Overview

Your research cockpit: portfolio pulse, attention queue, opportunities and market context.

One system, one research flow.

Noweir routes market data, technical structure, fundamentals, catalysts, portfolio risk, discovery and model governance through the same engine. Use the command bar for the fastest path.

TODAY'S OPPORTUNITIES

Run Noweir for the current 3–5 name research set.

V5.1

No full-market run cached yet.

Research only. READY NOW can be empty when strict compliance, cash, event, portfolio, sizing or evidence gates are incomplete.

Setup Trigger Alerts

Tracking watchlist, re-entry names and current opportunity candidates.

V5.3

Live Market Board

Latest available intraday · Yahoo/yfinance

Loading live charts…

Portfolio Pulse

No run yet

What Needs Attention

Opportunity Radar

Watchlist Research Memory

Loading local memory…

Research Change Alerts

Comparing point-in-time memory…

Decision Review Queue

Linking journal decisions to research memory…

Process Learning

Analyzing documented decisions…

Position Lifecycle

Loading execution-event history…

Execution Quality

Waiting for lifecycle evidence…

Portfolio Attribution

Run portfolio to attribute capital.

Market Context

Model Governance

Stock Research

Deep single-name research with clear separation between evidence, timing, valuation and portfolio fit.

Chart terminal

Ticker

Profile

Live Price + Technical Charts

Select a range · auto-refreshes every 60s while open

Open full chart terminal Open all advanced research tools

Run research to load event markers.

RSI 14

MACD

Research Event Intelligence

Earnings · SEC · current news

Run research to load historical event reactions.

Canadian Primary Evidence · SEDAR+

Official sedarplus.ca documents only

Official SEDAR+ document URL

Issuer

Filing type

Paste a direct official SEDAR+ document URL. The document is treated as untrusted research input and cannot trigger orders.

Live Charts

Multiple latest-available intraday charts with portfolio-aware defaults. Data may be delayed and is not exchange-grade streaming.

Advanced chart terminal

Symbols · comma or whitespace separated

Refresh

Latest available intraday bars + quote state

Open this page to load live charts.

Live means latest available provider data.

Noweir polls Yahoo/yfinance for the newest available intraday bars and quote state. It is useful for research and monitoring, but it is not a direct exchange feed and should not be treated as brokerage-grade execution data.

Visual Analytics

Ask for a comparison, benchmark view, valuation/growth map, sector heatmap or catalyst calendar. Charts are descriptive research tools, not trade signals.

Visual prompt

Symbols

Period

Benchmark

Ready.

View name

Saved views

Normalized Performance

Each series starts at 100; useful for relative performance.

Current-Weight Portfolio Reference vs Benchmark

Snapshot-weight backcast · not realized account performance

Run or save your portfolio to build this reference comparison.

Valuation vs Growth

Forward P/E vs revenue growth

Run analytics to populate.

Sector / Industry Heatmap

1-month return grouped by current provider classification

Run analytics to populate.

Earnings + Catalyst Calendar

Provider dates should be re-verified with company IR before event-sensitive decisions.

Run analytics to populate.

Research Event Intelligence

Connect price action to dated earnings, SEC filings and current news evidence without claiming that nearby events caused the move.

Full evidence workspace

Symbol

Historical window

Descriptive event study only.

Historical Earnings Reactions

Returns are from nearby trading-close anchors; not causal estimates.

Date	EPS est.	Reported EPS	Surprise	1D	5D	20D	5-session range
No event study run yet.

SEC Filing Reactions

Primary filing dates + market prices

Current Event Timeline

Evidence proximity ≠ causation

Interpretation Boundary

Portfolio-aware Discovery

Named research modes feed an explicit 10-gate sequence. A strong company can still end at NO TRADE or PORTFOLIO BLOCK.

Discovery Controls

V4.0 · V1.5 evidence

Theme

Universe · comma or whitespace separated

Portfolio symbols · auto-filled from Portfolio when available

Confirmed halal/compliance · SYMBOL=halal, one per line

Cash budget

Behavioral gate

Zero candidates is a valid result.

Opportunity Matrix

Research map only · gates decide eligibility

Run discovery to plot candidates.

Decision Shortlist

No opaque BUY score

Ticker	Decision	Mode / why	Business	Trend	Setup	Valuation	Portfolio	Event	Evidence	First block	Size cap
No discovery run yet.

Explicit Gate Order

First non-pass stops the candidate

Theme Dependency Map

Model Validation

Value Finder

Generate potential value candidates from the US and Canadian markets, then run the same V4 portfolio, compliance, event and entry gates. Cheap is research evidence, not a trade instruction.

Market Scan

V4.1 · Yahoo custom screens

Markets

Deep candidates

Minimum market cap

Minimum price

Minimum 3-month average volume

Confirmed halal/compliance · SYMBOL=halal

Cash budget

Behavioral gate

No scan yet. Zero names is a valid result.

What V4.1 Searches

Transparent candidate generation

Earnings Value

Positive earnings with P/E ≤ 22x before deeper peer/history checks.

Growth-Adjusted Value

Positive earnings, P/E ≤ 35x and PEG ≤ 1.5.

Enterprise Value

Positive EV/EBITDA ≤ 14x before cash/debt and peer validation.

Positive FCF Value

Positive levered FCF with P/E ≤ 30x; V1.5 then computes FCF yield.

These screens only create a research universe. V1.5 then checks peer-relative valuation, historical valuation, FCF yield and relative implied ranges. V4.0 can still return NO TRADE or PORTFOLIO BLOCK.

Value Scan Pulse

No opaque BUY score

Potential Value Candidates

Research candidates, not intrinsic-value claims

Ticker	Value evidence	Found by	P/E	EV/EBITDA	PEG	FCF yield	Peer lenses	Relative range	Trap flags	V4 decision	First block
Run the value scan to generate candidates.

Interpretation Boundary

A low multiple can reflect genuine mispricing, cyclicality, deteriorating fundamentals, leverage, accounting effects or a structurally weaker business. V4.1 therefore shows value evidence and value-trap warnings separately. Halal status is never inferred from market-data metadata.

Autonomous Opportunity Finder

Generate market, recent-earnings and re-entry candidates automatically, then require the unchanged V4.0 mode rules and 10-gate stack to validate them.

Autonomous Scan

V4.2 · seed ≠ validation

Markets

Deep candidates

Minimum market cap

Minimum price

Minimum 3-month average volume

Confirmed halal/compliance · SYMBOL=halal

Cash budget

Behavioral gate

No autonomous scan yet. Zero validated modes is a valid result.

Research Hypotheses

Coarse seeds · exact V4 rules run later

Quality Pullback

Profitable + growing + quality seed. Deep validation requires RSI 38–58 and intact trend.

Trend Continuation

Positive 52-week momentum seed. Deep validation requires momentum, technical strength and price above SMA200.

Oversold but Structurally Intact

Quality names falling today. Deep validation requires RSI ≤ 40 with quality/growth intact.

Diversifier

Lower-beta quality seed. Deep validation uses actual 252-day correlation to your holdings.

Breakout Confirmation

Strong 52-week momentum seed. Deep validation requires price within 4% of the 52-week high.

Canadian Quality

Profitable Canadian quality screen. Only runs against Canadian exchanges.

Post-Earnings Reset

Recent provider earnings-calendar events. Deep validation checks event timing and business quality.

Re-entry Candidate

Open sale/trim history from the Re-entry Board. Current V4 entry gates still apply.

A market screen hit only explains why a ticker entered the research queue. It is counted as a validated opportunity mode only when the unchanged V4.0 rules match after deeper V1.5 analysis.

Scan Pulse

No forced candidates · no opaque BUY score

Mode Coverage

Seeded vs deeply validated

Mode	Seeded	Deep analyzed	Validated	Review
Run the autonomous scan to compare seed hypotheses with V4 validation.

Autonomous Research Queue

V4 modes + gate stack decide what survives

Ticker	Seed hypothesis	Validation	V4 mode	Business	Trend	Setup	Valuation	Portfolio	Event	Evidence	Decision	First block
No autonomous scan yet.

Boundary

The autonomous layer creates a research universe. It does not infer halal status, does not turn a screen hit into a trade, does not bypass portfolio overlap, and does not place orders. Provider market and earnings-calendar data remain secondary evidence.

Framework Setup Finder

Search specifically for the two entry types in your documented framework: BREAKOUT or PULLBACK. The tool shows the six framework gates separately from the broader V4 decision gates.

Setup Scan

V4.3 · exact framework rules

Optional tickers · blank = autonomous US + Canada scan

Markets

Technical candidate limit

Deep V1.2/V4 checks

Confirmed halal/compliance · SYMBOL=halal

Proposed new position %

Cash budget

Behavioral gate

No setup scan yet. Zero complete setups is a valid result.

Your Entry Contract

No opaque setup score

BREAKOUT

Break above clear resistance, preferably after consolidation and with elevated volume. +4% surge/chase days are marked WAIT.

PULLBACK

Established uptrend pulling to EMA10, EMA20, first EMA50 touch, or a prior breakout retest. Trendline support remains manual-only in V4.3.

20-day gate

Price below EMA20 fails unless a specific exception is documented. A first EMA50 touch below EMA20 is shown as an unresolved framework conflict.

Risk

Structural invalidation is shown explicitly. The strict entry gate blocks stops wider than 10%; >15% is separately flagged as hard-wide risk.

Reward / risk

At least 2R using evidence-backed technical targets. The exact 2R threshold price can be calculated, but it is labeled as a threshold, not a forecast.

Halal + position cap

Compliance must be explicitly confirmed. Existing + proposed effective exposure must stay at or below 5%.

Automated moving-average references use the existing Noweir EMA10/EMA20/EMA50 series. The original framework says 10/20/50-day moving averages without specifying SMA versus EMA, so V4.3 keeps that implementation choice visible.

Setup Pulse

Framework review ≠ order instruction

Setup Queue

Six framework gates + broader V4 gates

Ticker	Setup	Signal	EMA20	EMA50	First 50D	Breakout / RVOL	Stop risk	R/R	Framework	First framework block	V4 state
Run the setup finder.

Boundary

V4.3 does not place orders, does not infer halal status, does not fabricate a target to force 2R, and does not auto-draw discretionary trendlines. Quarter-Kelly remains withheld unless a reviewed win probability exists because the current model-governance layer does not claim a calibrated probability of profit.

Setup Watch & Alerts

Track your existing watchlist against the V4.3 breakout/pullback framework and surface state changes when a setup appears, degrades, reaches 2R, touches the first EMA50, confirms a breakout, or becomes portfolio-blocked.

Watch Settings

V5.3 · browser/client persistence

Watchlist · shared with existing Noweir watchlist

Include open Re-entry Board names

Check due after

Mark stale after

Auto-check due names while dashboard is open

Auto-check max names

Confirmed halal/compliance · SYMBOL=halal

Proposed new position %

Behavioral gate

No setup-watch analysis yet.

What triggers an alert

Attention signal · not expected return

Setup appeared

NONE → BREAKOUT or PULLBACK.

Framework review reached or lost

All six user-framework gates start passing, or a previously passing setup degrades.

First EMA50 touch / breakout retest

Specific setup structures from your documented entry rules.

20-day gate / 2R transition

Reclaim or loss of the 20-day gate, and reward/risk crossing the 2R minimum.

Risk or portfolio block

Structural stop widens above 10%, +4%/parabolic chase state appears, or effective exposure breaches the 5% cap.

Near a level

Info alerts when price is within 1.5% below breakout resistance or within 2% above EMA20/EMA50 support. These are watch conditions, not entries.

This reuses the existing Noweir watchlist philosophy and client-side memory. There is no hidden background trading daemon. Auto-check runs only while the dashboard is open; durable cross-device scheduling would need an explicit server-side scheduler later.

Setup Watch Pulse

Current states + due cadence

Current Watch States

Exact levels, not a BUY score

Ticker	Freshness	Price	Setup	Nearest level	EMA20	EMA50	First 50D	Stop risk	R/R	Framework	V4	Alerts
No setup-watch history yet.

Setup Alert Inbox

Boundary

Alerts tell you what changed and what deserves review. They do not estimate probability of profit, infer halal status, place an order, change allocation, or override the broader V4 event/portfolio gates.

Portfolio Command Center

Base-currency weights, P/L, cash rules, theme exposure, correlation, risk contribution and action queue.

Base currency

Cash

Cash floor %

Max position %

Portfolio Pulse

Risk Contribution

Theme Exposure

Correlation Map · 252D

Action Queue

Upcoming Portfolio Events

Run the portfolio to load upcoming catalysts.

Holdings

Ticker	State	Weight	Unrealized	Halal	GTS	R/R	Catalyst
No portfolio run yet.

Markets & News

Current world-news context from the existing evidence stack. Headlines are research leads, not proof of causation.

Open full news page

Recent Headlines

Research Rules

Direct evidence first

Company filings and direct company reporting outrank thematic headline proximity.

Clusters, not headline counts

Repeated coverage of one story should not masquerade as multiple independent confirmations.

Timing ≠ causation

A nearby headline can explain what to investigate, not prove why a stock moved.

Model Lab

Validation, change control and model disagreement remain visible instead of being hidden behind a score.

Governance Manifest

Regression Suite

Prediction Boundary

Profit probability remains disabled.

Noweir may rank research candidates and report held-out model metrics, but it does not convert those outputs into a calibrated probability of profit without matured labels, out-of-sample validation, calibration and drift monitoring.

Watchlist Longitudinal Memory

Track how research states, evidence, setup quality and conflicts evolve across your watchlist. Priority here means “review this research,” not “buy this stock.”

Watchlist · shared with the original Intelligence workspace

Research profile

Capture due after

Mark stale after

Auto-capture due names while dashboard is open

Auto-capture max names

Automatic capture only runs while this dashboard is open.

Coverage + Freshness

Point-in-time memory, not reconstructed history

Research Priority Queue

Staleness + evidence/state drift + conflicts

Symbol	Freshness	State	Age	Evidence	Trend	Setup	Changes	Priority	20D outcome
Load watchlist memory.

Needs Review

Longitudinal Price

Captured prices only

Points are the market price stored with each point-in-time research capture, not a reconstructed daily price history.

Research Scores Through Time

Evidence · trend · setup

Research State Transitions

Latest transition per watchlist name

Research Change Alert Inbox

System-generated alerts compare point-in-time research snapshots and surface material evidence changes. They complement, but do not replace, the user-defined Smart Alert Center in the full Intelligence workspace.

Open Smart Alert Center

Inbox Summary

Research attention only · no buy/sell signals

Filters

Status

Severity

Alert History

Acknowledgement state is browser-local; alerts can be regenerated from memory.

No alerts loaded yet.

Interpretation Boundary

V2.3 ranks research changes by review priority. The score does not estimate expected return, profit probability, trade attractiveness, or brokerage action. A change in valuation, estimates, technicals, news counts, conflicts, or research state is a reason to inspect the evidence—not a reason by itself to buy or sell.

Thesis Drift + Decision Review

Connect each documented decision to its point-in-time research baseline, show what objectively changed, and keep the thesis verdict explicit and human.

Full Thesis Tracker

Review Summary

No automatic thesis verdicts.

Review Queue

Highest objective drift first

No review queue loaded yet.

Decision Detail

Select an entry.

Select a journal entry to inspect its baseline, current evidence, and prior manual reviews.

Boundary

V2.4 can compare measurable evidence, research-state drift, conflicts, alerts and process documentation. It cannot reliably decide whether your free-text thesis is true or whether an invalidation condition has been met. “Supported / weakened / invalidated” remains your explicit saved judgment, not an automated trading output.

Process Learning + Decision Analytics

Learn from documented decisions without letting small samples, wins, losses, or hindsight silently rewrite the live model.

Process Pulse

Diagnostics only · live decision weight 0

Process Quality Through Time

Documented decisions only

Recurring Process Gaps

Frequency, not severity of future returns

Structured Gate Reliability

Original full Intelligence journal only

Outcome-Bias Checks

Process ≠ outcome

Cohort Diagnostics

Outcome rates suppressed below the minimum sample

Cohort	Group	Decisions	Reviewed outcomes	Favorable rate	Mean reviewed return	Mean process
Run process analytics.

These cohorts are descriptive. They do not promote setups, change risk limits, or become expected-return estimates.

Decision Diagnostics

Latest documented decisions

Date	Symbol	Source	Action	Setup	Process	Outcome	Baseline event	Conflicts	Issues
Run process analytics.

Learning Governance

Position Lifecycle + Exit Learning

Record actual entries, adds, trims, invalidations and exits, then compare the documented plan with what happened afterward. This complements the decision journal; it does not infer brokerage executions.

Record Lifecycle Event

Browser-local execution log

Ticker

Event

Date / time

Price

Quantity / shares

Currency

Execution fees (trade currency)

Trade notional in portfolio base (optional)

Fees in portfolio base (optional)

Setup family

Setup ready?

Planned stop

Planned target

Portfolio value at event (optional)

Portfolio value currency

Active position limit % (optional)

Reason

Notes

ENTRY / ADD need price + quantity for reconstructed average-cost analytics. EXIT / STOP can omit quantity only when the recorded lifecycle makes a full close unambiguous. For V3.2 historical contribution, foreign-currency trades need the actual historical trade notional in the portfolio base currency; current FX is never backfilled. Reconstructed P/L is diagnostic, not broker/tax accounting.

Lifecycle Summary

Run analysis after recording events.

Price + Lifecycle Markers

Markers are the timestamps you recorded. Daily history is provider data; an after-exit rise or fall is descriptive and does not prove the decision was right or wrong.

Open / Reconstructed Positions

Average-cost diagnostic

Execution Timeline

Post-Exit Path

5D · 20D · 63D trading-session follow-through

Date	Symbol	Event	Exit price	5D	20D	63D	20D max rise	20D max drawdown	Re-entry	Review
No exit analysis yet.

Lifecycle Integrity + Framework Checks

No automatic orders

Permanent Re-entry Board

Every recorded SELL, TRIM or STOP becomes a persistent re-entry memory with sale context, post-sale excursion and live V1.2 entry state.

Re-entry Pulse

V4.0 · no automatic orders

Sale / Trim Memory

EMA references use the prior completed daily session

Sale date	Symbol	Type	Execution	Amount	Reason	EMA20	EMA50	Max rise	Max drawdown	Current	V12 zone	Distance	Current R/R	Later re-entry	Re-entry timing
Record or import SELL/TRIM events to build the board.

Interpretation Boundary

Post-sale maximum rise/drawdown is hindsight evidence. Re-entry price improvement compares sale price with a later re-entry price only. It is not market alpha and does not authorize an order.

Execution Quality + Trade Outcome Attribution

Measure documented risk, realized R, MFE/MAE and where execution friction occurred—entry, sizing, risk planning, holding or exit—without converting hindsight into live trading rules.

Execution Pulse

Diagnostics only · live model weight 0

Where the Friction Is

Counts of review diagnostics, not causal blame

Cycle-Level Execution

R / MFE / MAE require documented risk

Open	Symbol	Setup	Status	Planned R/R	Risk coverage	Realized R	MFE	MAE	Exit capture	Post-exit 20D	Diagnostics
Record lifecycle events first.

Recurring Execution Reviews

Frequency only

Setup / Exit Cohorts

Performance suppressed below minimum sample

Cohort	Group	Cycles	Completed	R samples	Positive R rate	Mean R	Median R
No cohort analysis yet.

Interpretation Boundary

Execution analytics use hindsight market paths. They are for process review, not for generating orders, probabilities or automatic strategy changes.

Portfolio Performance + Capital Attribution

Separate current P/L contribution, capital concentration, modeled risk contribution and allocation quality without fabricating a historical portfolio return.

Attribution Pulse

Run Portfolio Command Center first.

Capital Concentration

Current allocation · descriptive only

Position Attribution

Current unrealized P/L contribution + modeled risk share

Symbol	Weight	Market value	Unrealized P/L	Unrealized %	P/L contribution	Risk share	Risk/capital	Risk minus capital	State
Run portfolio attribution.

P/L Dependency

Observed unrealized P/L only

Risk vs Capital

Historical volatility contribution

Allocation Buckets

Halal status + research state

Execution Capital Learning

Uses V2.7 only when lifecycle evidence exists

Attribution Boundaries

Portfolio Ledger + Performance Readiness

Build the dated cash-flow and valuation history required before Noweir calculates real portfolio performance.

Record Ledger Event

Browser-local history

Base currency

Benchmark (optional)

Event

Date / time

External flow amount

Portfolio value

Cash balance

Invested value

Note

Deposits and withdrawals are external cash flows, not trades. Valuation snapshots should use total account value in one explicit base currency. V2.9 never invents historical FX.

Ledger Coverage

Add history, then analyze.

Performance Readiness

Evidence gate before calculation

V2.9 records and audits the required history. V3.0 Historical Performance calculates TWR only when external flows are close enough to recorded valuation boundaries, while MWR/XIRR uses the complete dated cash-flow history.

Ledger Timeline

External flows + valuation snapshots

Date	Event	Amount	Portfolio value	Cash	Invested	Currency	Source / note
No ledger events yet.

Cash-Flow Boundary Audit

Pre/post valuation coverage

Integrity Boundary

Performance stays descriptive and disabled until the data is sufficient. No inferred historical FX, no automatic reallocation, no model-weight changes.

Historical Portfolio Performance

Separate actual investment growth from deposits and withdrawals, then benchmark the active experiment against SPUS with explicit historical FX when the account is CAD.

Portfolio base currency

Benchmark symbol (optional)

Benchmark currency

Cash yield assumption % / year

V3.0 does not infer historical FX. Benchmark-relative performance is calculated only when the benchmark currency and portfolio base currency are the same. Cash drag is a benchmark opportunity-cost estimate, not proof that holding cash was wrong.

Capital Bridge

Uses the V2.9 ledger.

Return Measurement

Contribution-adjusted

Active Experiment vs SPUS

TWR · XIRR · drawdown · volatility · turnover · realized P/L · active relative return · cash drag · re-entry timing

Run once enough broker/valuation history exists. CAD accounts use explicit historical USD/CAD data to convert SPUS into CAD before comparison.

Benchmark Comparison

Auto-adjusted provider history

Cash Allocation Drag

Counterfactual estimate

TWR Subperiods

Adjacent recorded valuations

Start	End	Start value	End value	Deposits	Withdrawals	Subperiod return	Wealth index	Drawdown
Add enough ledger history to calculate TWR.

Data Quality + Interpretation

Performance Timeline + Decision Context

Review monthly, quarterly and YTD performance beside the decisions and execution evidence recorded during the same dates.

Portfolio base currency

Benchmark symbol (optional)

Benchmark currency

Calendar boundary tolerance

Cash yield assumption % / year

Co-occurrence is not causation. V3.1 places documented decisions and execution evidence beside performance windows, but it does not assign portfolio return to a trade without direct attribution evidence. Historical FX remains uninferred.

Timeline Coverage

Uses V2.9 ledger + V2.5/V2.7 process evidence.

Contribution-Neutral Equity Curve

TWR wealth index

Monthly Performance

Calendar-boundary gated

Month	TWR	Benchmark	Relative	Investment gain	Decisions	Process	Lifecycle events	Closed cycles	Closed-cycle R
Add valuation history to build monthly performance.

Quarterly Performance

Same evidence rules

Quarter	TWR	Benchmark	Relative	Investment gain	Decisions	Process	Lifecycle events	Closed cycles	Closed-cycle R
No quarterly period yet.

YTD / Year Performance

Period	TWR	Benchmark	Relative	Investment gain	Decisions
No YTD period yet.

Best / Weakest Months

Descriptive only

Drawdown + Recovery Episodes

Snapshot-resolution

Decision + Execution Context

Not causal P/L attribution

Data Quality + Boundaries

Position-Level Historical Contribution

Record what the account actually held at each valuation, neutralize internal trades, reconcile cash, and only then assign position-level return contribution.

Backfill Position Snapshot

Creates a linked V2.9 total valuation + V3.2 composition snapshot.

Base currency

Date / time

Total portfolio value

Cash balance

Positions · one per line

Optional note

Enter market values already converted into the portfolio base currency. V3.2 does not reconstruct historical FX or old holdings from today's account.

Position Cash Flow

Dividends, distributions, withholding taxes, cash interest.

Symbol

Type

Date / time

Amount in base currency

Note

Contribution Readiness

Need at least two reconciled position snapshots.

Linked TWR Contribution

Only when every consecutive interval passes

Component	Linked contribution	Attributed gain across reconciled intervals	Intervals
Build enough reconciled position history.

Top Contributors / Detractors

Exact linked contribution when available

Position Snapshots

Position + cash must reconcile to total value

Date	Positions	Position value	Cash	Total	Linked V2.9 snapshot	Source
No V3.2 position snapshots yet.

Attribution Intervals

Dollar reconciliation + exact adjacent TWR gate

Interval	Portfolio gain	Attributed before residual	UNALLOCATED	Trades	Cash-flow events	TWR	Status
Need two position snapshots.

Latest Reconciled Interval Components

Internal buys/sells are transfers, not profit

Component	Opening value	Purchases	Sales	Cash flows	Ending value	Attributed gain	TWR contribution	Quantity check
No reconciled interval yet.

Evidence Gaps + Attribution Boundary

Broker CSV Import + Historical Backfill

Stage brokerage exports into the existing V2.6 trade history, V2.9 portfolio ledger and V3.2 position-attribution ledger. Preview does not persist anything.

Import Source

Nothing is written until you commit.

Base currency

Source profile

Date order

CSV file

Source name

CSV text

V3.3 recognizes common broker headers but does not pretend every broker export has the same schema. Foreign historical values need recorded base-currency amounts. Ambiguous rows remain unresolved.

Preview Summary

Review before commit

Recognized Mapping

Header aliases → Noweir fields

Accepted Rows

Staged only

CSV row	Destination	Action	Symbol	Time precision	Review note
Preview a CSV first.

Unresolved / Excluded

Never silently guessed

Commit Plan

Existing IDs are skipped

Commit writes only the normalized accepted bundle into your existing browser ledgers. The raw CSV is not sent to permanent server storage by this import endpoint.

Import Audit

Local provenance history

Historical Reconciliation + Repair Ledger

Find quantity, valuation and attribution gaps after broker import. Candidates have zero accounting effect until confirmed from actual evidence.

Reconciliation Status

Waiting for analysis.

Issue Queue

After confirmed repairs

Severity	Issue	Symbol / period	Detail
Run reconciliation first.

Repair Candidates

Suggestions only

Candidates have zero accounting effect until confirmed. Verify broker or issuer evidence before creating a repair.

Confirm Evidence-Backed Repair

Stored separately from raw broker history.

Repair type

Effective date / time

Symbol

Ratio

Quantity

Quantity delta

Value in base currency

From symbol

To symbol

From quantity

To quantity

Evidence source

Evidence note

Repairs never overwrite the original imported event or position snapshot. V3.2 replays the raw history plus this separate confirmed repair ledger.

Confirmed Repair Ledger

Auditable corrections

Before / After Accounting Effect

Only confirmed repairs may change the after column

Reconciliation Boundary

No automatic corporate actions

A 2:1-looking quantity jump is only a candidate until a real broker or issuer record confirms it.

No invented history

Noweir does not manufacture missing trades, dividends, fees, FX rates, transfer values or symbol conversions to make an interval balance.

Raw evidence remains immutable

Repairs are stored beside imported data, not inside it, so every correction remains traceable.

Durable Portfolio History + Cross-Device Sync

Keep the accounting history local-first while optionally preserving immutable versions in a dedicated cloud database. No automatic conflict winner.

Cloud Connection

Checking V3.5 adapter…

Workspace key · reuses the browser-held V2.1 key

Automatic durable sync

Device ID

The service-role credential stays server-side. V3.5 will not reuse an unrelated Supabase project automatically.

Sync Status

Local-first inventory

Local Durable Inventory

Existing V2.6-V3.4 stores

Record type	Local count	Purpose
Loading local history…

Cross-Device Conflicts

Explicit resolution only

If two devices contain different payloads under the same logical record ID, both immutable versions remain in cloud history. Choose a winner only after checking the actual source evidence.

Sync Audit

Local activity log

Durability Boundary

Raw cloud versions are append-only

A changed record becomes a second version. The old version is not overwritten.

No automatic conflict winner

A newer timestamp or another device does not automatically become truth.

No automatic cloud deletion

Deleting a local row does not silently erase its durable cloud history on every device.

Dedicated database required

The currently connected img-command-center Supabase project is unrelated and is deliberately not reused for Noweir portfolio history.

Portfolio Data Quality + Coverage Control Center

Measure whether the historical evidence is complete enough for each accounting output. Data quality is not investment quality.

Evidence Quality

Waiting for audit.

Downstream Data Gates

Metric-specific readiness

Evidence Issue Queue

Blocking vs review

Status	Issue	Symbol / period	Detail
Run the audit first.

Monthly Evidence Coverage

What was actually recorded each month

Month	Portfolio valuations	External flows	Position snapshots	Trades	Position cash flows	Reconciliation issues
Run the audit first.

Control Boundary

Data quality is not investment quality

The V3.6 score describes evidence coverage and accounting integrity only. It is not a forecast, a trade signal, or an expected-return score.

Unsupported metrics stay gated

Missing cash-flow boundaries, unresolved quantity gaps, rejected records or cloud conflicts can block a metric rather than being silently imputed.

Research remains available

A failed historical-accounting gate does not stop manual company research. It only prevents unsupported historical portfolio claims from being treated as complete.

Evidence-Gated Analytics Enforcement

V3.7 turns V3.6 readiness into display rules. Unsupported derived metrics are withheld, while raw evidence and manual research remain available.

Enforcement Status

Waiting for V3.6 evidence audit.

Analytics Surface Gates

AVAILABLE vs WITHHELD

Enforcement Boundary

WITHHELD · V3.7 evidence gate

When a V3.6 gate is not ready, the unsupported derived number is removed from the display-safe copy rather than shown with a warning beside it.

Raw engine result remains auditable

V3.0, V3.1 and V3.2 raw outputs are preserved separately for diagnostics and regression review.

Independent metrics remain independent

For example, MWR/XIRR may remain visible while TWR is withheld if only the TWR boundary evidence fails.

Manual research stays available

Historical accounting gates never block company research, news review or watchlist work.

Research Memory

Point-in-time snapshots preserve what Noweir actually knew at capture time instead of reconstructing the past with today's data.

Capture Snapshot

Stored in IndexedDB on this browser.

Symbol

Profile

Tags · comma separated

Automatic research snapshots

Optional note / thesis context

V2.1 remains local-first. Automatic captures are compacted from the exact research packet you just ran, so the system does not need to re-query fundamentals simply to remember them.

Memory Status

Optional Cloud Memory

Checking cloud adapter…

Workspace key · browser-held secret, never the Supabase service-role key

Auto cloud sync after capture

The service-role credential stays server-side. A dedicated Supabase/Postgres database still has to be connected before cross-device sync can become active.

Snapshot Timeline

No snapshots captured yet.

Evidence Delta

Compare two point-in-time captures of the same symbol.

No comparison yet.

Experiment Lab

Test whether point-in-time research features line up with matured future outcomes. Diagnostics remain shadow-only and never auto-reweight production decisions.

Run Memory Diagnostic

Uses captured + matured snapshots

Feature

Outcome horizon

Optional symbol filter

A diagnostic becoming available does not make it production-ready. V2.0 keeps live decision weight at 0 and blocks automatic promotion.

Diagnostic Result

Waiting for matured observations.

Research Experiment Registry

Inherited V1.5 research ideas + V2.0 memory governance

Research Journal

Separate decision quality from outcome quality. Stored locally in this browser.

Add Journal Entry

Ticker

Action

Outcome

Thesis / evidence

What would falsify this?

Expected catalysts / checkpoints

Red-team case / strongest counterargument

Timeline

All Tools

Nothing from the original Intelligence workspace was removed. This page is the organized index back to every advanced tool.

Open full Intelligence workspace Market Terminal World News

V4.0 connects research quality to portfolio fit, execution memory and benchmarked results.

Noweir keeps every prior research, memory, lifecycle and accounting layer, then adds a permanent Re-entry Board, portfolio-aware discovery, explicit decision gates, broker-history bridging, Canadian primary-document evidence and a benchmarked active-account experiment. NO TRADE is a valid outcome.

Decision Workflow

One-Click Stock Analysis · Noweir Decision Engine · Goal Path Planner · Model Entry Intelligence · Historical Calibration · Dividend / ex-dividend analysis

FULL INTELLIGENCE

Integrity + Learning

10-gate workflow · Personal Learning V2 · adaptive weights shadow mode · V1.2 integrity · cross-layer conflicts · news quality · SEC intelligence · AI research packet

FULL INTELLIGENCE

Discovery

Theme Discovery + Asymmetry Lab · Advanced Screener · institutional workflow · peer comps · valuation history · expectation intelligence · chronological 21D/63D model validation

V1.5 + FULL INTELLIGENCE

Evidence

Why is this moving? · SEC Filing Monitor · Market Memory · Evidence Graph + Conflict Lens · Evidence Delta · Watchlist Impact Radar

FULL INTELLIGENCE

Watchlist + Attention

Watchlist Workspace · Research Attention Queue · Watchlist Event Radar · Smart Alert Center · Alert Audit Trail · V2.2 longitudinal point-in-time memory · staleness detection · state transitions · due-capture queue · V2.3 research-change inbox

V2.3 ALERTS

Portfolio Risk

Portfolio Decision Packet · Portfolio Concentration Detail · ETF Look-Through + Theme Collision Radar · Portfolio Shock Simulator · model disagreement

V1.3 / V1.4

Process + Storage

Decision Journal + Thesis Tracker · Workspace Backup · local portfolio save · browser-local journal · V2.0/2.1 Research Memory · V2.4 thesis drift · V2.5 process learning · V2.6 actual position lifecycle · V2.7 execution quality · V2.8 portfolio performance + capital attribution · V2.9 dated portfolio ledger + performance readiness · V3.0 TWR/MWR historical performance · V3.1 monthly/quarterly/YTD timeline + decision context · V3.2 position snapshots + trade-neutralized contribution · V3.3 reviewed broker CSV import + historical backfill · V3.4 reconciliation queue + explicit repair ledger · V3.5 durable immutable-version portfolio history + cross-device conflict handling · V3.6 portfolio data-quality score + evidence readiness gates · V3.7 display-safe analytics enforcement · V3.8 unified visual system + responsive/runtime smoke checks · V3.9 workspace navigation + production hardening · V4.0 permanent re-entry memory + disciplined discovery + broker/benchmark evidence bridge

V4.0 SYSTEM

Charts + Visual Analytics

Market Terminal · candlesticks · EMA10/20/50 · SMA200 · volume · RSI · MACD · watchlist scanner · multi-chart live board · normalized comparisons · benchmark views · valuation/growth maps · sector heatmaps · catalyst calendar · event markers · historical earnings/filing reaction views

V1.9 EVENTS

News + Market Context

World News dashboard · market pulse · themes · source mix · current company news and world-news mapping through the research stack

NEWS

System

One place to inspect active versions, API health and the unified research boundary.

API Health

Architecture

Single-name research

V1.1 + V1.2.1 evidence, integrity, news, SEC, Market Memory and historical calibration.

Portfolio intelligence

V1.3 portfolio decision layer + V1.4 risk/model governance.

Discovery

V1.5 factor, valuation, expectation, chronological-model and portfolio-fit research.

Event intelligence

V1.9 connects earnings, SEC filings and news evidence to descriptive historical reactions without asserting causation.

Research memory

V2.0 captures point-in-time snapshots, matures realized outcomes and evaluates diagnostics at zero live decision weight. V2.1 adds automatic compact capture and optional cloud sync. V2.2 organizes that memory longitudinally across the shared watchlist and surfaces stale or materially changed research. V2.3 turns material point-in-time changes into a persistent research alert inbox with acknowledgement state. V2.4 links documented journal decisions to exact or nearest point-in-time baselines, separates process quality from outcome quality, and requires a manual thesis judgment. V2.5 aggregates recurring process gaps, gate discipline, setup/context cohorts and early-vs-recent documentation quality while keeping the findings shadow-only. V2.6 records actual position events, reconstructs average-cost lifecycle diagnostics, tracks time after recorded invalidation, and measures post-exit 5D/20D/63D price paths. V2.7 adds documented-risk R-multiples, MFE/MAE, stop/target adherence review, entry hindsight diagnostics and attribution across entry, sizing, risk plan, holding and exit. V2.8 explains current P/L, capital concentration and risk-vs-capital. V2.9 adds dated external cash flows, portfolio valuations and cash-history readiness so later performance metrics are based on real history. V3.0 calculates contribution-adjusted capital growth, TWR, MWR/XIRR, same-currency benchmark-relative performance, snapshot-series drawdown and a clearly labeled benchmark cash-allocation drag estimate. V3.1 organizes valid returns into monthly, quarterly and YTD windows, builds a contribution-neutral wealth curve, identifies drawdown/recovery episodes, and places V2.5/V2.7 process evidence beside the same dates without claiming causation. V3.2 adds actual point-in-time position composition, internal-trade neutralization, quantity-path checks, cash/position cash-flow reconciliation, exact adjacent-subperiod TWR contribution and visible UNALLOCATED residuals rather than forcing missing evidence into a holding. V3.3 stages broker CSV data into those same ledgers with header mapping, source hashing, unresolved-row review, duplicate-safe browser commit and no historical FX guessing. V3.4 audits those imported histories for quantity, valuation, cash-flow and attribution mismatches, surfaces corporate-action candidates without applying them, and allows only explicit evidence-backed repairs in a separate immutable repair ledger. V3.5 adds optional local-first durable portfolio-history sync using immutable content-addressed cloud versions, device provenance, explicit conflict resolution and a dedicated database boundary. V3.6 measures evidence coverage and accounting integrity across those histories, shows month-by-month gaps, and explicitly gates unsupported MWR, TWR and position-contribution outputs without turning data completeness into an investment score. V3.7 enforces those gates in the display path: unsupported derived metrics are withheld, while raw engine output remains separate for audit and manual research remains available.

V4.0 decision discipline

The dashboard unifies re-entry memory, named discovery modes, explicit gates, portfolio correlation/theme/ETF overlap, broker history and SPUS benchmarking. It still does not place orders or invent compliance evidence.