← Back to Quintarth

Disclosures & Methodology

Last Updated: May 8, 2026

1. Nature of Services

Quintessentia Network Inc. (operating as “Quintarthai”) is a Canadian-Controlled Private Corporation incorporated in Ontario and headquartered in Niagara Falls. We operate a financial information and research platform that provides algorithmically generated analysis of publicly available market data. See /about for full operating-entity details.

We are NOT: a registered investment advisor (SEC/OSC), a broker-dealer (SEC/FINRA), a securities exchange or ATS, an investment dealer (IIROC), an exempt market dealer (OSC/CSA), or a crypto trading platform (CSA NI 21-329).

All analysis is impersonal, uniform, and of general circulation, consistent with the publisher's exemption under SEC Interpretive Release IA-1092 and comparable Canadian securities law provisions.

2. Data Sources & Freshness

Data TypeSourceFreshness
Equity PricesYahoo Finance (yfinance)May be delayed up to 15 minutes
SEC FilingsEDGAR EFTS API (Public)Real-time (as filed)
Canadian FilingsSEDAR+ / SEDI (Curated)Daily updates
Exchange RatesYahoo Finance / Bank of CanadaDaily
AI Analysis (LLM)Groq (US-hosted, primary inference) · Anthropic (US-hosted, fallback for tool-call paths) · Ollama (Canadian-hosted, air-gap mode for sensitive queries)On-demand generation
AI grounding (RAG)Qdrant 1.9 vector database, quintarth-regulations-v1 collection, 768-dim Ollama nomic-embed-text embeddings — all Canadian-hostedRefreshed on corpus update
Sentiment classifierFinancial-domain NLP model (transformer encoder) · Canadian-hostedPer-article on ingest

IMPORTANT: Market data displayed may be delayed. Delayed data is NOT suitable for real-time trading decisions.

Cross-border AI processing notice (Quebec users — Law 25 / PIPEDA): Some queries are processed by US-hosted LLM providers (Groq Inc., Anthropic PBC). Query text and the retrieved RAG context are transmitted to those providers under their published data-processing agreements; we do not opt in to any “use customer data to train your model” features. Quebec users (and any user concerned about cross-border processing) can switch Quinn to Air-Gap Mode in their account settings, which routes all inference through the Canadian-hosted Ollama Llama 3 8B model. Full per-engine architecture, data flows, and limitations: /legal/ai-transparency.html.

3. Scoring Methodologies

3.1 Delisting Risk Score (0-100)

Composite weighted score derived from: bid price compliance (20%), market cap (15%), float concentration (10%), promotional NLP detection (15%), audit opinion risk (15%), filing timeliness (10%), shell/reverse merger flags (10%), revenue trajectory (5%). A high risk score does NOT mean a company will be delisted.

3.2 Cross-Border Arbitrage Spread

Mathematical computation: Spread = (CA Price × FX Rate - US Price) / US Price × 100. Estimated transaction costs of ~1.5% are deducted. Spread Direction indicators represent observed price differentials — NOT buy/sell recommendations.

3.3 Insider Activity Analysis

Insider Buy/Sell ratios calculated from SEC Form 4 filings (US) and SEDI reports (Canada). Signal badges are statistical classifications. Insider purchasing does not guarantee positive future returns.

3.4 Stock Report Card (Composite Intelligence Score)

The Composite Intelligence Score (0-100) aggregates five sub-scores:

Sub-ScoreWeightRangeSource
Risk (inverted)20%0-100Delisting Risk Model
Fundamental30%0-100P/E, P/B, margins, ROE, cash flow
Momentum25%0-100Price change + 52-week position
Volume15%0-100Current vs average volume ratio
Insider Activity10%0-100SEC Form 4 buy/sell cluster analysis

Confidence Band: ±3 points (large-cap), ±5 (mid-cap), ±8 (small-cap), ±12 (micro-cap), ±18 (nano-cap). Wider bands indicate less reliable scoring.

Interpretation: 80-100 = Strong, 60-79 = Moderate, 40-59 = Neutral, 20-39 = Weak, 0-19 = Distressed. These labels describe signal strength — NOT buy/sell recommendations.

Non-GAAP Notice: The Composite Intelligence Score is a proprietary non-GAAP metric. It is a mathematical model output, not a statement of fact about any issuer's value or prospects.

4. Non-GAAP Metrics Notice

Certain metrics including Cash Burn Rate, Composite Risk Score, and Net Spread are non-GAAP derived metrics not defined under IFRS or US GAAP. These may not be comparable to similarly titled measures used by other platforms. See individual module descriptions for calculation definitions.

3.5 Technical Score Methodology

Technical indicators are computed using standard, widely-published formulas applied to publicly available price and volume data:

IndicatorParametersSignal Logic
SMA 50/20050, 200 dayGolden Cross (bullish) / Death Cross (bearish)
EMA 12/2612, 26 dayCrossover direction
RSI (14)14-period>70 overbought, <30 oversold
MACD12, 26, 9Signal line crossover + histogram
Bollinger Bands20-period, 2σBand position + volume context
OBVCumulativePrice/volume divergence

Weighting: Trend 40%, Momentum 30%, Volume 20%, Volatility 10%. Scores are mathematical outputs — NOT buy/sell recommendations. Uses closing price data which may be delayed up to 15 minutes.

3.6 Sentiment Analysis Methodology

News sentiment is computed using financial-domain NLP models applied to publicly available news and regulatory filings.

Source Weighting: SEC/SEDAR+ filings (highest), major financial newswires (high), industry publications (moderate), general news (lower), issuer press releases (lowest — inherent positive bias).

Temporal Decay: Exponential decay with half-lives from 3 days (breaking news) to 60 days (NI 43-101 exploration results).

Output: Entity-level sentiment score from -1.00 to +1.00. Scores are automated NLP outputs — NOT analyst opinions or research reports under FINRA Rule 2241 or IIROC Rule 3400.

3.8 Analyst Consensus & Third-Party Data Disclosure

Source. Quintarthai displays sell-side analyst consensus data — including mean / median / high / low 12-month price targets, analyst count, and recommendation labels (e.g. "Strong Buy", "Buy", "Hold") — sourced from the public yfinance library, which aggregates from Yahoo Finance's public API. Quintarthai does not independently survey analysts, does not generate research, and does not receive direct feeds from Refinitiv, FactSet, S&P Global, or other commercial data providers. All analyst data is third-party redistribution under Yahoo Finance's general-public terms.

Not a research report. Quintarthai is not a broker-dealer and is not registered with FINRA, IIROC, the SEC, or the OSC. The display of analyst consensus data on this platform is informational redistribution of publicly available aggregated data. It does not constitute:

Publisher's exemption posture. Quintarthai is a bona fide publisher of general-circulation financial information. Identical data is shown to every user querying a given ticker; no personalized portfolio recommendations are made. We rely on the SEC's IA-1092 publisher's exemption (and Canadian equivalents under NI 31-103) for the impersonal, general-circulation distribution of analyst consensus data alongside our own multi-factor models.

Limitations of analyst consensus data.

What Quintarthai adds. We display analyst consensus alongside our own independent multi-factor synthesis: 5-Year DCF, Forward P/E Benchmark, PEG=1.0 (Lynch), and Graham Number methods (see §3.7). Wide disagreement between analyst consensus and our models — or wide spread among analysts — is itself decision-support signal, not noise. We never aggregate or weight our own score against the analyst label; the two views are presented independently.

Removal. If you are a regulator, registered firm, or licensed broker-dealer who believes display of any specific analyst data on this platform is in conflict with your firm's licensing terms or regulatory requirements, please contact hello@quintarthai.com with the ticker and field name. We will remove the specific data within 1 business day pending review.

3.7 Fair Value Estimation Methodology

The Report Card displays a Blended Fair Value derived from up to 5 independent methods. Each method is computed independently and the blended value is the median of applicable methods (Graham Number excluded from the blend when more than 2 other methods apply, since it systematically lowballs growth stocks). All methods are non-GAAP model outputs and not investment recommendations.

Method 1: Analyst Consensus (mean)

Mean of sell-side analyst 12-month price targets, sourced from yfinance (which aggregates data from Refinitiv / FactSet feeds). Confidence is high (≥15 analysts), medium (5–14), or low (<5). We display the raw consensus — Quintarthai does not weight or curate the analyst list.

Method 2: Forward P/E Benchmark

Forward EPS × sector-typical forward P/E. Sector multiples are 10-year long-run medians per the S&P sector composites table:

Limitations: Sector medians smooth across very different sub-industries (semiconductor capex cycle vs. mature SaaS) and don't adjust for individual-company growth/margin profiles. Use as a directional anchor, not a precise target.

Method 3: PEG=1.0 (Lynch)

Forward EPS × earnings growth %. Based on Peter Lynch's heuristic that a fairly-valued growth stock has PEG ratio = 1.0. Excluded when earnings growth is negative, zero, or implausibly high (>50% YoY, often an extrapolation artifact).

Limitations: Highly sensitive to growth assumption. A small change in forecast growth produces large changes in implied fair value. Most useful for moderate-growth (10–35%) established companies.

Method 4: 5-Year Discounted Cash Flow (DCF)

Free cash flow per share projected over 5 years, plus terminal value via Gordon growth perpetuity.

Limitations: Single-rate discount and uniform growth ignore sector-specific capital structure. Companies with negative FCF (early-stage, capital-intensive, or one-time impairments) are skipped. The growth cap means DCF systematically underweights hyper-growth names; this is by design — a 50% growth rate extrapolated 5 years compounds to absurd valuations.

Method 5: Graham Number (informational only)

√(22.5 × trailing EPS × book value per share). Benjamin Graham's deep-value formula. Almost no growth or tech stock passes this test — it's presented as a "deep value floor" benchmark, not a target. Excluded from the Blended Fair Value when other methods apply (it would skew the median).

Blended Fair Value & Verdict

The displayed "Blended Fair Value" is the median of applicable methods (typically 3–4 of the 5). Median over mean prevents one outlier (e.g., Graham underweighting a growth name) from dominating the verdict. The range shown alongside (low–high) reflects the cross-method dispersion — wide ranges indicate genuine valuation disagreement, narrow ranges indicate consensus across methods.

The Verdict label compares current market price against the blended fair value with a ±10% band:

What Fair Value is NOT

5. Mining & Resource Disclosures

Mineral resource estimates referenced on this platform are derived from NI 43-101 or S-K 1300 technical reports filed by issuers. Quintarthai has NOT independently verified any resource estimates. Resource categories (Measured, Indicated, Inferred) are as classified by the Qualified Person. Inferred resources are speculative and are always displayed separately.

6. AI-Generated Content

Quinn and blog content are generated by language models using retrieval-augmented generation (RAG) from publicly available filings and market data. AI-generated content may contain errors, omissions, or outdated information. It is NOT investment advice and NOT a research report.

7. Limitations & Risks

8. Conflicts of Interest

Quintarthai does not receive compensation from any issuer for coverage or scoring. Quintarthai does not receive commissions or referral fees from any broker-dealer.

9. Regulatory Contacts

USA: SEC — sec.gov/investor | FINRA — finra.org/investors
Canada: OSC — osc.ca | Provincial regulators — securities-administrators.ca | Privacy Commissioner — priv.gc.ca

© 2024-2026 Quintessentia Network Inc. (operating as Quintarthai). All rights reserved.