Capital: deals, debt and who reports them

The money side of the database: 508 dated sales, listings, auctions, loans, special-servicing and foreclosure events on 351 sites from public reporting, plus a new layer that asks SEC EDGAR which commercial-mortgage trusts and public owners report each of the 894 US malls (query: the mall name plus its city; filers classified by SEC industry code: 6189 asset-backed trust, 6798 and 65xx real-estate owner): 162 malls text-match a CMBS trust or depositor filing (15 with the mall name printed right before its city, as in a loan table) and 306 a public owner's filing (171 name-before-city); 161 match only tenants or other companies (in the CSV, not the tables); 69 have names that another record shares or that return over 200 hits and are excluded from the tables. The groups overlap. A text match is a candidate, not a confirmed loan or owner. Spot checks against the filing text on 20 Oct 2026 (30 links, one per site): strong trust links 1 of 8 false, weak trust links 3 of 8, strong owner links 1 of 8, weak owner links 3 of 6. Small samples: the strong tier is better, not clean (one strong link was the same mall name in another state). Computed at build 2026-10-20; counts describe the database, not the market. Deal table · Where each number comes from.

Deal events per year, by kinddeal table, 2000 to date; tap or hover a year for the split

Events per year. Systematic news collection began in 2024 (dashed line); earlier rows come from Wikipedia text and are sparse, so the rise is mostly coverage, not the market. 2026 is a partial year. 29 undated and 88 pre-2000 events are not charted.

Sale prices by yearsales with a reported price since 2010; medians, n in brackets

US malls only, and only sales whose supporting sentence carries the price and a sale verb; a keyword heuristic drops parcels, anchor boxes, former department stores, parking lots, portfolios, developments, loans and listings, and an acreage test drops deals under half the site's acres: 76 rows left out. One row per site and price across years. The sum is of the handful of prices reported in news, not market volume. No year yet has five sales with size data, so no $/sq ft or $/acre is shown.

Loan events reported in news and Wikipedialoan, special-servicing and foreclosure rows from the deal table, any year

CMBS trusts and depositors naming these mallsSEC EDGAR full-text search (2001 to date), forms 10-K/10-D/8-K/10-Q/ABS-EE; 'strong' = the mall name is printed right before its city in the filing; depositor corporations issue many trusts

When a mall loan is securitised, the trust or its depositor files an 8-K with the prospectus that names the property, and the trust then files annual 10-Ks. A mall named in those filings may be collateral in the pool, or may be mentioned as a neighbour or a tenant location: the trust name dates the pool and names the deal to look up, and the filing itself has to be read to confirm the rest. This is the raw public trail behind what Trepp and CoStar package, without the loan-level numbers (those sit in exhibits EDGAR does not index for search). Residential and consumer-loan trusts are filtered out by a name pattern, which can miss some.

Trust vintages (the year in the trust name) for strong mall-to-trust links on sites that pass the common-name filter; 11 strong links to depositors without a year in the name are not charted. EDGAR full-text search starts in 2001, so earlier deals appear only through later filings. If a loan ran ten years, a 2014 to 2016 vintage matures in 2024 to 2026; five- and seven-year terms and refinancings exist, so read the prospectus. Several trusts on one mall can mean one loan split into pari-passu notes, or several false matches.

Candidate loan vintages (strong matches only)the year in the trust name, for malls whose name is printed right before the city in the trust's filing; most recent first

4 malls; all shown. What EDGAR can and cannot say: the trust vintage is the year the pool was issued, which usually dates the loan; the term, servicer, balance and current status are not in the searchable text, so a maturity has to be read from the prospectus. Weak matches (name and city anywhere in the same filing) are left out of this table and the chart; they are in the CSV. Whether the loan is still in the pool, in special servicing or paid off sits in the trust's monthly 10-D exhibits, which are not full-text indexed; read the trust's latest distribution report for that. Counts by year: 2014: 1, 2007: 1, 2006: 1, 2004: 1.

Public companies named in filings (not necessarily current owners)REITs and real-estate companies (SEC industry codes 6798, 65xx) whose filings name the mall; a filing can describe a mall sold years ago; one company files under several registrants, so counts here differ from the owner hubs, which count recorded roles

Lenders, buyers and sellers named in deal rowsdeal table; rows and distinct sites

Downloads: deals.csv · edgar.csv (one row per mall-to-filer link). Method and limits: data sources. Loan-level numbers (balance, maturity, DSCR, appraisal, servicer status) now come from the trusts' ABS-EE Exhibit 102 where one exists (trusts issued since late 2016), and owner-reported GLA, occupancy, sales and anchors from REIT 10-K property tables: sec_facts.csv, shown on each record under “Securitised debt and public filers”.