Methodology
Data Sources & Methodology
Every figure on this site traces to a named source with a date. This page documents where each dataset comes from, how often it refreshes, and how to cite it.
Last updated:
County deed records
Our lending-activity rankings come from commercial deeds of trust recorded in county public records — starting with Harris County, Texas. We collect recorded documents, identify the lender of record, classify the property type, and capture the loan amount when the instrument discloses it. Collection is ongoing and rankings are compiled quarterly; each ranking page states the quarter it covers.
Recorded volume measures activity, not quality: a deed of trust proves a loan was made, on that property, by that lender, in that period. No lender pays to appear, and no lender can pay to be removed.
FFIEC Call Reports
Bank balance-sheet data — CRE concentration, loan book composition, capital ratios — comes from quarterly Call Reports filed with the FFIEC. We refresh after each quarterly filing deadline, and every figure is labeled with its reporting quarter. Field definitions and what each metric means for borrowers are in our CRE Finance Glossary.
NCUA filings
Credit union data comes from quarterly NCUA 5300 Call Reports, handled the same way: refreshed quarterly, labeled with the reporting period.
Enforcement actions
Regulatory enforcement data comes from the public enforcement databases of the federal banking regulators (FDIC, OCC, Federal Reserve, NCUA). We report what the regulator published — action type, date, and status — and link the public record where available.
Hiring signals
Public job postings (commercial lenders, CRE credit officers, portfolio managers) are a leading indicator of lending appetite. We track postings from public hiring feeds and careers pages, note them with the date observed, and treat them as signals — a posting means a lender is staffing up, not that any specific loan will be made.
Community-discussion summaries
Where borrowers and brokers discuss lenders publicly, we summarize the discussion in our own original words and link the source. Verify-then-publish: a claim from a discussion thread is either corroborated against records or clearly framed as reported sentiment. We do not republish others' content wholesale.
First-party pulses (observed lender terms)
As an operating commercial mortgage brokerage, RefiLoop talks to lenders about live deals. Pulses are dated observations from those real conversations: what a lender quoted, required, or declined, recorded at the time. Every pulse shows its observation date and ages visibly — reported terms are observations subject to change, never offers of credit. Lenders can dispute or update any pulse through our Corrections & Removal Policy.
Cite this data
You are welcome to cite any figure on this site — in articles, research, newsletters, or reports — with attribution to "RefiLoop Lender Data" and a link to the page the figure appears on. Many pages include a copy-ready citation block. A good citation names the figure, the period, and the source, for example:
According to RefiLoop Lender Data, [lender] recorded [N] commercial deeds of trust in Harris County in [quarter]. Source: [page URL]
For bulk data access, custom cuts, or methodology questions, email david@refiloop.com.