Desk State Advisor 2.12

PSRC Desk State Panel v2.12 Unified is a TradingView decision-context system built to compress multiple independent market domains into a compact desk panel. Its purpose is not to generate a stream of attractive signals. Its purpose is to make the market environment legible before a trader commits risk.

The panel combines higher-timeframe structure, an adaptive causal regime model, auction profile behavior, prior-session value rotation, liquidity acceptance and rejection, engine-native flow, volatility, execution quality, a manual risk gate, and a directional Desk Power meter. The final command state can approve, watch, wait, or veto without forcing a trade when evidence conflicts.

Download Desk State AI [PSRC]

Download the user manual:

PSRC_Desk_State_v2_12_Professional_User_Manual.pdf

Download PDF • 311KB

Download the technical whitepaper:

PSRC Desk State Technical White Paper.pdf

Download PDF • 224KB

Also vailable on:

• Ninjatrader

• QuantConnect (Lean Engine)

• Python Stack (andas, scikit-learn, XGBoost/LightGBM, and PyTorch)

STYLE-AGNOSTIC BY DESIGN

A scalper, swing trader, breakout trader, ICT/SMC trader, VWAP trader, liquidity-reversal trader, mean-reversion trader, discretionary trader or systematic operator can keep the same entry logic. Desk State provides the context in which that logic is deployed.

What the tool is

·       A multi-domain market-context dashboard.

·       A causal regime and auction-state framework.

·       A disciplined conflict detector that treats WAIT as a valid decision.

·       A common language for discretionary traders, quants and risk reviewers.

·       A forward-testable decision layer whose states can be logged and analyzed.

What the tool is not

·       It is not a claim to directly observe the intentions of banks, funds, dealers or market makers.

·       It is not a guaranteed prediction engine.

·       The Desk Power percentage is not a probability of profit.

·       The Universal OHLCV flow model is not equivalent to exchange-native aggressor order flow.

·       The execution cell is a chart-data proxy; it does not know every broker-side spread, depth or slippage condition.

·       The manual PASS risk setting is not a substitute for an actual risk process.

Operating principle

A retail indicator often asks: “Is this indicator bullish or bearish?” PSRC Desk State asks a different sequence: What regime are we in? Where is price relative to value? Is value migrating? Has liquidity been accepted or rejected? Does flow support the move? Is the execution environment healthy? Is risk authorized? Only then does it consider whether a directional READY state is justified.

The Desk State AI is a submodule across all our trading systems that protect the account

2. The Institutional-Thinking Framework

The word institutional in this manual describes a decision process, not privileged access. The panel is institutional-style because it emphasizes context, state uncertainty, auction location, liquidity, execution and risk before trigger selection.

Retail-style shortcut	PSRC institutional-thinking alternative
Count bullish indicators	Separate independent evidence domains; do not double-count the same price information.
Trade every signal	WAIT, WATCH and VETO are first-class outcomes.
Assume trend is stable	Measure regime persistence, entropy and duration-conditioned exit hazard.
Treat price as direction only	Ask where price is trading relative to POC, VAH, VAL, VWAP and prior-session value.
Call every sweep “smart money”	Distinguish rejection, acceptance and unresolved auction behavior.
Ignore execution quality	Gate READY when volatility/liquidity proxies are degraded.
Let conviction override risk	Keep a separate manual risk authority that can veto the trade.

Why conflict is valuable

A sophisticated desk should not force a direction merely because one domain is strong. A bullish Markov trend can coexist with bearish higher-timeframe structure, lower value, opposing flow or poor execution. PSRC Desk State surfaces that disagreement as CONFLICT rather than averaging it into false confidence.

DESK RULE

A strong signal in the wrong environment is still the wrong trade for many styles. Context does not replace edge; it determines whether the edge is being asked to work under favorable or hostile conditions.

Desk State on Near Protocol

3. Panel Anatomy and One-Glance Interpretation

The panel is intentionally compact. The first row answers what the market is doing. The second row answers whether acting now is sensible. The power bar summarizes directional alignment without overriding any gate.

DESK BIAS	MARKOV	PERSIST/HZ	ENTROPY	STRUCTURE	AUCTION
BEAR	RANGE	92% / HZ 8%	23% LOW	D↓ H4↓ H1↓	BEAR EXT / INSIDE VALUE
LIQUIDITY	FLOW	VOL	EXEC	RISK	COMMAND
PDL -0.54A	FP MIX 0%	COMPRESS 10%	AVOID PROXY	NOT REVIEWED	WAIT ALIGN
DESK POWER  BEAR 60%   ■ ■ ■ ■ ■ ■ □ □ □ □

Cell	Question it answers
Desk Bias	What directional thesis survives the domain gates?
Markov	What causal regime is active now?
Persist/HZ	How persistent is the state, and how likely is it to exit given its age?
Entropy	How uncertain are the outgoing regime transitions?
Structure	What is the confirmed Daily/H4/H1 directional architecture?
Auction	Where is price relative to current and prior value, and is value accepting/rejecting/migrating?
Liquidity	What prior-day/week boundary is being approached, accepted or rejected?
Flow	Does the selected engine show supportive, opposing or mixed participation?
Vol	Is volatility compressed, normal or expanding?
Exec	Is the chart-derived execution environment good, degraded or avoid?
Risk	Has the external risk process passed, vetoed or not reviewed the trade?
Command	What is the highest-priority desk action or blocker?
Desk Power	How strongly are the available directional domains aligned?

4. Core Logic: How the Engine Thinks

4.1 Higher-timeframe structure

The structure layer uses confirmed higher-timeframe data to reduce repaint risk. The default mode is H4 EMA20 + Daily EMA50. An alternate mode uses H4 EMA200. H1 EMA20 is displayed as tactical alignment context.

In the default mode, the directional structure gate requires H4 and Daily agreement. H1 can disagree and will remain visible so the operator can distinguish strategic direction from tactical pullback.

Desk State on USDJPY

4.2 Adaptive Markov regime

The Markov block is not a fitted hidden Markov model. It is a causal five-state classifier followed by an exponentially decayed first-order transition model. The states are Bull Trend, Bear Trend, Range, High Volatility and Transition.

State	Core condition / interpretation
Bull Trend	Trend strength above threshold, fast regime EMA above slow EMA, positive fast-EMA slope.
Bear Trend	Trend strength above threshold, fast regime EMA below slow EMA, negative fast-EMA slope.
Range	ADX at or below the range threshold.
High Volatility	ATR percentile at or above the high-volatility threshold; this state has precedence.
Transition	Conditions do not cleanly qualify as trend, range or high-volatility expansion.

Transition observations are exponentially discounted using the Adaptive Transition Decay setting. This prevents thousands of old observations from permanently dominating current behavior.

Persistence, entropy and hazard

·       Persistence estimates how strongly the current regime tends to remain in itself.

·       Entropy summarizes uncertainty across the five possible outgoing transitions. Lower entropy means a more concentrated transition distribution.

·       Hazard estimates the probability that a regime exits on the next bar conditional on how long it has already persisted. The engine learns separate duration bins.

·       The regime must have sufficient effective samples, adequate persistence, acceptable entropy and—when enough duration data exists—an exit hazard below the veto threshold before it is considered stable.

4.3 VWAP and value migration

Session VWAP and anchored Weekly VWAP provide a second auction/location perspective. The engine asks whether price is above or below both, whether value is migrating up or down, whether price is near a preferred pullback zone, and whether the market is too extended to chase.

4.4 Liquidity acceptance and rejection

Prior-day and prior-week highs/lows are treated as observable reference levels—not as guaranteed “resting liquidity.” A wick through a level followed by rejection is different from repeated closes beyond that level. The engine explicitly separates rejection from acceptance.

4.5 Execution quality

Execution quality uses available chart information such as relative volume, volatility percentile and LVN context to label conditions GOOD, DEGRADED or AVOID. This is deliberately conservative. It is not a substitute for broker spread, depth, slippage, margin or fill-quality checks.

5. Auction Profile Engines: Footprint vs Universal

v2.12 is a unified release with a Profile Engine dropdown. Both modes feed the same POC/VAH/VAL, value-area, prior-session and command logic. The difference is how the profile and flow evidence are built.

Feature	Footprint	Universal OHLCV
Profile source	TradingView footprint rows aggregated into a session map.	Confirmed chart OHLCV distributed across fixed price rows.
Flow	Normalized footprint delta + buy/sell imbalance-row edge.	Candle/volume-pressure proxy + relative volume.
POC / VAH / VAL	Yes	Yes
Prior-session value rotation	Yes	Yes
Order-flow fidelity	Higher where footprint data is meaningful.	Proxy only; not aggressor-classified flow.
Best use	Primary engine when footprint data is available and appropriate.	Fallback / robustness comparison / symbols where chart volume is the available source.

IMPORTANT

The unified source contains a footprint request. Therefore the unified release requires an account/symbol environment that supports TradingView Footprint functionality even when Universal OHLCV is selected. A standalone Universal build is the compatibility edition for environments without Footprint access.

Profile concepts

Term	Meaning in PSRC Desk State
POC	The profile row with the greatest volume.
VAH / VAL	Upper and lower boundaries of the selected value area (70% by default).
HVN	Current row volume is high relative to POC volume; often associated with acceptance / two-way trade.
LVN	Current row volume is low relative to POC volume; can indicate thin traversal or poor execution context.
VA Acceptance	Repeated closes beyond VAH or VAL by the configured ATR buffer.
VA Rejection	A move beyond value that fails and closes back inside.
POC Migration	Developing POC moves meaningfully up, down or remains stable.

Prior-session value rotation

The engine retains the completed prior session POC, VAH and VAL and compares the developing current session against them. The resulting context can be HIGHER VALUE, LOWER VALUE, INSIDE VALUE, OVERLAP HIGH, OVERLAP LOW, ABOVE VALUE ACCEPT, BELOW VALUE ACCEPT, OVERLAP or PRIOR MIXED.

When Require prior value alignment is enabled, an explicitly opposing prior-value rotation can block READY without forcing an opposite trade.

6. Desk Bias and Conflict Logic

Desk Bias is an institutional-style inference, not a claim of direct institutional intent. It is domain-gated rather than score-voted. The engine considers higher-timeframe structure, adaptive regime, auction/location, prior value, liquidity, and participation flow. A domain may confirm, remain neutral or contradict.

NO INDICATOR DEMOCRACY

Five indicators derived from the same price series are not five independent pieces of evidence. Desk State avoids a generic confluence count and instead works with distinct information domains.

Bias states

State	Interpretation
STRONG BULL	Structure and Bull Trend regime align; auction is bullish; no bearish participation contradiction; directional profile/liquidity/flow evidence is present.
BULL	Bullish trend and auction domains survive the gates but do not meet the stronger alignment standard.
NEUTRAL	Directional evidence is insufficient without a direct contradiction.
CONFLICT	One or more major directional domains explicitly disagree.
BEAR / STRONG BEAR	Mirror image of the bullish logic.

7. Desk Power: The 10-Segment Alignment Meter

Desk Power is a signed directional alignment index displayed as ten uniform segments. Green represents bullish alignment; red represents bearish alignment. Ten lit segments represent maximum modeled alignment in that direction. The number is deliberately called power, not confidence, because it is not statistically calibrated as a probability of profit.

Domain	Maximum contribution	Notes
HTF Structure	25	Directional structure vote.
Adaptive Markov	Up to 25	Weight is reduced by regime reliability derived from persistence, entropy and exit hazard.
Auction / Location	25	Current profile, value, liquidity and location direction.
Prior-session value	10	Included only when a completed prior profile is available.
Engine-native flow	15	Included only when flow is enabled and available.

The score is normalized across available domains. If the desk detects an explicit conflict, absolute power is capped at 60%. This prevents the cosmetic meter from displaying extreme conviction while the primary panel says CONFLICT.

DO NOT TRADE THE METER

Desk Power does not change BUY READY, SELL READY, risk, profile, execution or regime logic. It is a fast diagnostic summary. A 100% reading means maximum modeled alignment—not a 100% chance of a winning trade.

8. Command States and Universal Alerts

The command cell is a blocker-first state machine. It reports the highest-priority reason the system is not ready rather than hiding useful information behind a generic “wait.” On a live, unclosed bar, BAR OPEN is appended as a modifier so the operator sees the real blocker while still respecting closed-bar confirmation.

Command	Meaning
USE ≤ H1	Chart timeframe is above the supported operating range.
NO FOOTPRINT / NO VOLUME	Selected engine lacks usable data.
WAIT SESSION	Custom-session profile is outside its active window.
VETO	Manual risk state or event blackout blocks the trade.
WARMUP	Regime statistics do not yet have enough effective data.
WAIT PROFILE	Profile has not matured or directional profile evidence is absent.
WAIT REGIME	Markov persistence/entropy/hazard conditions are not stable enough.
WAIT CONFLICT	Major evidence domains disagree.
WAIT BIAS	No directional desk thesis survives the gates.
WAIT ALIGN	Bias direction and trend-regime direction are not compatible.
WAIT LOCATION	Directional auction/location state is absent.
WAIT CHASE	Price is too extended from session value.
WAIT INV	Structural invalidation is too distant or unsuitable.
WAIT PRIOR VA	Prior-session value rotation explicitly opposes the direction.
WAIT FLOW	Directional flow is required but does not confirm.
WAIT EXEC / VETO EXEC	Execution proxy is degraded or unacceptable.
BUY WATCH / SELL WATCH	The setup survives market gates but the manual risk gate has not passed, or the live bar is still open.
BUY READY / SELL READY	All required model, execution and manual risk gates pass on a confirmed bar.

Universal alerts

BUY READY and SELL READY alerts are driven by the common command engine, so they work regardless of whether Footprint or Universal OHLCV is selected. Alerts fire on the transition into READY, not repeatedly on every subsequent bar.

·       PSRC BUY READY

·       PSRC SELL READY

·       PSRC VETO

·       PSRC PROFILE ACCEPT ABOVE

·       PSRC PROFILE ACCEPT BELOW

·       Any alert() function call for dynamic messages containing ticker, timeframe, selected engine, bias and prior-value context.

Best practice: create READY alerts as once-per-bar-close alerts and keep Require closed bar enabled during formal forward testing.

9. Risk Gate and Human Authorization

Risk is intentionally separated from the signal engine. The dropdown has three states: NOT REVIEWED, PASS and VETO. High-impact event blackout can override the manual state and veto the setup.

Setting	Use
NOT REVIEWED	Default for observation and discretionary desk workflow. Prevents READY authorization.
PASS	Use only after the operator’s actual risk process has cleared exposure, correlation, leverage/margin, event and execution constraints.
VETO	Explicitly blocks execution regardless of directional alignment.

Two valid forward-test protocols

·       Signal-engine test: set PASS for the full session to isolate the quantitative engine. This tests READY quality without human risk-gate variability.

·       Desk-workflow test: leave NOT REVIEWED until a human performs the defined risk review, then select PASS. This tests the complete human + system process.

Do not mix those two datasets. Their missing READY events have different causes.

10. Using Any Trading Style Within Institutional Context

THE OPERATING PHILOSOPHY

Your trading style chooses the setup. PSRC Desk State chooses the context. The panel is not asking every trader to become a trend trader; it is asking every trader to know what regime, value, liquidity, flow, execution and risk environment they are trading into.

This is the most important concept in the manual. The same panel can support very different trade construction methods because its job is contextual rather than stylistic.

Trading style	How to use PSRC Desk State
Scalping	Use M5 as the primary desk context. Execute your existing lower-timeframe trigger only when the broader state is understood. A scalp against strong Desk Power is a deliberate counter-context trade, not an accidental one.
Intraday trend	Favor directional structure + Bull/Bear Trend regime + supportive auction migration. Use your own pullback, breakout or continuation trigger.
Swing	Use H1 as the reference panel while Daily/H4 structure supplies strategic context. Your holding period may extend beyond the reference bar; validate this separately.
ICT / SMC	Keep your liquidity raid, FVG, order-block or market-structure entry rules. Use Desk State to distinguish whether the setup is aligned with regime, value rotation and auction acceptance/rejection.
VWAP / profile	Use session/weekly VWAP, POC/VAH/VAL migration, value acceptance/rejection and prior-session value to contextualize your existing profile setups.
Liquidity sweep reversal	A PDH/PDL or PWH/PWL rejection is context, not an automatic reversal. Check regime, value, opposing acceptance and flow before acting.
Breakout	Look for acceptance beyond value rather than a single wick. Rising/higher value and supportive flow strengthen the contextual case.
Mean reversion	Use Range / balanced auction / HVN / value context to identify environments where your reversion method belongs. Note that BUY READY / SELL READY are intentionally conservative trend-regime authorization states, so a range trader should use the panel context rather than demand a READY signal for every mean-reversion setup.
Systematic / algorithmic	Use the discrete states and READY alerts as features or gates in a separately validated strategy. Do not assume the dashboard itself establishes expectancy.

A practical way to think about it

Suppose your preferred method is a bullish fair-value-gap retracement. Desk State does not decide whether that FVG is technically valid. It tells you whether the FVG is occurring in Bull Trend or Transition, above or below value, during higher or lower value rotation, with or against flow, in a good or poor execution environment, and under a passed or unreviewed risk state. The trader keeps the setup; the panel upgrades the context.

11. Timeframes and Workflow by Holding Horizon

The script supports chart timeframes up to H1. Above H1 it displays USE ≤ H1. The internal context still includes H4 and Daily information through confirmed higher-timeframe requests.

Reference timeframe	Best use	Guidance
M1-M3	Experimental micro-execution	Possible, but noisier. Profile-row calibration and regime transitions become more sensitive. Not recommended for initial FT1 testing.
M5	Scalping and active intraday	Primary tactical timeframe. Excellent for liquidity raids, breakout acceptance and active session flow.
M15	Intraday / default research baseline	Recommended primary baseline. Good balance between responsiveness and state stability.
H1	Slower intraday / swing context	Use when the trade thesis is broader. Daily/H4 structure remains the strategic anchor.
H4 / D1 chart	Not supported as panel timeframe	Use H1 or lower; the script already imports confirmed H4/Daily context.

Suggested workflow by style

1.     Choose one reference timeframe for the panel and keep it fixed for the test period.

2.     Define your own setup rules before reading Desk State.

3.     Read the panel from left to right: Bias → Regime → Uncertainty → Structure → Auction → Liquidity → Flow → Volatility → Execution → Risk → Command.

4.     Use Desk Power only as a fast alignment summary, never as a substitute for the panel.

5.     If your style requires a lower execution timeframe, execute there while preserving the approved reference-chart context.

12. Best Practices and Operating Discipline

·       Respect WAIT. A quiet panel is a feature, not a failure.

·       Do not change Markov, profile or flow thresholds in the middle of a forward-test cohort.

·       Do not convert a 90–100% Desk Power reading into larger position size unless that sizing rule has been independently validated.

·       Prefer Footprint where the data is meaningful; keep Universal as a robustness comparison, not a competing vote.

·       Treat CFD/FX volume according to its actual feed limitations. Do not describe proxy volume as centralized institutional flow.

·       Keep Require closed bar enabled for formal tests and live alerting unless you are explicitly studying intrabar behavior.

·       Verify the custom session timezone if you use Custom Session profile scope.

·       Use profile row size appropriate to the instrument. If a single bar spans excessive rows, increase ticks per row to control runtime.

·       Use NOT REVIEWED by default in desk workflow. PASS is an explicit authorization state, not a confidence button.

·       Log every READY event, including losers and signals you would personally have skipped. Selective screenshots destroy test validity.

·       When the panel says CONFLICT, identify the conflicting domains. Do not simply wait for the color to change without understanding why.

·       Version every meaningful logic change. Never mix v2.12 results with a later tuned build in the same statistical sample.

13. Recommended Baseline Settings

The defaults below are the recommended FT1 baseline. Keep them unchanged across the first human test cohort unless a setting is structurally incompatible with a symbol or feed.

Group	Setting	FT1 default
Display	Panel position	Bottom Center
Display	Value font size	12
Display	Require closed bar	ON
HTF	Filter	H4 EMA20 + Daily EMA50
HTF	H1 EMA	20
HTF	H4 fast / slow EMA	20 / 200
HTF	Daily EMA	50
Markov	Regime fast / slow EMA	20 / 50
Markov	DMI / ADX smoothing	14 / 14
Markov	Trend ADX / Range ADX	22 / 16
Markov	ATR / Volatility rank	14 / 100
Markov	High-volatility percentile	80
Markov	Minimum effective samples	30
Markov	Minimum persistence	0.55
Markov	Maximum entropy	0.70
Markov	Transition smoothing alpha	0.50
Markov	Adaptive decay	0.995
Markov	Hazard age bins	3 / 8 / 16
Markov	Minimum hazard exposure	8
Value	Session / Weekly migration bars	3 / 5
Value	Migration threshold	0.03 ATR
Value	Preferred pullback zone	0.60 ATR
Value	Maximum chase distance	1.50 ATR
Value	Maximum invalidation distance	2.50 ATR
Profile	Engine	Footprint
Profile	Ticks per row	25
Profile	Value area	70%
Profile	Footprint imbalance	300%
Profile	Universal max bins/bar	160
Profile	Scope	Exchange Day
Profile	Custom session	09:30-16:00 America/New_York
Profile	Minimum confirmed bars	8
Profile	Migration bars / threshold	3 / 0.05 ATR
Profile	Acceptance closes / buffer	2 / 0.03 ATR
Profile	Rejection memory	4 bars
Profile	LVN / HVN thresholds	25% / 65% of POC volume
Profile	Require profile for READY	ON
Profile	Require prior-value alignment	ON
Liquidity	Rejection memory	6 bars
Liquidity	Acceptance closes / buffer	2 / 0.05 ATR
Flow	Source	Engine Native
Flow	Require directional flow for READY	OFF
Flow	Footprint delta threshold	0.12
Flow	Minimum imbalance edge	1
Flow	Universal smoothing / RVOL baseline	5 / 20
Flow	Universal pressure / min RVOL	0.12 / 1.00
Execution	Minimum RVOL for GOOD	0.65
Execution	GOOD max ATR percentile	90
Execution	AVOID ATR percentile	95
Risk	External risk state	NOT REVIEWED
Risk	Event blackout	OFF
Visuals	Optional chart levels	OFF by default

14. Forward Testing Protocol

v2.12 is best treated as a frozen human forward-test candidate. The purpose of FT1 is to measure behavior before optimizing it.

Recommended initial universe

Start with XAUUSD, NAS100/USTEC, US30, EURUSD and GBPUSD on M5 and M15. Add instruments only after the data collection process is stable.

Record every READY event

Field	What to capture
Timestamp	Exact signal close and timezone.
Instrument / timeframe	Ticker and reference chart timeframe.
Engine	Footprint or Universal OHLCV.
Command	BUY READY / SELL READY.
Bias / Markov	Desk Bias, regime, persistence, hazard and entropy.
Structure / Auction	D/H4/H1 plus full auction and prior-value state.
Flow / Vol / Exec	Full displayed states.
Desk Power	Direction and percentage at the READY bar.
Entry reference	Close or separately defined execution price.
Structural invalidation	Level/distance according to the testing plan.
MFE / MAE	Maximum favorable and adverse excursion.
R outcomes	Whether +0.5R, +1R, +2R or stop was reached first.
Bars to outcome	Time-to-resolution in bars.
Notes	Optional human observation; never use notes to erase losing data.

Do not change thresholds because the first ten trades look poor or excellent. Seek enough events to evaluate the overall signal and the major conditional subgroups. A practical minimum for directional conclusions is roughly 100 READY events overall, with larger samples preferred before changing specific subgroups.

What to analyze

·       READY expectancy by instrument and timeframe.

·       MFE/MAE asymmetry before imposing a single exit model.

·       Bull/Bear Trend versus Transition or Range context at the time of setup formation.

·       Higher Value / Lower Value / overlap states.

·       Low-entropy versus high-entropy regimes.

·       Footprint BUY/SELL/MIX states and imbalance edge.

·       Prime location versus merely acceptable location.

·       Desk Power deciles as a descriptive variable—not as a pre-assumed probability.

15. Limitations, Data Quality and Interpretation Boundaries

Boundary	Operational implication
No privileged institutional feed	Desk Bias is an inference from observable market data, not direct knowledge of institutional intentions.
Footprint depends on source	The quality and meaning of footprint volume depend on symbol, venue and TradingView data availability.
FX/CFD volume can be proxy volume	Do not assume centralized exchange volume where the instrument does not provide it.
Universal is reconstructed	POC/VAH/VAL are derived from distributing chart-reported volume across price rows; this is an approximation.
Execution proxy is incomplete	Real spread, order-book depth, slippage, broker constraints and gaps may differ from chart conditions.
Risk state is manual	PASS means the operator says risk review was completed; the script cannot verify portfolio exposure on its own.
Desk Power is not win probability	Treat it as directional alignment only.
READY is trend-oriented	Mean-reversion and range styles should use the contextual states rather than expect READY to authorize every valid non-trend trade.
Forward edge is unproven until tested	Operational correctness is not statistical proof of profitability.

16. Quick Reference

Bullish institutional context

Typical ingredients: positive HTF structure, Bull Trend regime, stable persistence/low entropy, higher or accepted-above value, rising POC/value, VAL rejection or bullish liquidity acceptance, supportive flow, healthy execution, passed risk. Your own bullish setup still decides the entry.

Bearish institutional context

Mirror image: negative HTF structure, Bear Trend regime, stable state, lower or accepted-below value, falling POC/value, VAH rejection or bearish liquidity acceptance, supportive sell flow, healthy execution and passed risk.

Balanced / mean-reversion context

Typical ingredients: Range regime, price inside value, HVN/POC proximity, low directional profile vote and mixed flow. This may be excellent context for a separately validated mean-reversion method even though BUY READY / SELL READY may remain unavailable.

Conflict context

The desk is telling you the evidence is not coherent. Either wait, reduce the importance of the setup, or explicitly treat the trade as a counter-context strategy if that behavior is part of your tested plan. Desk Power is capped at 60% during explicit conflict.

ONE-SENTENCE OPERATING RULE

Never ask “What does Desk State tell me to trade?” Ask: “What does Desk State tell me about the environment in which I am about to use my trading method?”

Appendix A. Technical Logic Summary

A.1 Markov stability gate

The regime classifier uses current chart data causally. Transition weights are decayed by the configured factor on confirmed bars. The current regime is stable only when the effective sample count is sufficient, persistence clears the minimum, entropy is below the maximum, and—if enough duration observations exist—the duration-conditioned exit hazard remains below the hard hazard threshold.

A.2 Desk Power reliability adjustment

The Markov contribution to Desk Power is reduced when the regime itself is uncertain. Conceptually, reliability combines persistence, inverse entropy and inverse exit hazard. A regime with high persistence but high transition uncertainty therefore cannot receive the same influence as a clean, durable state.

A.3 READY gate sequence

1. Supported timeframe and usable profile data.

2. Active session, when Custom Session is selected.

3. No manual/event risk veto.

4. Sufficient adaptive-regime samples and mature profile.

5. Stable Markov state.

6. Directional Desk Bias with no unresolved conflict.

7. Bull Trend for BUY or Bear Trend for SELL.

8. Directional location without excessive chase or invalidation distance.

9. Directional profile evidence, when required.

10. No opposing prior-session value rotation, when required.

11. Flow confirmation if the operator elects to require it.

12. Execution quality acceptable.

13. Manual risk state PASS.

14. Confirmed bar, if closed-bar confirmation is enabled.

Appendix B. Glossary

Term	Definition
Acceptance	Repeated closes beyond a reference level by the configured buffer; distinct from a one-bar breakout.
Auction	The process by which price explores and establishes value across price levels.
Desk Bias	The directional thesis that survives PSRC domain gates.
Desk Power	Signed 0-100 alignment index across available directional domains; not probability of profit.
Entropy	Normalized uncertainty across possible next Markov states.
Exit Hazard	Estimated chance the current regime ends on the next bar given its current duration bucket.
Footprint	TradingView footprint-derived row data used for profile and flow in Footprint mode.
HVN	High-volume node relative to current POC volume.
LVN	Low-volume node relative to current POC volume.
Persistence	Estimated self-transition tendency of the current regime.
POC	Point of Control; the profile row with the most volume.
Prior Value	Completed prior-session POC/VAH/VAL context compared with the current developing profile.
Rejection	Price probes beyond a reference but fails to maintain acceptance there.
VAH / VAL	Value Area High / Low for the configured percentage of profile volume.
VWAP	Volume-weighted average price used as a value/location proxy.
WATCH	Market gates support a direction, but final READY authorization is not complete.

PLAZO SULLIVAN ROCHE CAPITAL