Data Methodology

AfriTerminal aggregates publicly available financial data from multiple sources and presents it in a unified interface for African capital markets professionals. All data is sourced from official exchange feeds, government publications, and reputable public APIs. No proprietary or licensed data is used without permission.

Data is retrieved by automated Python scrapers running as GitHub Actions workflows. Outputs are committed as static JSON and CSV files served via GitHub Pages. There is no real-time streaming infrastructure - all data represents end-of-day or intra-day snapshots with the delays noted in the table below.

The Market Context card shown on individual stock detail views and the Intelligence tab is computed by bayesian_signals.py. It has three parts, and each is scoped to exactly what it can honestly claim.

1. Regime classifier. Deterministic thresholds on real MPR, inflation, external reserves, and Brent crude inputs classify the current macro regime (e.g. restrictive/accommodative rates, high/low inflation, oil-positive/negative) and map it to a plain-English sector implication.

2. Composite score (descriptive, not predictive). A base rate of up-days over the last 30 trading days (Laplace-smoothed), adjusted by real filing activity, NGN/USD correlation, S&P 500 correlation, and market breadth. Composite score = base_rate × filing × FX × global × breadth × 100. This is a summary statistic describing recent history, shown as "highest score," never as "validated" or "significant" on its own.

3. Directional significance (a real hypothesis test, added Phase 2, August 2026). A stock's daily direction is tested against the null hypothesis that it is no better than a coin flip, using an exact two-tailed binomial test, with a real Benjamini-Hochberg FDR correction at q=0.10 across every ticker tested that run. A zero-change day (no real trade printed a new price) is excluded from the count entirely rather than treated as a "down" day - a stock needs at least 15 real (nonzero-change) trading days in its trailing 30 before a test is attempted at all. A stock only shows a "Directionally significant" badge when it actually passed this test; every other stock shows no badge, not a false negative.

Every score on this card - composite score, regime label, and significance result alike - carries a confidence label and the number of trading days it was computed from. A score is never shown without that context. None of this is investment advice, and no part of it constitutes a buy, sell, or hold signal.

AfriTerminal publishes two families of proprietary sector indices, Phase 1A (launched May 2026):

• NGX Sector Indices (ATBI-NG, ATCI-NG, ATOI-NG, ATTI-NG, ATCGI-NG, ATII-NG, ATAGI-NG): 7 equal-weighted sector indices across NGX-listed equities. Base level: 1,000 on launch date.
• Global Sector Benchmarks (ATGBI, ATGMI, ATGEI, ATGTI, ATGCI, ATGII, ATGAI): 7 market-cap weighted indices tracking the same sectors' largest global companies via Yahoo Finance.

Index level formula: Level(t) = Level(t−1) × Σ(weight_i × Price_i(t) / Price_i(t−1)). Weight constraints: maximum 35% per single constituent; minimum 2%; top-3 combined maximum 65%. Applied iteratively (up to 10 passes).

Phase 1A uses equal weighting for NGX indices because reliable free-float adjusted market capitalisation data is not yet available via the public NGX feed. Market-cap weighting with free-float adjustment is planned for Phase 2A, subject to an NGX data partnership.

AfriTerminal maintains a structured database of NGX company fundamentals (profit after tax, EPS, revenue, dividends) extracted from official filings. Sources are strictly limited to the filed documents themselves: annual reports and results announcements from NGX Group's document library. News articles and third-party aggregators are never used as a source for a financial figure.

Extraction verification. Every extracted figure is verified against the literal text printed in the source PDF before it can be stored. Figures pass through relational validation checks (profit against revenue, EPS arithmetic, year-on-year magnitude bands, unit and scale reconciliation between kobo and naira). Each figure carries an extraction confidence level (high, medium, or low). Figures below the confidence threshold are withheld from display and routed to manual review; they are never published. Where a displayed figure carries less than high confidence, a visible badge marks it.

FY-only standard. All displayed comparisons (financial history cards, peer comparisons, the relative value screener, implied valuations) use full-year (FY) figures exclusively. Interim periods (quarterly, half-year) are stored but never mixed into displayed comparisons, because NGX interim reporting is too inconsistently labelled to support reliable trailing-twelve-month construction.

Peer comparisons group companies by NGX subsector (falling back to sector where a subsector is thin) and show each company's fundamentals side by side. Implied valuation multiplies a company's statistically projected EPS (bear, base, and bull scenarios derived from its own filing history) by the trailing P/E median of its sector peers. Peers with implausible multiples (outside a 0.3x to 60x band) are excluded from the median. Where the resulting figure sits far from the current market price (beyond 3x in either direction), it is still shown, marked with an explicit out-of-band note and capped at Low confidence, rather than hidden.

Where you record positions in the Portfolio view, AfriTerminal computes four analytics directly from your recorded positions and its own live/historical market data - never a fabricated or estimated figure:

• Concentration: largest single position and top-3 combined, as a percentage of total recorded portfolio value at live prices.
• FX sensitivity: a value-weighted average of each holding's own NGN correlation (from the Market Context Layer, Section 5). If not every holding has an FX signal yet, the card reports partial coverage explicitly (e.g. "2 of 3 positions") rather than silently excluding the rest.
• Benchmark-relative return: each position's return since it was added is compared against the NGX All-Share Index's return over that same holding period (not one shared window - different positions have different start dates), then rolled up into a value-weighted portfolio-level figure. The NGX ASI series is limited to its trailing 30 recorded data points; a position held longer than that has no matching starting point yet and is reported as "not yet comparable," not silently dropped.
• Correlation matrix: pairwise Pearson correlation of daily price changes across your holdings, computed only where a pair has at least 15 overlapping trading days of price history - a thinner pair shows "n/a" rather than a number computed on too little data.

Portfolio positions are personal record-keeping data you enter yourself; see the Privacy Policy for how they are stored, and Section 9 (Free Access & Future Subscriptions) of the Terms of Use for the record that this is not a brokerage or execution service.

For methodology questions, data corrections, or institutional access enquiries:

info@afriterminal.com