Methodology & Sources — How GroundVision Sources Every S-REIT Figure

We built this the way we build research for clients: every number traces to a source, the pipeline is deterministic, and nothing is guessed. Here's exactly how the data gets here.

The principle: no unsourced numbers

The dataset behind this site is generated by a deterministic pipeline into a database where every datapoint carries its source URL, the date it is valid for, and a status flag (market data, computed, or under-review). The site generator refuses to render a metric that isn't in the database — which is the structural reason you won't find a fabricated figure here.

Two data tracks

1 · Market & valuation data — automated

Price, market cap, NAV per unit, distribution yield, distribution history and the 52-week range are collected programmatically from Yahoo Finance (SGX <code>.SI symbols). This data is delayed / end-of-day, so every value is stamped with an as-of date. We compute the ratios ourselves — yield, P/NAV, premium/discount — so the derivation is transparent and reproducible, and we cross-check the headline yield against an independent trailing-distributions calculation. Where the two diverge (usually a recent listing or a special distribution), we flag the figure as indicative rather than presenting it as verified.

2 · Operational & balance-sheet data — from filings

Gearing, interest coverage, cost and tenor of debt, the debt-maturity ladder, WALE, occupancy and tenant concentration are not available from any free market feed. They live inside each trust's SGX results announcements and business updates. We extract them from those filings for the deal brief and cite the document each figure came from.

Why some sources aren't used as the source of record

Free aggregator tables exist, but they publish derived numbers without per-figure citations, so we use them only as cross-checks — never as the authoritative value. The SGX JSON price endpoint is more authoritative than Yahoo, but its bot-protection blocks server-side collection; the pipeline is built to switch to it when run from an allowlisted environment.

What this demonstrates

This entire property — the scraper, the provenance database and ~50 source-linked pages — is generated by GroundVision's agentic research tooling. It's the same machinery we point at a client's portfolio, deal pipeline or target market to produce source-backed briefs at speed. If you'd like that pointed at your data, talk to us.

Questions

Is the market data live?
It's delayed / end-of-day, sourced from Yahoo Finance, and stamped with an as-of date. For real-time, licensed feeds we integrate SGX's commercial data products on client engagements.
How often is it refreshed?
Market data refreshes on a schedule (the pipeline re-runs and re-generates the site); operational metrics refresh when a new SGX filing is published, each tagged with its reporting period.