Market summary (volume, units, rate, loan amount, credit score, LTV, DTI, income, etc.) with previous-period comparison

The last calendar date (YYYY-MM-DD) this query's matching records actually reach. Records arrive after the events they describe, so a period ending today is normally backed by data ending days or weeks earlier, and any bucket extending past this date is built from a partial period. Measured from THIS query's own matched set, not a global figure: the delay varies by geography and moves over time, so two requests with different filters can legitimately report different dates.

Average appraised property value. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average combined LTV (including subordinate liens). Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average credit score. Excludes values outside 300–850 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average back-end DTI ratio across loan APPLICATIONS matching the filters — every outcome, on the same population as units. A denied application carries a much higher ratio than a funded loan, often above 100, so this figure sits several points above what closed borrowers actually carry. For originations use originatedAvgDti. Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average front-end DTI ratio. Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average household annual income. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average LTV ratio. Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average loan amount. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average loan term in months. Excludes values outside 0–1,200 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average purchase price. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average note rate. Excludes values outside 0–25 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Per-measure count of documents this query's sentinel guard excluded — documents that carry the measure's field but whose value is an out-of-range fill. Compare against units to see what share of the matched population each average was actually computed over. Absent for units (a document count, not a guarded metric).

Average back-end DTI ratio across ORIGINATED loans only — the subset of applications whose hmdaActionTaken is "loan originated". Prefer this over avgDTI for any question about what borrowers who actually got a loan are carrying. avgDTI averages applications of EVERY outcome, and a denied application legitimately carries a far higher ratio — often above 100 — because that is frequently why it was denied, so the two figures differ by several points on the same filters and neither is a correction of the other. null when the matched set contains no originations. Excludes values outside 0–100 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Loans ORIGINATED — the subset of units whose hmdaActionTaken is "loan originated". This is the count most callers mean by "number of loans"

Total dollar volume of ORIGINATED loans — the subset of volume whose hmdaActionTaken is "loan originated". Pair with originatedUnits for an originated average loan size. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

How the matched applications split by outcome — a complete PARTITION of units and volume, so the two headline numbers can be reconciled against their parts without a second query. Roughly half of all rows are not originated loans and the share varies widely by geography, so no fixed correction ratio exists. Summary only: on the chart use slice: "hmdaActionTaken", on the time-series segment: "hmdaActionTaken".

Loan APPLICATIONS matching the filters — every outcome, not just originations. Withdrawn, denied, closed-for-incompleteness and secondary-market purchases are all counted here; roughly half of all rows are not originated loans, and the originated share varies enough by geography that no fixed correction ratio exists. For originations use originatedUnits; outcomes reports the full breakdown, and hmdaActionTaken filters on it

Total dollar volume of loan APPLICATIONS matching the filters — every outcome, on the same population as units, not just originations. For originated dollars use originatedVolume. volume ÷ units is an application average, not an originated average. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average appraised property value. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average combined LTV (including subordinate liens). Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average credit score. Excludes values outside 300–850 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average back-end DTI ratio across loan APPLICATIONS matching the filters — every outcome, on the same population as units. A denied application carries a much higher ratio than a funded loan, often above 100, so this figure sits several points above what closed borrowers actually carry. For originations use originatedAvgDti. Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average front-end DTI ratio. Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average household annual income. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average LTV ratio. Excludes values outside 0–999 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average loan amount. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average loan term in months. Excludes values outside 0–1,200 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average purchase price. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Average note rate. Excludes values outside 0–25 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Per-measure count of documents this query's sentinel guard excluded — documents that carry the measure's field but whose value is an out-of-range fill. Compare against units to see what share of the matched population each average was actually computed over. Absent for units (a document count, not a guarded metric).

Average back-end DTI ratio across ORIGINATED loans only — the subset of applications whose hmdaActionTaken is "loan originated". Prefer this over avgDTI for any question about what borrowers who actually got a loan are carrying. avgDTI averages applications of EVERY outcome, and a denied application legitimately carries a far higher ratio — often above 100 — because that is frequently why it was denied, so the two figures differ by several points on the same filters and neither is a correction of the other. null when the matched set contains no originations. Excludes values outside 0–100 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

Loans ORIGINATED — the subset of units whose hmdaActionTaken is "loan originated". This is the count most callers mean by "number of loans"

Total dollar volume of ORIGINATED loans — the subset of volume whose hmdaActionTaken is "loan originated". Pair with originatedUnits for an originated average loan size. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.

How the matched applications split by outcome — a complete PARTITION of units and volume, so the two headline numbers can be reconciled against their parts without a second query. Roughly half of all rows are not originated loans and the share varies widely by geography, so no fixed correction ratio exists. Summary only: on the chart use slice: "hmdaActionTaken", on the time-series segment: "hmdaActionTaken".

Loan APPLICATIONS matching the filters — every outcome, not just originations. Withdrawn, denied, closed-for-incompleteness and secondary-market purchases are all counted here; roughly half of all rows are not originated loans, and the originated share varies enough by geography that no fixed correction ratio exists. For originations use originatedUnits; outcomes reports the full breakdown, and hmdaActionTaken filters on it

Total dollar volume of loan APPLICATIONS matching the filters — every outcome, on the same population as units, not just originations. For originated dollars use originatedVolume. volume ÷ units is an application average, not an originated average. Excludes values outside 0–100,000,000 — the source encodes unknown values as an out-of-range fill, and averaging them raw is wrong by orders of magnitude. The bound is applied before the metric and is not part of the public filter surface, so an equivalent range filter will not reproduce this figure exactly.