What data is included?

Mortgage Lender Data for HubSpot writes to its own fields on your company records, and you choose which ones. Everything is optional except identity: skip a field and it's never created, so your company record doesn't fill up with fields nobody uses.

Every field starts with lm_, and each field's HubSpot description states what it measures and its main caveat, so the definition travels with the data. This article covers all of it: the standard set that comes with every engagement, the four optional packs, and what every single field means.

The standard set

Included in every engagement: seventy-four fields across seven groups, plus three bookkeeping fields the sync uses to mark duplicates and record what it created. That's seventy-seven in all.

Identity (12 fields) — Who the lender is, their regulator, their identifiers, where they are
Hierarchy (6 fields) — Parent and ultimate parent, and how many filing entities share it
Operational (3 fields) — Bookkeeping fields the sync uses to mark duplicates and record what it created
Volume and rank (14 fields) — Loans, dollars, applications, denials, conversion, national rank
Trend (9 fields) — Last year's figures, the change, and a growth direction
Product mix (14 fields) — Conventional, FHA, VA and USDA share, lien position, occupancy, manufactured housing
Denial analysis (15 fields) — Denials broken out by reason, with dollar volume on the two most actionable
Demographics (4 fields) — Applicant income, and lending in lower-income and majority-minority areas

Most teams take all of it. The groups exist so you can leave out what you won't use.

The optional packs

Four packs are sold separately, because each serves a specific kind of buyer rather than everyone.

Credit model detail · 17 fields

Every lender's originations, counted by the credit scoring model used on the primary applicant. One field per model, plus a single field naming whichever model a lender uses most, so you can sort and filter without reading sixteen numbers.

You see exactly which models a lender pulls and in what volume. A lender running mostly on one generation of model, or splitting across several, becomes visible and filterable.

Who this is for. Credit bureaus, scoring providers, verification companies, and anyone whose product depends on which model a lender relies on.

What to know before selecting it. Five of the models only became reportable in 2024, so they read zero in earlier years rather than being unused. Counts are on the primary applicant only, so a joint application counts once, not twice. The sixteen counts add up to the lender's total originations, which is how we check the numbers are right.

Which credit model a lender uses is only available in this pack. It's not part of the standard set.

Lending footprint · 70 fields

A loan count for every state, plus concentration measures: how many states, counties, and metro areas the lender works in, which ones lead, and what share of the business each represents.

The state counts answer one question: does this lender originate where you operate, and at what volume. A state's field is only filled when the lender actually lends there, so filtering on "Texas originations is known" finds every Texas lender, and sorting on it ranks them. A separate field lists each lender's material states, so one filter covers all the states you're licensed in at once.

Who this is for. Title and settlement companies, and any vendor whose licensing, pricing, or coverage is drawn on a map.

What to know before selecting it. These are loan counts, not dollars, because counts are exact and dollars are approximate. Each state carries one number. Finer cuts, like purchase dollars inside your licensed states or county-level product mix, come from the underlying data as part of the engagement rather than living as fields.

Channel and balance sheet · 14 fields

What kind of institution a lender is, and what happens to its loans after closing.

Disposition: the share kept on the balance sheet, sold to Fannie Mae or Freddie Mac, securitized through Ginnie Mae, securitized privately, or sold as whole loans, plus a single field naming the primary channel. Institution type: bank or thrift, credit union, independent mortgage company, or bank-affiliated mortgage company. And loan-to-value: the median, the share of conventional loans above 80% LTV, and a count and dollar figure for loans that carry mortgage insurance.

Who this is for. Mortgage insurers, warehouse lenders, capital markets desks, and non-QM investors.

What to know before selecting it. The portfolio-retained figure is an upper bound: the federal filing only records a sale made in the same calendar year the loan closed, so a loan sold the following January reads as kept. Loan-to-value can be legally omitted by smaller banks and credit unions, so the pack includes a coverage field showing how much of each lender's data carries it, and the LTV measures go blank below 50% coverage rather than showing a number built on too little data. The ratio is combined across all liens on the property, which slightly overstates the mortgage insurance figures where piggyback second liens are common.

Institution profile · 26 fields

Who the institution is beyond its name. Regulator and charter class, total assets with year-over-year growth, the most recent CRA exam (date, rating, and method), minority- or women-owned designation, deposit insurance, Federal Home Loan Bank membership, the ownership organization's name and size, how long it's filed federal lending data, and whether the legal entity is still active. That last part includes a registered street address and a headquarters street address, which federal lending data alone doesn't carry.

Who this is for. Anyone selling to banks and credit unions on compliance or relationship grounds, and anyone who needs to know an account's size, health, and corporate family before the first call.

What to know before selecting it. CRA fields apply to banks and thrifts only. Credit unions and independent mortgage companies aren't CRA-examined, so those fields are blank for them, correctly. Asset figures are as reported to regulators, in thousands of dollars. The filing-status field updates at the annual refresh: a lender that stops appearing in federal data is marked "No longer filing," which usually means an acquisition.

What each field means

All counts are exact. All dollar fields are approximate, because the government publishes each loan amount as the midpoint of a $10,000 band.

Coded values never reach your portal. The federal files use numeric codes (regulator 3 means FDIC, for example), and every code gets translated to its readable label before writing. Fields with a fixed set of values arrive as dropdowns whose options are those labels. A code outside the documented federal list shows as "Unknown" rather than a guess.

A blank is an answer, not an error. It means the measure doesn't apply, or the source didn't carry it: a lender with no denials has no denial shares, a state a lender doesn't lend in has no count, and a measure below its coverage floor gets blanked rather than shown on thin data. Zero and blank never mean the same thing.

Identity (12 fields)

lm_lei: The Legal Entity Identifier: a permanent 20-character code assigned to the legal entity itself. The key everything joins on. Never changes with a rebrand, move, or acquisition.
lm_legal_name: The legal name as registered. Frequently differs from the brand name your team uses; the service doesn't "correct" it.
lm_domain: The lender's website domain, from its federal registration.
lm_rssd: The Federal Reserve's ID for the institution. Blank for most independent mortgage companies, which have none.
lm_fdic_cert: FDIC certificate number. Banks only.
lm_ncua_charter: NCUA charter number. Credit unions only.
lm_tax_id: Federal tax ID as filed. Stored as text; self-reported, so occasionally malformed.
lm_hq_city: Headquarters city as reported. Not a mailing address.
lm_hq_state: Headquarters state, two-letter code.
lm_hq_zip: Headquarters ZIP code, as filed with federal regulators.
lm_data_year: The activity year the numbers describe.
lm_last_refreshed: The date the service last wrote this record.

Hierarchy (6 fields)

lm_parent_rssd: The direct parent company's Federal Reserve ID. Blank when the lender has no parent.
lm_parent_name: The direct parent's name.
lm_top_holder_rssd: The ultimate parent's Federal Reserve ID: the top of the ownership tree.
lm_top_holder_name: The ultimate parent's name. The field that makes "top 300 lenders" answerable at the relationship level instead of the filing-entity level.
lm_is_top_holder: True when this record is the top of its own family.
lm_family_entity_count: How many filing institutions share this record's ultimate parent. A value above 1 flags a duplicate risk in any CRM, because those entities are usually one commercial relationship.

Operational (3 fields)

lm_duplicate_of_lei: Set on a record identified as a duplicate: the identifier of the surviving record it duplicates. Blank on every clean record.
lm_record_source: Marks records the service created, so they are always distinguishable from records your team made.
lm_match_method: How this record was matched: an identifier you already had, website domain, name and state, or your manual confirmation.

Volume and rank (14 fields)

lm_originations_count: Loans originated in the data year. Zero is a valid value.
lm_originations_amount: Total dollars originated. Approximate.
lm_purchase_count: Home-purchase loans originated (as opposed to refinances and other purposes).
lm_purchase_amount: Home-purchase dollars originated. Approximate.
lm_purchase_share: Home-purchase loans as a percent of all originations. Blank when there are no originations.
lm_first_lien_purchase_count: Home-purchase originations in first-lien position.
lm_avg_purchase_amount: Average home-purchase loan size. Approximate. Blank when there are no purchase loans.
lm_applications_received: Applications received: originated, approved but not accepted, denied, withdrawn, or closed for incompleteness. Excludes purchased loans and preapproval requests.
lm_applications_denied: Applications denied.
lm_denial_rate: Denials as a percent of applications received.
lm_purchase_denial_rate: Denial rate on home-purchase applications only.
lm_pull_through_rate: Originations as a percent of applications received.
lm_rank_by_count: National rank by loans originated, 1 = largest.
lm_rank_by_amount: National rank by dollars originated. Ranks an approximate measure; prefer rank by count.

Trend (9 fields)

All blank for a lender with no prior-year filing.

lm_originations_count_prior: The prior year's origination count.
lm_originations_change: This year's count minus last year's. Can be negative.
lm_originations_change_pct: The same change as a percent of the prior year.
lm_originations_amount_change: The dollar-volume change. Approximate.
lm_rank_by_count_prior: Last year's national rank by count.
lm_rank_change: Rank movement. Positive means the lender climbed: rank 200 to 150 shows +50.
lm_denial_rate_change: The denial-rate change, in percentage points.
lm_pull_through_change: The pull-through change, in percentage points.
lm_growth_direction: One word: Growing (up at least 5%), Flat (within 5%), Shrinking (down at least 5%), or Stopped filing.

Product mix (14 fields)

Shares are percents of originations. Blank when there are no originations.

lm_conventional_share: Conventional loans as a percent of originations.
lm_fha_share: FHA-insured share.
lm_va_share: VA-guaranteed share.
lm_usda_share: USDA share.
lm_gov_backed_purchase_share: Government-backed (FHA, VA, USDA) share of home-purchase originations.
lm_conv_to_gov_ratio: Conventional originations divided by government-backed originations. Blank when there are no government-backed loans.
lm_first_lien_share: First-lien share of originations.
lm_subordinate_lien_share: Second-lien and other subordinate share.
lm_closed_end_share: Closed-end loans as a share of originations where the lender reported the loan structure.
lm_open_end_share: Open-end lines (HELOC-style) on the same basis.
lm_owner_occupied_share: Share of originations on the borrower's primary residence.
lm_site_built_1_4_share: Share on site-built one-to-four-unit homes.
lm_manufactured_count: Manufactured-home loans originated.
lm_manufactured_amount: Manufactured-home dollars. Approximate, and proportionally most approximate here because these are small-balance loans.

Denial analysis (15 fields)

Reason shares are computed against denials that carry a reason, and blanked when lm_denial_reason_coverage falls below 50%. See "What do the denial reasons mean?"

lm_denial_reason_coverage: The percent of this lender's denials that report a reason. Smaller lenders may legally omit reasons. Blank when there are no denials.
lm_denial_dti_count: Denials citing debt-to-income ratio.
lm_denial_employment_count: Denials citing employment history.
lm_denial_credit_count: Denials citing credit history.
lm_denial_collateral_count: Denials citing collateral.
lm_denial_cash_count: Denials citing insufficient cash for down payment or closing costs.
lm_denial_unverifiable_count: Denials citing unverifiable information.
lm_denial_incomplete_count: Denials citing an incomplete application.
lm_denial_mi_count: Denials citing mortgage insurance denied.
lm_denial_other_count: Denials citing "other".
lm_denial_cash_amount: Dollar volume of denials citing insufficient cash. Approximate.
lm_denial_unverifiable_amount: Dollar volume of denials citing unverifiable information. Approximate.
lm_salvageable_denial_count: Distinct denials citing any fixable reason: debt-to-income, insufficient cash, unverifiable information, or incomplete application. Distinct, so a denial citing two of them counts once.
lm_salvageable_denial_amount: Dollar volume of those denials. Approximate.
lm_top_denial_reason: The most-cited reason. Blank when no denials carry a reason.

Demographics (4 fields)

lm_lmi_tract_share: Share of originations in lower-income areas: census tracts with median family income below 80% of the metro's. The Community Reinvestment Act definition.
lm_lmi_borrower_share: Share of originations to lower-income borrowers: the applicant's own income below 80% of the area median. Computed over originations with a usable income figure; business-purpose and multifamily loans report no income and are excluded.
lm_minority_tract_share: Share of originations in census tracts with a majority-minority population. The 50% threshold is a convention, not a regulatory definition.
lm_avg_applicant_income: Average applicant income on originations, in thousands of dollars as filed. Joint applications carry combined household income.

Credit model pack (17 fields, optional)

A count of originations per credit scoring model used on the primary applicant. The sixteen counts sum to the lender's total originations, which is how the numbers are checked. Model fields are ordered by federal code; no model is ranked above another. Five models became reportable in 2024 and read zero in earlier years rather than being unused.

lm_orig_equifax_beacon_5: Equifax Beacon 5.0
lm_orig_experian_fair_isaac_2: Experian Fair Isaac v2
lm_orig_fico_classic_04: FICO Risk Score Classic 04
lm_orig_fico_classic_98: FICO Risk Score Classic 98
lm_orig_vantage_2_0: VantageScore 2.0
lm_orig_vantage_3_0: VantageScore 3.0
lm_orig_multiple_models: More than one model
lm_orig_other_model: A model outside the federal list
lm_orig_no_model: No model used
lm_orig_fico_9: FICO 9
lm_orig_fico_8: FICO 8
lm_orig_fico_10: FICO 10
lm_orig_fico_10t: FICO 10T
lm_orig_vantage_4_0: VantageScore 4.0
lm_orig_model_exempt: Loans where the lender is exempt from reporting the model
lm_orig_model_unknown: Loans where the model field is blank or unrecognized
lm_primary_credit_model: The model this lender uses most, computed from the counts above.

Lending footprint pack (70 fields, optional)

Fifty-seven origination-count fields plus thirteen concentration measures. State counts are exact loan counts, filled only where the lender actually lends, so a blank means no lending there and filters find real activity.

lm_orig_count_: One field per state and territory, 56 in all: loans originated in that state. Blank where the lender made none.
lm_orig_count_no_state: Originations with no state reported. The catch-all that makes the 57 counts sum to total originations, so nothing is silently lost.
lm_states_active: The lender's material states: at least 1% of originations or at least 100 loans. One filter covers every state you are licensed in.
lm_state_count: How many states the lender originated in.
lm_top_state: The lender's largest state by loan count.
lm_top_state_share: That state's share of originations.
lm_top3_state_share: The top three states' combined share.
lm_county_count: How many counties the lender originated in.
lm_top_county: The largest county, shown by name: "St. Louis County, MO (29189)".
lm_top_county_share: That county's share of originations.
lm_msa_count: How many metro areas the lender originated in.
lm_top_msa: The largest metro area, shown by name.
lm_top_msa_share: That metro's share of originations.
lm_top5_msa_share: The top five metros' combined share.
lm_nonmetro_share: The share of originations outside any metro area.

Channel and balance sheet pack (14 fields, optional)

The six disposition shares sum to 100 by construction.

lm_portfolio_share: Share of originations kept on the balance sheet. An upper bound: a loan sold in a later calendar year still reads as kept, because the filing only records same-year sales.
lm_gse_share: Share sold to Fannie Mae or Freddie Mac.
lm_ginnie_share: Share securitized through Ginnie Mae.
lm_private_securitization_share: Share securitized privately. The non-QM signal.
lm_whole_loan_share: Share sold as whole loans to banks, credit unions, mortgage and finance companies, or insurers.
lm_other_purchaser_share: Share sold to any other purchaser type.
lm_primary_disposition: The largest of the six shares, as one word. Computed from the shares, so the two can never disagree.
lm_institution_type: One of five values: Credit union, Bank or thrift, Independent mortgage company, Bank-affiliated mortgage company, or Unknown. Never blank.
lm_ltv_coverage: The percent of first-lien originations that carry a loan-to-value ratio. Smaller banks and credit unions may legally omit it.
lm_median_cltv: The median combined loan-to-value on first-lien originations. Blanked below 50% coverage.
lm_high_ltv_conv_share: The share of conventional first-lien originations above 80% LTV: the mortgage insurance trigger line. Blanked below 50% coverage.
lm_ltv95_share: The share of first-lien originations at 95% LTV or higher: the down-payment-assistance and high-LTV-program signal. Blanked below 50% coverage.
lm_mi_addressable_count: Conventional first-lien originations above 80% LTV: loans that carry mortgage insurance. A floor, not blanked, because a floor built on partial coverage is still a floor.
lm_high_ltv_conv_amount: The dollar volume of those loans. Approximate.

Institution profile pack (26 fields, optional)

lm_regulator: The lender's federal regulator (OCC, Federal Reserve, FDIC, NCUA, HUD, or CFPB).
lm_institution_class: Charter-based institution class (commercial bank, thrift, credit union, independent mortgage bank, and their variants).
lm_total_assets: Total assets in thousands of dollars as of December 31 of the data year.
lm_prior_year_assets: Total assets one year earlier, thousands of dollars.
lm_asset_growth_pct: Year-over-year asset growth, percent. May be negative.
lm_cra_exam_date: Date of the most recent CRA examination. Banks and thrifts only.
lm_cra_rating: Most recent CRA rating: Outstanding, Satisfactory, Needs to improve, or Substantial noncompliance.
lm_cra_exam_method: The exam method used (large bank, small bank, strategic plan, and others).
lm_minority_owned: Minority- or women-owned designation, by category.
lm_occ_charter: OCC charter number, national banks only.
lm_thrift_id: Thrift institution identifier.
lm_deposit_insurance: FDIC, NCUSIF, Other, or None.
lm_fhlb_district: Federal Home Loan Bank district, when a member.
lm_fhlb_id: FHLB member identifier.
lm_org_name: The ownership organization this lender belongs to.
lm_org_assets: The whole organization's total assets, thousands of dollars.
lm_org_1_4_family_holdings: The organization's 1-4 family mortgage holdings, thousands of dollars.
lm_exemption_status: Whether the lender is exempt from reporting certain fields, and how they filed.
lm_quarterly_filer: Whether the lender is large enough to file quarterly (roughly the 37 largest).
lm_first_filed_year: First year in the federal data we carry.
lm_last_filed_year: Most recent year the lender filed.
lm_filing_status: Active filer, or No longer filing (updated at the annual refresh; usually means an acquisition).
lm_entity_status: Whether the legal entity is active, per the global LEI system.
lm_lei_registration_status: Maintenance status of the LEI registration. LAPSED only means the renewal wasn't paid.
lm_registered_address: Registered legal address (can differ from the operating headquarters).
lm_gleif_hq_address: Street-level headquarters address, from the global LEI system.