Invest Vietnam
Search a listed company and trace every material figure to the document, page, table and cell that supports it. Opportunity discovery applies deterministic, versioned screening conditions over persisted evidence — and states explicitly when evidence is missing instead of inventing a value.
Counts are read from the canonical universe and persisted screening snapshots. Fresh market data is recomputed at request time from the stored price timestamp under the canonical 72-hour policy; a stored freshness label is never trusted on its own. A company without a screening snapshot counts as missing evidence, not as a zero-signal company.
Opportunity discovery
Every category is a versioned, fixed condition over persisted market and financial evidence. Selecting one opens the same canonical set of matched companies that the public opportunity discovery service returns.
Current signals
Up to five companies with the most matched signal categories, in the public opportunity discovery service's canonical order (most matched categories, then ticker). Missing evidence fails closed: a refused or unavailable input never becomes a match.
Loading current opportunity signals…
Signals use fixed thresholds over persisted evidence. Check again after the next evidence refresh. Nothing is substituted or estimated.
No substitute results are shown.
These are deterministic screening conditions over persisted evidence, not investment recommendations, forecasts, ratings or advice. They are not personalized and never use portfolio, alert or account data.
Certified publications
Invest Vietnam research publications whose latest company report has passed internal certification, most recently approved first. This is not general market news and does not represent every market change.
Loading certified research…
Only reports that pass internal certification appear here, and each row links to the company research surface. Nothing is substituted or estimated.
No substitute items are shown.
Each opportunity category is a fixed, versioned rule over persisted evidence. The public service evaluates the same stored inputs for every caller, returns companies in a canonical order, and never re-sorts, scores or recommends on the client. Missing evidence fails closed rather than fabricating a match.
Extraction is a candidate; typed, validated evidence is the source of truth. Figures carry their document, page, table and cell so a conclusion can be traced, and every screening snapshot records the versions and timestamps used to build it.
Valuation comparisons are self-relative where the evidence allows it. Historical valuation availability is modeled from reporting-lag assumptions: modeled evidence is labelled, may be permitted only where a rule says so, and is never described as point-in-time certified.
A research report appears in the certified feed only after the repository's internal review records approval. Certification describes publication status and provenance; it is not a rating, a forecast or advice.
Missing, stale or inapplicable inputs are refused with a reason code instead of becoming a zero. UNKNOWN values are never converted to zero, and a refused condition is never counted as a matched signal.
Market recency is recomputed at request time from the stored price timestamp under the canonical 72-hour current and 168-hour aging policy. Fundamental and research freshness follow their own canonical windows and are shown alongside every signal.