S-REIT Metrics Extractor: DPU, Gearing, ICR, Occupancy and WALE from Singapore REIT PDFs

Turn Singapore REIT (S-REIT) results press releases, financial statements and results presentations into clean, structured JSON. Every number comes with the page number and the exact text it came from, so you can check it in seconds.

Built for investors, analysts, fintech builders and AI agents who need S-REIT fundamentals the day results come out, not weeks later when a screener updates.

What it extracts

Field	Example	Notes
dpu_cents	11.51	Distribution per unit (or per stapled security), in cents. unit says SGD cents or USD cents
aggregate_leverage_pct	41.4	Aggregate leverage / gearing
interest_coverage_x	2.6	Interest coverage ratio (ICR)
occupancy_pct	91.9	Committed / portfolio occupancy
wale_years	4.8	Weighted average lease expiry
avg_cost_of_debt_pct	3.52	Average / all-in cost of debt
fixed_rate_debt_pct	75.5	Share of debt at fixed or hedged rates
nav_per_unit	1.73	Net asset value per unit. unit says SGD or USD

It also returns the REIT name, the reporting period (e.g. 1H 2026, 1Q FY26/27), the document's reporting_currency and its document_type (report or slides).

Each metric includes:

• page: the page it was found on
• snippet: the source text around the number. On slides, a KPI tile (a big number with its label above or below it) is read by position, and its snippet starts with [tile] followed by the label and the value, e.g. [tile] Aggregate Leverage 44.9%.
• currency: SGD or USD for DPU and NAV; null for ratios
• confidence: high, medium or low. Values from slide decks are never high. Occupancy whose scope is unclear is low, because occupancy is often quoted for one property without saying so.
• scope: portfolio, segment_or_asset or unclear. REITs often quote occupancy or WALE for one mall, one country or their top 10 tenants, and slides show market figures (e.g. "Submarket Occupancy"). Those are flagged as segment_or_asset instead of being passed off as the portfolio figure, and a portfolio-level figure is always preferred when the document has one.

Which PDF to use

Best: the results press release (news release / media release) or the financial statements. These are written in full sentences and tables, and extraction is most reliable on them.

Presentation slides and business updates also work, and are free for now. Slide layouts (KPI tiles, multi-column tables, charts) are harder to read, so expect more gaps and check medium-confidence values against the quoted page. Slide decks stay free until their extraction accuracy matches that on press releases.

US-dollar REITs (e.g. Manulife US REIT, Prime US REIT) are supported: their DPU and NAV come back labelled USD. DPU and NAV from REITs that report in other currencies (EUR, GBP) are not returned yet, and the row carries a warning instead.

Example output

{
  "source": "https://links.sgx.com/.../897539_ESR-REIT%20-%201H2026%20Results%20News%20Release.pdf",
  "reit_name": "ESR-REIT",
  "period": "1H 2026",
  "reporting_currency": "SGD",
  "document_type": "report",
  "status": "ok",
  "billed": true,
  "metrics_found": 7,
  "metrics": {
    "dpu_cents": {
      "value": 11.51, "unit": "SGD cents", "currency": "SGD", "page": 1,
      "snippet": "…with Total DPU growing 2.4% y-o-y to 11.510 Singapore cents…",
      "confidence": "high", "scope": "portfolio"
    },
    "aggregate_leverage_pct": {
      "value": 41.4, "unit": "%", "currency": null, "page": 1,
      "snippet": "…Gearing at 41.4%; expected to reduce to 39.9% after rede…",
      "confidence": "high", "scope": "portfolio"
    },
    "fixed_rate_debt_pct": {
      "value": 75.5, "unit": "%", "currency": null, "page": 5,
      "snippet": "As at 30 June 2026, approximately 75.5% of ESR-REIT's debt is on fixed interest rates…",
      "confidence": "medium", "scope": "portfolio"
    }
  }
}

How to use

1. Paste one or more public PDF URLs into PDF documents. Optionally add reitName per document (slide decks often show the name only as a logo).
2. Run. You get one dataset row per document.
3. Export to JSON, CSV or Excel, or call it from the API, Make, Zapier or an AI agent through Apify's MCP server.

Tip: pair it with an SGX announcements Actor. Pull this week's REIT announcements, keep the results press release links, and pass them here.

Pricing

Pay per document, $0.03 for each press release or financial statement with at least 3 metrics found (values flagged as one segment or asset, or with low confidence, don't count).

Free:

• Slide decks and business updates: free for now. For best results, pass the press release or financial statements.
• Documents with fewer than 3 metrics, documents that aren't results announcements, scanned PDFs and downloads that fail.

Every row says whether it was charged: billed is true or false, and free rows carry a billing_note saying why.

If you set a maximum cost per run, the run stops before the next document once another charge wouldn't fit (free slide decks don't use up the limit), and the status message tells you how many documents weren't started.

Limitations

• Needs a text-based PDF. Scanned image PDFs aren't supported yet.
• Only the first 80 pages of a PDF are read.
• Tuned for S-REIT and Singapore business-trust reporting language. Other markets may work partially.
• When a document reports the same metric for several periods, the extractor aims for the document's own period, but check the quote.
• The number is a starting point, not investment advice. Check anything you act on against the quoted page.

Not affiliated with SGX, MAS or any REIT manager.