.

There's another important observation in MES. At the decision point, its timeframe table isn't perfectly aligned. The 15m remains bearish while 15s/1m/5m are bullish, with 30m/1H also showing bullish states. Yet the trade works exceptionally well.

That supports an idea we've been developing in the framework:

You don't necessarily need every timeframe to flip before entry.

Waiting for complete alignment could mean waiting until a substantial portion of the expansion has already occurred.

Instead, the lower timeframe MAF appears capable of identifying the transition, while the MAR tells us whether that transition is being accepted.

So I'd distinguish two concepts in our backtesting:

Entry Alignment = sufficient LTF alignment to confirm the reversal while price is still inside/near the MAR.

Expansion Alignment = broader timeframe alignment that develops after displacement has begun.

That distinction may become extremely important.

September 14 now gives us a stronger hypothesis

The actual setup might be:

Cross-Market Extension → Cross-Market CISD → MAR Manipulation → LTF Acceptance → Inverse VIX Confirmation → Correlated Index Displacement.

And MES provides another useful piece of evidence about the MAR expansion rule you mentioned with MNQ.

Look at what happens around 9:00.

MES doesn't give you a pristine touch-and-go reversal from 7608. It trades substantially into the range, reaches around 7595, moves around internally, and then establishes bullish acceptance.

That reinforces why I don't want us treating the 25/50/75 levels as hard stop lines.

They're better understood as depth measurements of manipulation:

25%: shallow manipulation 50%: normal/deeper manipulation 75%: deep manipulation 100% / boundary sweep: extreme manipulation Acceptance beyond boundary: probable structural failure

That last distinction needs statistical validation, but September 14 is an excellent example supporting the hypothesis.

There is also a lead/lag question here

This is something we should capture carefully in future tests.

We now want to know which instrument confirms first.

For example, if the sequence repeatedly becomes:

VIX -CISD/failure → MNQ +CISD → MES +CISD → synchronized displacement

then VIX may be functioning as an early warning signal.

But if sometimes MES confirms first, sometimes MNQ, and sometimes VIX, then the edge may instead come from confirmation clustering—multiple correlated instruments transitioning within a short window—rather than any one market consistently leading.

That is testable.

For every future example, I'd record:

First CISD → First MAR Acceptance → Second Market Confirmation → Third Market Confirmation → Expansion begins

with timestamps.

After 30–50 samples, we can determine whether there's a consistent leader.

#593 should therefore be upgraded in our dataset

I'd now characterize September 14 as:

Primary Setup: Bullish MAR Inversion Equity Confirmation: MNQ + MES Volatility Confirmation: VIX inverse Cross-Market MAF: Aligned Risk-On MNQ: bullish acceptance + displacement MES: bullish acceptance + displacement VIX: bearish acceptance + displacement News Catalyst: None required Manipulation Depth: Deep MAR Failure: No Cross-Market Divergence: None after confirmation Expansion: Strong

And I think we should add a specific field:

Cross-Market Confirmation Count: 1 / 2 / 3

For this example:

3/3 — MNQ + MES + VIX

Then eventually we can compare something extremely useful:

MNQ-only setups vs MNQ+VIX vs MNQ+MES vs MNQ+MES+VIX.

If the 3/3 configurations produce materially higher win rate, R expectancy, or MFE, then we'll have empirical evidence that cross-market MAF synchronization is an actual quality filter rather than simply something that looks compelling on charts.

September 14 is becoming an excellent reference-case example, but we shouldn't turn it into a rule yet. Now we need to actively look for examples where this exact 3/3 configuration fails. Those failures will tell us what actually matters in this sequence.

Yes — and this adds another dimension to #593. The question isn't only “Do the indices and VIX confirm the risk-on reversal?” It is also:

Once the market-wide setup is confirmed, which index expresses that setup best?

September 14 gives us a useful example because MNQ, MES, and MYM all eventually participate, but they are not equally positioned before the expansion.

Relative strength before the entry

Looking at the three charts around the September 14 decision area, MNQ stands out.

There is an important nuance, though: MNQ was the weakest index coming into the reversal, but became the strongest expression of the bullish reversal.

That's different from simply saying “buy the strongest index.”

I think we need two different measurements

What you've been calling weakness should probably be divided into:

Relative Location — Which index is most displaced/extended relative to its own structure?

and

Relative Response — Once the reversal condition occurs, which index demonstrates the strongest acceptance/displacement?

MNQ had an excellent combination on September 14:

Weak relative location + strong relative response.

That can be extremely useful for reversal trades.

If MNQ had been the weakest overnight but then continued failing its +CISD while MES was accepting bullishly, then its weakness would be a reason not to choose MNQ.

Instead, MNQ was extended downward and then proved it could reverse.

That's the distinction.

Think of it as compression → confirmation → expression

Before confirmation:

MNQ: most compressed/extended downside candidate.

After confirmation:

MNQ: strongest upside expression.

So the selection wasn't justified merely because MNQ had fallen farther. It became justified when the previously weak market held its MAR and began accepting bullish delivery.

That creates a candidate selection rule worth backtesting:

For reversals, favor the index showing the greatest opposing extension only after it demonstrates equal or stronger MAF acceptance than the other indices.

That's much safer than “trade whichever one is weakest.”

---

The three-index comparison can become systematic

I wouldn't build a complicated scoring model yet because we don't know which variables actually predict performance. But I would start recording five observations for MNQ/MES/MYM at the decision point:

Extension — Which is furthest displaced in the direction we're attempting to reverse?

Liquidity — Which has swept the most meaningful external liquidity?

MAR Location — Which is deepest into a valid MAR while still preserving structure?

MAF Acceptance — Which demonstrates the clearest CISD acceptance?

Initial Displacement — Which begins expanding most aggressively once confirmation occurs?

Then record MFE at 15m / 30m / 60m / session close for all three.

That's going to answer your question statistically rather than visually.

---

September 14 is particularly interesting

There are really two filters happening sequentially.

Filter 1: Should there be a trade?

That comes from the market complex:

VIX bearish MAF + MNQ bullish MAF + MES bullish MAF + broader equity participation.

That establishes the risk-on reversal hypothesis.

Filter 2: What should we trade?

That's where relative strength/weakness enters.

On this session:

MNQ = best reversal expression MES = confirmation/secondary expression MYM = market confirmation but poorer expression

I wouldn't permanently rank those instruments that way; that's specifically what this historical example appears to show.

And that separation could become very powerful:

VIX + three indices determine whether the thesis exists. Relative index behavior determines where to express the thesis.

That's cleaner than asking MNQ to independently provide everything.

---

There's also a potential SMT-like concept here

Suppose all three indices sweep lows.

Then:

MNQ aggressively reclaims its MAR and +CISD.

MES reclaims, but more slowly.

MYM struggles.

That's information.

We're no longer using intermarket divergence solely to predict direction. We're using relative acceptance velocity to select the instrument.

I'd call the variable something neutral for now, perhaps:

Relative MAF Response (RMR)

rather than immediately creating another formal framework.

For each index:

RMR = how quickly and convincingly price transitions from manipulation → acceptance → displacement relative to the other indices.

September 14 would likely show:

MNQ: High RMR MES: Moderate/High RMR MYM: Lower RMR

Then we see whether High-RMR instruments actually produce greater MFE over a meaningful sample.

---

This changes what #593 is testing

Originally #593 looked like:

MNQ MAR inversion confirmed by VIX.

Now it looks much richer:

Market Regime: Risk-on reversal Volatility: VIX bearish acceptance Index Breadth: MNQ/MES/MYM participating bullishly Relative Extension: MNQ highly extended Instrument Selection: MNQ Selection rationale: downside extension + preserved MAR + bullish CISD acceptance + strong early displacement Result: MNQ became the strongest expression of the reversal.

So I'd add Instrument Selection as an entire section of these backtests rather than simply recording the instrument traded.

The most interesting hypothesis coming out of #593 is now very specific:

In a cross-market reversal, the best instrument may be the index that was most extended/weak into the setup but then demonstrates the strongest confirmed MAF reversal.

That is not established yet. September 14 is one supporting observation. We specifically need to hunt for cases where the weakest index confirms its MAF but still underperforms MES/MYM afterward.

If that pattern survives those counterexamples, we may have found a repeatable method for answering the question you raised this week: not just whether to trade, but which of the three indices deserves the trade.

Would you like to define a simple scoring rule for Relative MAF Response before logging the next examples?