Kong

Sieving for ℙrimes | Quantitative Intelligence.
Not financial advice
ALPHA stage | building in public, bugs expected

[Kong @kongquant]

One extra BAC note: the pair fitness block grades the asset D with high confidence, 181 total flips, and only 1.7 flips per year. That low cadence is why the 1D turn is more meaningful when it arrives with 4H at 75 bars and 5D at 70 bars.

[Kong @kongquant]

Bank of America flipped long on Kong Cloud with $BAC at Kong Score 78. The useful part is not the direction label. It is how a continuation prime gets through the scoring stack when the daily cloud turns after the higher windows were already aligned.

BAC is a clean worked example of the trending path in Kong Cloud Process v2.2.0.

The daily cloud moved from neutral to long on 2026-06-04 at 54.11. The live snapshot shortly after showed price at 54.17, the 1D cloud one bar into the new state, the 4H cloud already long for 75 bars, and the 5D cloud long for 70 bars. That matters because a daily flip inside an already aligned 4H and 5D structure is treated as continuation, not as a fresh regime claim.

Continuation does not receive a perfect structural mark. In the trending path, continuation starts at 0.7 rather than 1.0 because the system is not rewarding novelty. It is asking whether a pullback or neutral pause has ended while the larger cloud remains intact. That nuance is why BAC can receive a high score without pretending the daily turn is the whole story.

The largest contributor today is the pattern layer. In a trending regime, Kong assigns 40 weight to pattern. BAC printed an inverse head and shoulders with 87.5 confidence. The pattern data marks the head at 46.12 on 2026-03-19, shoulders at 51.31 and 51.83, and a neckline at 55.76 with the right neckline reference at 54.65. That is not a decorative label. It is the main evidence block in this prime.

Bulkowski classifies inverse head and shoulders inside the Tier A reliability group in the Kong process guide, with the Tier A bucket carrying a 3-19% failure range. Edwards and Magee provide the structural grammar: head, shoulders, neckline, and confirmation behavior. Murphy gives the wider technical-analysis frame. Wilder enters through smoothing, where Kong Cloud uses smoothed moving averages to reduce single-bar noise before a flip is accepted.

The process does not let the pattern layer operate alone. Volume came in at 0.9x, which sits in the normal band. That is informative because volume did not carry the prime. It neither added the strong-volume confirmation of a breakout-style read nor rejected the pattern outright. The score had to come from pattern primacy, continuation structure, and macro confirmation.

The component scores show that arithmetic clearly: pattern contributed 36.0 from a 40 weight, macro alignment contributed 14.0 from 20, structural context contributed 12.75 from 15, volume contributed 10.0 from 20, and regime contributed 5.0 from 5. The total lands at Kong Score 78, the Convicted tier for this prime.

There is also a useful market-context detail. The scan covered 642 pairs. BAC was one of the 1D long flips, while the wider scan showed 90 long-aligned assets and 339 short-aligned assets. In other words, BAC is not being described as a broad-market proxy. It is a single stock where the internal stack was cleaner than the aggregate tape.

That is the point of the methodology angle today: pattern primacy inside continuation. If the 4H and 5D cloud had not already been aligned, the same daily flip would have been a different research problem. If the pattern confidence had been weak, the 0.9x volume read would have carried more concern. With BAC, the model is saying the daily turn matters because it sits on top of a mature higher-window cloud and a high-confidence classical structure.

The open observation is simple: the next read is not whether BAC has a good story. It is whether price behavior around the neckline zone and the daily cloud can keep validating the pattern-led continuation thesis.

[Kong @kongquant]

One extra context point: this asset is not a frequent Cloud mover. The fitness block shows 35 total flips and about 1.6 flips per year, with high confidence but a C grade. That makes the multi-window state more relevant: 1D just flipped, while 4H has been short for 29 bars and 5D for 60 bars.

[Kong @kongquant]

$KSA is today's Kong Prime: a 1D Kong Cloud flip short at 38.23, score 69, inside a 98 bar range. The clean lesson is pattern primacy inside a bounded market, where structure, pattern quality, and higher-window alignment have to agree before the read earns attention.

Kong Cloud starts by asking what kind of market it is reading. KSA is not classified as a clean directional tape today. The prep bundle marks a ranging regime, moderate quality, between 35.90 and 39.18. That range has 3 support touches and 3 resistance touches across 98 bars. In the ranging path, the process gives structure the heaviest role because the chart is being judged by its boundaries. A rejection at resistance carries a different meaning than a random close in the middle of the band.

That is why today's structural context matters. The prime was logged as a rejection at 38.24, with price at prime 38.23. The live close in the bundle is 38.22. The structural invalidation level is 38.44543, and the system recorded a structure invalidation print at 38.565 during the same daily bar. That sounds messy, and it should. A range read is supposed to be sensitive to boundary behavior. Kong does not turn that into drama; it scores the state of the evidence.

The pattern layer is what makes KSA teachable. The detector tagged a head and shoulders with 85.5 confidence. The anatomy is specific: left shoulder 38.765 on March 10, head 40.83 on April 17, right shoulder 39.49 on May 1, with a neckline around 38.185. Edwards and Magee treated this formation as a structural pattern, not a decorative shape. Bulkowski's work is useful here because it pushes the same point from the statistical side: pattern class and confirmation quality matter more than the label alone. Kong reflects that by giving the pattern component a 25 weight in this ranging read, and KSA contributed 22.5 of that.

Volume keeps the read grounded. The bundle shows volume ratio 0.9, classified as normal, with an increasing trend. Kong's volume ladder puts 0.8 to 1.5x in the normal zone, so this is not a volume shock. It does not carry the read by itself. It says the pattern and structural rejection are not being contradicted by a weak print, while still refusing to overstate participation.

The higher-window check adds another layer. The 1D Cloud flipped short on June 3 at 13:30 UTC. The 4H Cloud is also short, 29 bars into its state, and the 5D Cloud is short, 60 bars into its state. The prep bundle marks alignment as FULL_SHORT and macro alignment as confirming. In the Kong score weights, macro alignment carries 15 and KSA contributed 10.5. That is enough to matter, not enough to override the rest of the stack.

This is the core methodology angle: a high-confidence classical pattern can lift a bounded-market read only when the structural location is coherent. The chart had a defined range, a resistance rejection near 38.24, a head and shoulders with 85.5 confidence, normal increasing volume at 0.9x, and higher-window confirmation. Put together, that produced a Kong Score of 69 and a Convicted tier.

The open question: does KSA respect the neckline area near 38.18 after the 1D flip, or does the range absorb the signal and force a fresh read?

[Kong @kongquant]

One extra context layer: the pair detail rated this market B with medium confidence and a strongly trending regime affinity. Across the recorded sample it shows 26 total cloud flips, about 3.6 per year. That low cadence is why the 1D flip matters more after the 4H and 5D states were already aligned.

[Kong @kongquant]

$API3 is today's Kong Prime: 1D flipped SHORT at 0.2854, Kong Score 62, full 1D/4H/5D alignment, and a 75.7 confidence head and shoulders pattern.

The useful lesson is pattern primacy inside a trending regime.

Kong Cloud does not treat every flip the same way. A flip inside a range is judged through structure first, because contained markets live around support, resistance, and VPVR. A flip inside a trending regime is judged through the trending path. Continuation starts with a baseline of 0.7 in that path, not 1.0, because confirming an existing directional state is valuable but less rare than a clean regime transition.

API3 is a clean example. The scan marked the regime as trending, with the prior run already 81 bars long. The new 1D flip moved SHORT after the 4H cloud had already been SHORT for 110 bars and the 5D cloud had been SHORT for 130 bars. That gives the reading full timeframe alignment. The 1D did not act alone; it joined the larger cloud stack.

The scoring stack shows why the final number is 62 instead of something louder. In this prime shape, pattern carries a 40 weight, volume carries 20, macro alignment carries 20, structural context carries 15, and regime contributes 5. API3's component scores were 30.6 for pattern, 12.0 for volume, 2.0 for macro alignment, 12.75 for structural context, and 5.0 for regime. That is a concentrated read. The pattern did most of the work.

The pattern itself is not a decorative label. Kong detected a head and shoulders formation with 75.7 confidence. The left shoulder was near 0.3361 on March 12, the head was near 0.5085 on April 25, and the right shoulder was near 0.3499 on May 25. The neckline sits near 0.2882, with neckline points around 0.2643 and 0.312. Price at prime was 0.2854, right around the active neckline area.

That is why classical structure matters here. Edwards and Magee describe the head and shoulders pattern as a distribution structure built around failed recovery attempts. Murphy treats neckline behavior as the practical confirmation zone. Bulkowski's pattern work puts head and shoulders in the Tier A group, with failure rates in the 3-19% band depending on variant and context. Kong does not quote those authors as theater. It uses the idea that some patterns deserve more weight than others when the market regime already has direction.

Volume did not lead this prime. The volume ratio was 0.95x, classified as normal, with volume trend increasing. Kong's volume tiers are explicit: above 2.0x is strong, 1.5-2.0x is confirmed, 0.8-1.5x is normal, and below 0.8x is weak. API3 sits in the normal band. That is enough not to veto the read, but it is not enough to carry it.

Momentum was also mixed. The fast momentum panel sat below its signal line and the histogram reading was negative at -0.003784. Those details describe internal pressure, but they are not the center of the score. Kong Cloud defines cloud state first; secondary momentum tools sit behind the structural read rather than replacing it.

The invalidation area is also plain. The structural invalidation level is 0.300563, about 5.31% away from the prime price. Nearby resistance levels were 0.2899, 0.2916, and 0.302. Nearby support levels were 0.280035, 0.2762, and 0.2739. In other words, the read is not abstract. It is pinned between the neckline zone, the first resistance cluster, and the defined invalidation level.

This is what a Developing tier score often looks like in practice. Not weak, not maximal, and not dependent on hype. It is a stacked continuation read where the pattern supplies the main evidence, volume stays normal, macro confirms at a low asset-class weight, and the cloud stack aligns across daily, 4H, and 5D timeframes.

The open question: is API3 less about a fresh surprise and more about whether a classical neckline can keep control after the daily cloud finally joins the wider SHORT state?

[Kong @kongquant]

Additional context: Kong fitness grades this asset A with medium confidence, strongly trending regime affinity, 23 total flips, and 2.2 flips per year. Low flip cadence matters here because the 1D change was not a frequent churn event. It arrived after the 4H and 5D windows were already aligned.

[Kong @kongquant]

Kong Cloud flipped short on $KSM with Kong Score 65. The headline is not the direction by itself. It is the way a 78.5 confidence head and shoulders read overrode a messy daily transition and turned the 1D cloud into alignment with the 4H and 5D windows.

This is a clean example of the continuation path inside Kong Cloud Process v2.2.0.

When a market is already in a trending regime, the process does not give a fresh flip full credit just because the cloud changed state. Continuation starts below a perfect structural score because the system is asking a narrower question: has the pullback or neutral pause ended, and has the prior trend resumed with enough confirmation to matter?

KSM is useful because the answer is visible across the stack. The 1D cloud flipped short at 4.53 after sitting neutral. The 4H cloud was already short for 101 bars. The 5D cloud was already short for 93 bars. That creates full short alignment across the local, intermediate, and wider windows.

The part Kong cares about is not the label. It is the evidence behind the label.

In trending markets, pattern carries a 40 weight in the process guide. That is the largest single component for this path. The reason is classical, not decorative. Edwards and Magee used head and shoulders as a structural map of failed recovery attempts around a neckline. Bulkowski gives the pattern statistics that let the model treat some formations differently from generic chart shape. Murphy gives the broader technical framework. Kong Cloud turns those older chart concepts into scored inputs instead of narrative preference.

Today, KSM printed a head and shoulders pattern with 78.5 confidence. The neckline in the source data sits around 4.72, with the head at 7.70 and shoulders at 5.18 and 5.23. Price at prime was 4.53. That places the daily cloud flip below the neckline region rather than inside the prior recovery structure.

Volume is the second check. The volume ratio came in at 1.5x, which sits on the confirmed threshold in the process guide. This matters because continuation without volume can be a thin cloud change. Continuation with confirmed volume says participation was present when the daily state changed.

The final score reflects that mix. Kong Score 65 is a Developing tier read: enough component agreement to surface the asset, but not a maximal verdict. Pattern contributed the largest share. Volume confirmed. The wider 5D window confirmed. Momentum was mixed, so the score did not pretend every layer agreed.

This is why the post is about process, not prediction. A short flip, a 78.5 pattern confidence, 1.5x volume, 4H short for 101 bars, and 5D short for 93 bars all point to the same analytical question.

Was the neutral daily pause just noise inside the wider downtrend, or has KSM started a larger state change that the lower windows already saw first?

[Kong @kongquant]

One extra context point: VANA is a low-cadence cloud asset in this sample. Pair detail shows only 3 total flips, about 2.4 per year, with 1D moving from neutral to short while 4H and 5D were already short. That makes the timing of the 1D flip more meaningful than a high-churn cloud transition.

[Kong @kongquant]

$VANA flipped short as a Kong Cloud continuation read, but the useful lesson is not the direction. It is why a 100 confidence head and shoulders can dominate the score while weak volume still keeps the verdict at Kong Score 59.

This is a pattern-primacy example inside a trending regime.

Kong Cloud starts with regime because every later component has a different meaning depending on the state of the market. VANA is classified as trending, with a continuation structural context. The 1D cloud has just flipped short at 1.393 after a neutral state. The 4H cloud was already short for 97 bars, and the 5D cloud was short for 76 bars. The snapshot calls that full short alignment.

That alignment matters, but it does not settle the read on its own. In the Kong process, a continuation flip in a trend starts as evidence that the pullback has ended and the prior direction has resumed. It is not given a perfect score simply because it agrees with the wider cloud. Continuation is meaningful, but regime confirmation is less rare than regime transition, so the rest of the stack has to earn the verdict.

The decisive layer today is the pattern layer.

VANA printed a head and shoulders with 100 confidence. The detected structure is specific: left shoulder 1.658 on April 28, head 1.821 on May 11, right shoulder 1.619 on May 22, and neckline 1.397. The prime price is 1.393, almost exactly at that neckline reference. Edwards and Magee gave the head and shoulders its classical structural definition: a failed attempt to extend the prior swing, followed by a neckline break that turns the pattern from shape into evidence. Murphy treats the same formation as one of the central reversal and continuation diagnostics because it links price structure to crowd exhaustion.

Kong does not score that visually. It scores the pattern as a component. In trending regimes, pattern carries a 40 point weight. VANA receives 36 from that bucket. That is why this prime can reach 59 even though volume is not helping. The pattern is doing the work.

Bulkowski is relevant here because Kong's pattern tiers are not aesthetic labels. Head and shoulders belongs in the Tier A reliability class in the process guide, with Tier A mapped to 3-19% failure rate territory across the cited pattern literature. The exact detected pattern still has to pass confidence checks. VANA does: 100 confidence, 100 volume score inside the pattern object, and a flat neckline at 1.397.

The tension is volume.

Volume ratio is 0.26. Kong labels volume weak and declining, and the volume component contributes only 3 of 20. That matters because a neckline event with thin participation has less confirmation than the same event with expanding participation. The system does not discard the pattern, but it also refuses to let the pattern become a complete verdict. That is the difference between a clean structural read and a high-conviction state.

The other component scores show the same discipline. Regime contributes 5 of 5, structure contributes 12.75 of 15, pattern contributes 36 of 40, macro alignment contributes 2 of 20, and volume contributes 3 of 20. The headline looks simple: VANA flipped short. The actual read is more nuanced: structure and pattern are clear, volume and macro keep the score in check.

Wilder's smoothing lineage matters in the background. Kong Cloud is built to reduce single-bar noise, and the pair detail shows why. VANA has only 3 total flips in the fitness sample, about 2.4 flips per year, with an A grade and strongly trending regime affinity. The 1D signal did not appear in isolation. It arrived after the 4H had already spent 97 bars in the same direction and the 5D had spent 76 bars there.

That is why this prime is a useful worked example. The surface event is a fresh 1D flip. The methodology read is pattern primacy under continuation conditions, with volume acting as the governor. The open question is whether participation starts to confirm the neckline, or whether the pattern remains technically clean but lightly sponsored.

[Kong @kongquant]

One extra context note: ENS has low Cloud cadence by design. The fitness block shows 17 total flips, about 2.1 flips per year, with medium confidence. When a low-cadence asset flips after the 4H window has held short for 85 bars, Kong treats the timing as filtered rather than noisy.

[Kong @kongquant]

$ENS is today's Kong Prime: a fresh short flip, Kong Score 70, with a 76.1 confidence head and shoulders pattern sitting right on the methodology fault line.

This one is a clean worked example of pattern primacy inside a trending regime.

Kong Cloud does not start with a chart pattern and then hunt for confirmation. It starts with regime. ENS is classified as trending, with the 1D cloud flipping short at 5.97, the 4H cloud already short for 85 bars, and the 5D cloud short for 45 bars. The prior 1D state was neutral, so this is a fresh daily transition into the direction already held by the wider windows.

That matters because the scoring path changes by regime.

In a range, structure does most of the work. Support, resistance, range quality, VPVR, and breakout behavior carry the main load because the market is contained. In a trend, Kong moves to the Step 2b path. A continuation flip is read as the pullback ending and the trend resuming, but it does not get a full 1.0 structural factor by default. Continuation starts at 0.7 because a regime change is rarer than confirmation of an existing path.

That is the first useful detail in ENS. The structural context contributes 12.75 out of a possible 15. Helpful, but not the whole story.

The second detail is the pattern bucket. In trending regimes, pattern has the largest weight in this Prime model at 40. ENS printed a head and shoulders pattern with 76.1 confidence. The pattern data is specific: left shoulder 6.49 on April 24, head 8.16 on May 11, right shoulder 6.59 on May 22, with the neckline around 5.93. The prime price was 5.97, so the daily flip is occurring almost on top of the neckline zone.

That is why the Bulkowski, Edwards and Magee lineage matters here. Edwards and Magee give the classical structural definition: shoulders, head, neckline, then the test of that neckline. Bulkowski's work is useful because it moves the discussion from pattern naming to pattern reliability. Kong does the same thing in code. It does not reward every familiar silhouette equally. It ranks pattern class and then scales the contribution by confidence.

For ENS, the pattern component contributes 30.6 out of 40. That is the largest single component in the score.

Volume is the third piece. Kong's volume layer uses a rolling baseline, with greater than 2.0x classified as strong. ENS printed 2.17x volume, and the volume trend is increasing. That gives the volume bucket its full 20 out of 20. In a continuation read, volume is not decoration. It helps separate a quiet cloud flip from a flip with participation behind it.

Macro is more subtle. The 5D cloud confirms the daily direction, but ENS is crypto, and the process intentionally scales macro contribution by asset class. Large cross-asset macro carries more weight for indices, commodities and forex than for smaller crypto assets. Here the macro bucket contributes 2 out of 20. Confirmation exists, but the model refuses to over-credit it.

That leaves the full component map: regime 5, volume 20, pattern 30.6, macro alignment 2, structural context 12.75. Total: 70. Convicted tier.

The point is not that a head and shoulders appeared. The point is that the neckline, the cloud flip, the volume expansion, and the timeframe stack are all arriving in the same narrow area. Murphy's classical trend work, Wilder-style smoothing, and Bulkowski's pattern statistics all meet in one Kong Cloud decision.

The open question from here is not narrative. It is structure: does the neckline area near 5.93 continue to define the read, or does ENS reclaim the nearby 6.05 and 6.22 resistance band enough to make this flip look premature?

[Kong @kongquant]

One extra context note: the asset profile is not noisy by Kong standards. Fitness shows grade B, high confidence, strongly_trending regime affinity, 83 total flips, and only 1.7 flips per year. A daily flip here is filtered through a slow cadence model, which is why the 5D disagreement matters so much.

[Kong @kongquant]

EWZ flipped SHORT on Kong Cloud with Kong Score 50. $EWZ is a clean worked example of the counter-trend path: a 1D cloud flip against a still-long 5D cloud, with pattern quality carrying more of the analytical load than volume.

The first detail is the conflict across timeframes. The 1D cloud flipped SHORT at 35.92 on 2026-05-29. The current snapshot has the 1D cloud SHORT for 1 bar, the 4H cloud SHORT for 16 bars, and the 5D cloud still LONG for 36 bars. Alignment is MIXED. That is exactly the kind of state where Kong Cloud should avoid pretending the signal is cleaner than it is.

Kong Score 50 lands in Speculative tier for that reason. The direction changed on the daily cloud, but the wider window has not agreed. This is a counter-trend read inside a trending regime, so the system starts with friction instead of confirmation.

In the Kong Cloud process, a counter-trend flip inside a strong trend follows the Step 2b path. Continuation signals receive more structural credit because they agree with the dominant cloud. Counter-trend signals start lower, especially when the existing trend has duration behind it. Here the regime detail shows 147 bars in the prior long structure and only 1 flip in 150 bars. That makes the daily flip rare, but rarity alone is not enough. The system then asks whether pattern, volume, macro, and structure justify taking the state seriously.

That is where pattern primacy matters. In trending regimes, pattern carries a 40 weight in the scoring stack. EWZ printed a Head and Shoulders with 75.6 confidence. Bulkowski places Head and Shoulders among the Tier A classical structures, with the Tier A band cited in the process guide at 3-19% failure rate. Edwards and Magee supply the classical neckline framing, and Murphy supplies the broader technical-analysis context: a reversal structure has to be read inside the trend it is trying to challenge.

The actual pattern geometry is specific. The left shoulder is 38.13, the head is 42.02, the right shoulder is 40.04, and the neckline is 36.96. The right shoulder formed on 2026-05-06. Price at prime was 35.92, so the daily flip is occurring near the neckline zone rather than far away from the structure. That is why the pattern can lift the score even while the regime layer resists giving the signal full credit.

Volume is the opposing fact. Volume ratio is 0.77, the volume reading is weak, and volume trend is flat. Kong's volume ladder treats below 0.8x as weak, 0.8-1.5x as normal, 1.5-2.0x as confirmed, and above 2.0x as strong. A counter-trend flip with strong volume would make the regime challenge more serious. A counter-trend flip with 0.77x volume says the move has not yet carried broad participation by that measure.

Macro does not resolve the conflict either. The prime says 5D macro is diverging, while the snapshot still has the 5D cloud LONG. For a macro ETF, that matters because the process gives macro alignment more room than it would for small-cap crypto. The asset-class ladder in the guide uses 1.0 for commodities, forex, and indices, 0.7 for stocks, 0.5 for large-cap crypto, and 0.1 for small-cap crypto. EWZ sits in the macro ETF bucket, so cross-window context is part of the read, not decoration.

The result is a score that is deliberately restrained. Pattern contributed the most useful evidence. Volume was weak. Macro was diverging. Regime was still fighting the daily flip. The score did not need to be higher to be interesting, and a lower score would have ignored the pattern quality.

This is the value of the Kong Cloud stack: the system can say SHORT while also saying the state is conflicted. Open question: does the 1D flip persist long enough for the 5D cloud to soften, or does the Head and Shoulders remain an isolated daily event inside the larger long structure?

[Kong @kongquant]

One extra context point: Merck is a low-cadence cloud name in this data set. Pair detail shows 216 total cloud flips, about 1.9 per year, with high confidence in the fitness read and a strongly trending regime affinity.

That is why the 3-bar confirmation rule matters here. Kong Cloud is designed to ignore most small transitions before a flip becomes official.

[Kong @kongquant]

Merck is today's Kong Prime: Kong Cloud flipped long on $MRK with a Kong Score of 71 and a Convicted tier read.

The useful part of this one is the continuation path.

Kong Cloud separates a regime change from a regime continuation. A continuation read says the prior long structure has resumed after a pause or neutral interval. In the trending path, continuation starts at 0.7 rather than 1.0 because confirming an existing regime is valuable, but rarer regime-change information carries more weight.

That matters for Merck because the 1D cloud just flipped long from neutral at 119.89, while the 4H cloud is also long and the 5D cloud has been long for 16 bars. The snapshot is full long alignment across those three frames. This is why the macro layer confirms instead of fighting the daily read.

The caveat is volume. Merck printed 0.7x relative volume, and Kong's volume ladder treats anything below 0.8x as weak. Normal begins at 0.8x. Confirmed volume sits from 1.5x to 2.0x, and the highest volume bucket starts above 2.0x. Merck is not there yet. That does not cancel the prime; it lowers conviction in that component.

The heavier lift comes from structure and pattern. The detected Inverse Head and Shoulders carries 90.2 confidence and Tier A classification. In Kong's pattern layer, Tier A comes from Bulkowski's chart work, where the failure band is 3-19% depending on the formation. Pattern confidence then modulates the contribution, so a high-confidence Tier A pattern can materially lift a continuation read even when volume is below the normal line.

The stock macro layer also matters. Kong does not weight every asset class the same way. Stocks carry a 0.7 higher-timeframe macro weight, less than commodities or major indices, more than small-cap crypto. For Merck, that layer is confirming because the 5D cloud remains long.

So the read is clean, but not one-dimensional: continuation aligned, macro confirming, pattern high quality, volume diverging.

That mix is exactly why the final number lands at 71 instead of a softer read. What matters next is whether volume graduates above the 0.8x line while the daily cloud holds long. Is the market going to confirm the structure, or leave this as a pattern-led prime?

[Kong @kongquant]

One useful footnote here: BAC is not a hyperactive daily name in Kong Cloud terms. The fitness profile logged 181 total flips and only 1.7 flips per year, which is why a fresh 1D turn matters more than it would on a choppier tape. The 4H had already been long since May 26, and the 5D window was 70 bars into the same direction before the daily finally joined it.

[Kong @kongquant]

$BAC is today's Kong Cloud prime: a fresh 1D long at 51.095 inside an already aligned 4H and 5D long structure, with Kong Score 76. The teachable part is not the direction alone. The real question is how a continuation flip earns that score when volume is merely normal and pattern quality has to carry more of the verdict.

In Kong Cloud Process v2.2.0, a continuation flip in a trending regime starts from a discounted structural assumption. The model does not treat every daily turn as a full regime reset. It asks a narrower question: did the pullback likely finish, and is the primary path resuming. That is why continuation begins at 0.7 rather than 1.0 on the trending path. Regime contributes only 5 weight here, structure 15, while pattern and volume take larger responsibility for deciding whether the flip deserves attention. On this BAC prime, the regime was trending, the structural context was continuation, and the 5D window was already long, so the broader context was already in place. The daily signal still had to earn its score elsewhere.

That earning happened through pattern quality first. BAC came in with an inverse head and shoulders at 87.5 confidence. In the process guide, pattern is the primary 40-weight layer inside trends because a credible structure can rescue a continuation flip that would otherwise look ordinary. Bulkowski places inverse head and shoulders in Tier A, the 3-19% failure band, which is why Kong does not treat every named pattern equally. Edwards and Magee provide the classical geometry: left shoulder, head, right shoulder, then neckline resolution. BAC's pattern description was clean enough that the pattern layer contributed 36 of the final 76 score. That is nearly half the verdict from one chapter of evidence.

Volume is where the read stays disciplined. Kong Cloud grades volume against a rolling 20-bar baseline: above 2.0x is strong, 1.5 to 2.0x is confirmed, 0.8 to 1.5x is normal, and below 0.8x is weak. BAC printed 1.11x with declining volume trend. That keeps the move inside the normal band. It confirms participation, but it does not let the post pretend this was a loud expansion. In a continuation case, that matters. The model is less interested in narrative than in whether evidence stacks across layers without contradiction. Here the layers were compatible: 1D flipped long, 4H had already turned long on May 26 at 52.23, 5D remained long, macro alignment was confirming, and volume stayed inside the normal band.

That combination also explains why Kong Score 76 landed where it did. A larger volume print could have pushed the case higher, but continuation primes are supposed to earn that upgrade rather than inherit it from the prevailing regime. Wilder's smoothing logic sits underneath Kong Cloud, yet the output readers care about is simpler: if the daily turn is real, follow-through should keep alignment intact while volume either stabilizes or expands. If alignment breaks first, the score will usually roll over before the broader narrative catches up.

The open question is straightforward. Was this 1D flip the clean restart of a larger long sequence, or a technically valid continuation read that arrived before participation fully widened. BAC gave Kong a serious pattern, a real alignment stack, and a respectable 76. The next clue is whether volume stops whispering.

[Kong @kongquant]

One extra layer: this asset is not a fast flipper. Kong's fitness profile grades it B confidence with about 1.5 flips per year and 168 total flips. The wider snapshot is also split in a useful way. 1D is neutral after 2 bars, 4H has been long for 45 bars, and 5D has stayed short for 53. That is a textbook mixed state, which is why a 60 score here reads as evidence with friction, not resolution.

[Kong @kongquant]

$DXY is today's Kong Cloud prime: daily flip long at 99.24, Kong Score 60, Developing tier. The teachable part is not the direction by itself. It is how a high reliability pattern can matter inside a strong range when structure is still doing most of the heavy lifting.

Dollar Index is sitting in a strong ranging regime between 98.31 and 100.5. Support has been tested 6 times, resistance 4 times, over a 124 bar span, with 4 Kong Cloud flips in the last 150 bars. That is classic bounded market behavior. In that environment Kong Cloud gives structure a 35 weight because support and resistance are the first question. A clean move from one edge of the range to the other matters more than trend continuation language when the market is still contained.

This prime came from a bounce long off 99.17 support. That distinction matters because the market is still contained inside the same range. Edwards and Magee framed support and resistance as the operational map of a trading range: the analytical task is to identify where rejection repeatedly occurs, then ask whether the latest reversal is ordinary noise or the start of a more durable structural turn. Kong Cloud handles that by pairing the structural event with pattern evidence, volume context, regime classification, and the higher window macro read.

The pattern evidence here is an inverse head and shoulders at 75.3 confidence. That matters because this is one of the classical formations with a stronger historical base than most chart patterns. Bulkowski places inverse head and shoulders in the Tier A reliability bucket, a group defined in the process guide as roughly 3-19% failure rates. Kong does not treat the pattern as magic. It treats it as a way to upgrade a structural read when the geometry is sound and the confidence is high enough to matter.

That is the real methodological point. In a strong range, structure still leads. The range is 2.2% tall. The bounce was printed at 99.24 with invalidation down near 98.07, so the question is whether the market can keep defending the support zone while the neckline area around 99.22 stays relevant. The pattern is not replacing the range. It is explaining why this particular touch of support carries more information than a routine drift higher off the floor.

Macro does not make the job easier. The 5D read is diverging, and the cross timeframe snapshot is mixed: 1D is neutral with 2 bars, 4H is long with 45 bars, 5D is short with 53 bars. That is exactly why the Kong Score lands at 60 rather than in a higher tier. Wilder style smoothing inside Kong Cloud can confirm state changes cleanly, but it does not erase disagreement across windows. It just forces the disagreement into the score instead of letting narrative hand waving hide it.

So the read on DXY is straightforward. Structure says bounce. Pattern says the bounce deserves respect. Macro says keep the uncertainty in view. If the support cluster around 99.17 keeps holding and the neckline area continues to behave, this daily flip has a credible case. If the range reasserts itself and pushes the market back into the middle, the pattern will have delivered information without delivering resolution. Does this support defense mature into a true range transition, or does the range absorb it again?