Home/White Papers/FQHC NCR Framework
Edition 1, 2026 · 36 pages · FQHC finance

The FQHC NCR framework: net over gross.

For a PPS-paid FQHC, GCR is the wrong lens. The encounter rate has almost no relationship to gross charge. NCR is the right number. This is the working framework for the CFO, the billing director, and the senior accountant who own the FQHC finance reporting cycle.

Edition
1 · 2026
Length
36 pages
Audience
FQHC finance
Lens
NCR
Reference
UDS Table 9
Example
7-site FQHC

Executive summaryFive things to fix the FQHC finance lens.

FQHCs are paid differently from fee-for-service practices. The reporting frameworks borrowed from FFS billing produce numbers that mislead the board, the CFO, and the auditor. This framework is the working corrective.

01
GCR is structurally misleading for PPS-paid FQHCs.
02
NCR computed on expected reimbursement is the right lens.
03
T1015 denials concentrate in roughly ten root cause categories.
04
The wrap is reconciled to UDS Table 9 encounter counts.
05
Behavioral health and SUD encounters are segmented inside the NCR.

The FQHC encounter rate is the single most important number in FQHC finance. Medicaid and Medicare both pay an FQHC a contracted per-encounter rate that has almost no relationship to the gross charge that the EHR generates for the visit. A typical FQHC will see gross charges of three hundred dollars for an encounter that pays a two-hundred-twenty-dollar PPS rate. The GCR that arithmetic produces is roughly seventy-three percent. The reality is the encounter paid at one hundred percent of the expected rate. The reporting framework that uses GCR shows an FQHC in trouble that is actually performing perfectly. We have watched boards spend three full quarters chasing a phantom collection problem because the GCR was sliding while NCR sat at ninety-six percent. The encounter rate was unchanged. The charge master was updated. That was the entire story.

The NCR lens fixes this by dividing cash collected by expected reimbursement on the actual payer mix. For an FQHC encounter paid on PPS plus wrap, the expected reimbursement is the PPS rate. The NCR for a perfectly performing FQHC sits near one hundred percent. When NCR drops, the framework knows where to look. When GCR drops, the framework cannot distinguish a real problem from a structural distortion in how the EHR posts charges. The five takeaways above carry through the rest of this paper. Each one is a working hypothesis the framework tests against twelve months of remits, the prior UDS submission, and the open AR snapshot the day the engagement starts. The number that moves first is almost always the eligibility denial dollar pool on T1015, because real-time eligibility verification at check-in is the single highest-leverage process change an FQHC can make in week one. The number that moves last, and matters most, is the aggregate NCR on the Medicaid MCO book where wrap reconciliation lags by sixty to one hundred and twenty days. That is the work the framework is built to organize.

The pattern repeats across the FQHCs we have audited. The framework in this paper is the working corrective: it retires GCR from the board pack, anchors expected reimbursement to the state PPS schedule and the wrap, segments BH and SUD encounters into a sub-metric, and locks UDS Table 9 as the source of truth for encounter counts so the wrap reconciliation has somewhere honest to land.

Landscape todayThe FQHC industry in 2026.

The Health Resources and Services Administration counts roughly fourteen hundred federally qualified health centers operating across the United States in 2026, delivering primary care, dental, behavioral health, vision, enabling services, and pharmacy through approximately ten thousand delivery sites. Those fourteen hundred organizations served an estimated thirty-one million unique patients in the most recent UDS reporting year, with seventy percent at or below one hundred percent of the federal poverty level and roughly half on Medicaid. The FQHC sector is the largest single primary-care safety-net network in the country, and the encounter-rate economics that drive every one of those visits run on PPS, not fee-for-service.

PPS rates vary by state, by service line, and by the annual CMS Medicare Economic Index update. Medical encounter rates in 2026 sit in the two-hundred to two-hundred-eighty-dollar range, with California, New York, and Massachusetts at the top of that band. Dental encounter rates run roughly one-hundred-fifty to two-hundred-twenty dollars. Behavioral health encounter rates run one-hundred-eighty to two-hundred-forty dollars, though states with mature BH carve-outs (Pennsylvania CBH, Washington, Oregon CCO) often pay BH encounters through a separate Medicaid BH MCO and the wrap math runs against a different schedule. Every FQHC CFO should know the current PPS rate for each service line at each site to the dollar. Those rates are gazetted by the state Medicaid agency under federal Section 1902(bb) authority.

The payer mix that drives the encounter pool in 2026 typically runs forty-five to sixty percent Medicaid (split between FFS and managed care), eight to fifteen percent Medicare, eighteen to twenty-eight percent self-pay on sliding fee, and the balance commercial. Behavioral health and substance-use-disorder programs at FQHCs that operate them often run a different payer mix, weighted heavily to Medicaid behavioral health carve-outs and self-pay. The wrap reconciliation cycle (the lag between encounter and full PPS realization for managed-care members) introduces sixty to one-hundred-twenty days of cash exposure on every Medicaid MCO encounter. That single structural feature is why FQHC cash days outstanding read higher than fee-for-service practices even when collection performance is excellent. The NCR lens is the only honest way to separate the wrap-lag effect from a real collection problem.

The lensGCR vs NCR for an FQHC.

Four rows that explain why GCR is the wrong lens and NCR is the right one. The table reads as the corrective the CFO can hand the board.

The math underneath the table is intentionally simple. GCR equals cash collected divided by gross charges posted; NCR equals cash collected divided by expected reimbursement on the actual payer mix. For an FQHC encounter, expected reimbursement equals the state PPS rate (which already bundles every service rendered during the encounter, per the bundled-rate rule under 42 CFR 405.2462). The gross charge that the EHR posts is whatever the charge master says, which for most FQHCs reflects a multiple of the Medicare physician fee schedule or a legacy historical rate that nobody touches because it does not affect cash. The GCR ratio of cash-to-charges is therefore not a measure of collection performance; it is a measure of how aggressively the charge master is set relative to the PPS floor.

Targets vary by payer because the contractual mechanics vary. Medicare PPS for FQHCs runs on a cost-based encounter rate under the Medicare Benefit Policy Manual Chapter 13, and an FQHC that bills clean should expect NCR of ninety-seven to ninety-nine percent against the Medicare expected number. Medicaid FFS PPS, where the state pays the encounter rate directly, also targets ninety-four to ninety-eight percent. Medicaid MCO with state wrap is the lower band, target eighty-eight to ninety-four percent because the wrap reconciliation runs quarterly in most states and there is always a tail of encounters that have not yet pulled their wrap dollars. Commercial sits at ninety to ninety-five percent because commercial payers do not pay PPS and the contracted rate is whatever was negotiated. Behavioral health carve-outs (Community Behavioral Health in Pennsylvania, Beacon, Magellan in other states) run lowest at eighty to ninety percent because BH credentialing friction and prior-authorization rules generate a steady leakage of encounter denials that take longer to work than medical denials.

Dimension
Why GCR mislead
What NCR shows
Numerator
Cash collectedThe number is the same in both lenses. Cash is cash.
Cash collectedSame numerator. The split happens on the denominator.
Denominator
Gross chargesThe EHR posts a charge that has no relationship to PPS. The denominator is structural noise.
Expected reimbursementPPS rate plus wrap. The denominator is the dollars the FQHC is contractually owed.
Healthy value
VariableA healthy FQHC will show GCR anywhere from forty to seventy percent depending on charge-posting habits.
Near 100 percentA healthy FQHC will sit near full collection on the expected reimbursement.
What it diagnoses
Almost nothing usefulGCR cannot distinguish a charge-posting choice from a collection problem.
Real collection performanceNCR isolates the gap between owed and collected on the same encounter pool.

The cash waterfallPPS + interim + wrap.

An FQHC encounter for a Medicaid managed-care member generates two payments. The MCO pays its contracted fee schedule. The state pays the wrap to bring the encounter up to the PPS rate. The waterfall below shows the dollars on a typical encounter and where the framework reconciles each payment.

The wrap reconciliation cycle is where most FQHCs leak the largest single dollar pool, and it is the part of the cycle the EHR is least equipped to track. State practices vary. Pennsylvania reconciles quarterly through CBH wrap files for behavioral health and the medical assistance wrap process for primary care. California runs an annual DHCS PPS reconciliation, settled in arrears after the audit. New York runs a semi-annual APG-to-PPS comparison. Texas, Florida, and Georgia each have their own cadence. The cash-flow exposure on a typical Medicaid MCO encounter therefore runs sixty to one hundred and twenty days from date-of-service to full PPS realization. An FQHC reporting AR aging without separating wrap-pending from genuinely-overdue dollars will overstate distress on every monthly board pack.

The framework treats the wrap as a second receivable. Every Medicaid MCO encounter generates two open AR lines from day one: the interim claim AR (worked normally) and the wrap AR (worked on the state's reconciliation cycle). UDS Table 9 is the anchor because it forces an annual count of encounters by payer that the wrap math has to land against. If Table 9 says twenty-eight thousand Medicaid MCO encounters and the wrap dollars received during the year reconcile to twenty-three thousand encounters at the PPS rate, the gap (five thousand encounters worth of wrap) is the recovery target. That dollar amount, computed against the current PPS rate, is the single largest one-time cash recovery the framework routinely identifies in the first ninety days of an engagement.

ENCOUNTER · T1015 PPS rate $220.00 expected MCO INTERIM PAYMENT Contracted fee schedule $95 STATE WRAP PAYMENT PPS rate less MCO paid $125 ENCOUNTER TOTAL PPS realized $220.00 100% OF EXPECTED RECONCILIATION TO UDS TABLE 9 Encounter counts on Table 9 = sum of MCO interim claim counts + state wrap encounter counts. Wrap dollars rounded to PPS rate × Table 9 encounter count. Variance = recovery target. SOURCE OF TRUTH → UDS TABLE 9

Denial taxonomyThe T1015 pareto.

Ten root cause categories explain the bulk of FQHC T1015 denials. The pareto below ranks the categories by dollar impact on a representative seven-site FQHC. The top two categories alone account for roughly fifty-eight percent of denial dollars.

A deeper read of the taxonomy is the reference the billing director keeps on the wall. The framework tracks fifteen root causes in total, and every one of them maps to a specific prevention rate that an in-house workflow can deliver. Rendering provider not credentialed with the payer for date-of-service is a one-hundred-percent preventable denial if the credentialing matrix is current; we recommend a weekly cross-check between the provider roster and the active payer panels for each site. Provider NPI not on file with the MCO panel is the same class of prevention; it is a credentialing and enrollment failure, not a billing failure, and the fix is a payer-onboarding checklist that runs the day a new provider's start date is confirmed. Place-of-service mismatch (POS 11 office versus POS 50 federally qualified health center) is ninety-five percent preventable through a single claim edit at the EHR level that pins the POS to the FQHC default for any provider rendering at an FQHC site.

Missing or invalid encounter date is one-hundred-percent preventable through claim-edit pre-bill rules. T1015 billed without the required E&M companion code is ninety-five percent preventable; many payers require an E&M (typically 99202-99215 for primary care, 90791-90837 for behavioral health) on the same claim as the T1015, and the claim edit should reject any T1015 that lacks a companion. Member not enrolled with the MCO on date-of-service is ninety percent preventable through real-time eligibility (RTE) verification at check-in via the X12 270/271 transaction; the remaining ten percent is mid-month MCO reassignment that no front-desk workflow can catch in real time and that has to be worked retro. Duplicate T1015 same date-of-service same member is one-hundred-percent preventable through claim-edit pre-bill duplicate detection. Service line missing modifier per state Medicaid spec (HE for behavioral health, HF for substance-use disorder, HD for pregnant member) is ninety percent preventable through encounter-type to modifier mapping in the EHR.

NDC missing on injectable companion code is eighty-five percent preventable through pharmacy-EHR integration. PCP assignment mismatch is roughly sixty percent preventable up front and forty percent recoverable retro through the inbound PCP-attribution authorization process. Sliding fee and income verification missing for the self-pay class is one-hundred-percent preventable through UDS Table 4 capture at registration. Provider type not aligned with state encounter rules (LCSW versus LMFT versus LPC eligibility) is ninety percent preventable through a provider-eligibility matrix maintained against the current state Medicaid manual. Service date prior to FQHC effective date with the payer is one-hundred-percent preventable; the fix is a contract-load audit. ICD-10 not aligned with covered diagnosis policy is eighty percent preventable through code-set enforcement at the EHR diagnosis-picker level. Authorization not on file for services requiring auth (MAT, IOP, partial hospitalization) is ninety-five percent preventable through a pre-visit auth-check process.

These root causes concentrate in such a narrow set because FQHCs have a much smaller covered-service surface than a hospital or a multi-specialty group. The T1015 line is the single transaction that drives the encounter rate. The first two categories (eligibility-on-DOS and wrap-not-posted) routinely account for fifty-five to sixty-five percent of denial dollars; the first five account for roughly eighty-five percent. That math drives the rule of thumb that an FQHC denial program can cut T1015 denial dollars by sixty to seventy percent in six months by aggressively working the first three categories alone.

T1015 denial pareto · 12 months · representative FQHC
Source: ASP-RCM FQHC engagements
Missing eligibility on DOS $248K · 32% Wrap not posted by state $202K · 26% Auth required for visit type $92K · 12% PCP attribution mismatch $68K · 9% Service not covered under PPS $52K · 7% Duplicate same-day encounter $38K · 5% Modifier missing on T1015 $26K · 3% Provider not credentialed $20K · 2.5% UB-04 vs 1500 routing error $14K · 1.8% Other $10K · 1.7% $0 $770K total

Worked exampleAnonymized seven-site FQHC.

The worked example below is composited from a real engagement and presented at a level of detail that an FQHC CFO or controller can read against their own book. The health center is a seven-site community health center on the Eastern seaboard with a heavy behavioral-health and substance-use-disorder program, serving a Medicaid-dominant patient population through the state's behavioral health carve-out (a Community Behavioral Health MCO model) plus a primary care Medicaid managed-care plan. Annual encounter volume sat at roughly fifteen thousand across the seven sites, split sixty-two percent medical, twenty-four percent behavioral health, and fourteen percent dental. Pre-engagement aggregate GCR was forty-seven percent (the board read distress); pre-engagement aggregate NCR computed on actual PPS expected reimbursement was seventy-six percent (the framework read a real but addressable problem, concentrated in two payer pools). The diagnostic delta between those two numbers, twenty-nine percentage points, is roughly the typical gap an FQHC will see between GCR and NCR when its charge master sits at a three-to-one multiple of PPS. That gap is structural; it tells the CFO nothing about collection performance and the framework retires it from the board pack on day one.

The NCR drag concentrated in two places. First, the Medicaid MCO wrap had not been reconciled to UDS Table 9 in three consecutive years, leaving an estimated $1.8M annualized in unposted wrap dollars sitting on the state ledger. Second, T1015 denials on the behavioral health book were running fourteen-point-two percent, driven by a credentialing gap on three LCSWs who had been rendering encounters at two sites for nine months without active panel enrollment with the carve-out MCO. The framework's first ninety days isolated those two pools, ran a recovery filing on the wrap, ran a corrected-claim re-bill on the denied BH encounters once credentialing was retro-dated, and moved board reporting onto NCR with the wrap reconciliation appendix. By month six, aggregate NCR moved from seventy-six to ninety-two percent, T1015 denial rate fell from fourteen-point-two to five-point-one percent, eligibility denial dollars dropped from $248K to $72K, cash days outstanding moved from fifty-eight to thirty-four, and annualized wrap recovery hit $1.8M. None of those moves required additional staff. Each one required the lens to be right first.

Anonymized worked example

Seven-site FQHC. NCR 76 to 92 in six months.

  • Pre-engagement baseline. GCR sat at 47 percent. NCR sat at 76 percent. T1015 denials concentrated in eligibility and wrap.
  • Lens reset, month 1. Board reporting moved to NCR. UDS Table 9 was anchored as the encounter source of truth.
  • Denial taxonomy, months 2-4. The top two categories worked aggressively. Eligibility verification moved to a pre-visit step.
  • Wrap reconciliation, months 3-6. Wrap dollars reconciled to Table 9. Recovery filings on aged wrap encounters.
  • Steady-state at month 6. NCR 92 percent. Annualized wrap recovery $1.8M. Eligibility denials down 71 percent.
NCR · 6-MO TRAJECTORY M 0 M 6 76% 92%
ASP-RCM · FQHC senior partner team NCR trajectory · monthly

Dashboard viewWhat the CFO actually sees.

A live view of NCR by payer with the wrap reconciliation status and the top T1015 denial categories joined in. The CFO and the senior partner share one view. Monthly board reporting drops out of this dashboard.

NCR · by payer · 7 sites · 12 months refreshed monthly
Medicaid FFS2,840 encounters
78%
95%
Medicaid MCO A4,210 encounters · PPS wrap
71%
93%
Medicaid MCO B3,560 encounters · PPS wrap
68%
91%
Medicaid MCO C1,820 encounters · PPS wrap
74%
94%
Medicare980 encounters · cost-based
82%
96%
Commercial620 encounters · contracted
88%
97%
Self-pay sliding fee1,140 encounters
62%
81%
Aggregate NCR15,170 encounters
76%
92%
7-Site FQHCAnonymized · representative engagement
WORKED EXAMPLE

KPI movement before vs after the framework.

Aggregate NCR
76%
92%
T1015 denial rate
14.2%
5.1%
Annualized wrap recovery
$0
$1.8M
Eligibility denial dollars
$248K
$72K
Wrap reconciled to UDS Table 9
No
Yes
Cash days outstanding
58d
34d
Edition 1 · 2026 ASP-RCM FQHC team

The day the board saw NCR instead of GCR, the conversation changed. We had been explaining a structural quirk for three years. The new lens showed the same FQHC was actually collecting almost all the money it was owed. The problem we thought we had was a measurement problem.

CFO · 7-site FQHC · anonymized

Implementation checklistLand the NCR lens in 90 days.

Eight steps to move FQHC finance reporting onto the NCR lens with the wrap reconciliation anchored to UDS Table 9.

A few practitioner notes that the checklist depends on. AR aging is reported by date-of-service, not by date-of-claim, because the lag between encounter and clean claim submission at most FQHCs runs seven to forty-five days and DOC-based aging hides the real risk in the under-thirty bucket. The standard buckets are zero-thirty, thirty-one-sixty, sixty-one-ninety, ninety-one-one-twenty, one-twenty-one-one-eighty, one-eighty-one-three-sixty-five, and over-three-sixty-five days since DOS. An over-ninety percentage above thirty percent of total AR signals operational distress and the framework escalates immediately. UDS Table 9 (encounters by payer) plus UDS Table 4 (patients by income, used for sliding fee discount class) plus UDS Table 5 (services, staffing, productivity) are the three reports every FQHC files with HRSA annually and that the framework cross-references at month three to confirm the encounter pool and the discount-class distribution are reconciling. Sliding fee discount classes follow PIN 2014-02: Class A (one-hundred-percent discount at or below one hundred percent of federal poverty level, but with a nominal fee), Class B (seventy-five-percent discount, one-oh-one to one-twenty-five percent FPL), Class C (fifty-percent discount, one-twenty-six to one-fifty percent FPL), Class D (twenty-five-percent discount, one-fifty-one to one-seventy-five percent FPL), Class E (nominal fee only, one-seventy-six to two-hundred percent FPL), and Full Pay (above two-hundred percent FPL).

For FQHCs participating in the 340B drug pricing program, the framework adds three additional control points. Duplicate-discount avoidance: a 340B drug billed to Medicaid cannot also receive the Medicaid rebate, and the claim must carry the appropriate modifier (JG or TB per CMS) so the state can suppress the rebate. HRSA OPAIS database verification of covered-entity status, contract pharmacy registrations, and child-site listings on a quarterly recertification cadence. Auditable patient definition under HRSA 1996 guidance, which requires that 340B drugs be dispensed only to patients with an established documented relationship with the covered entity for whom the covered entity maintains records of care. For FQHCs with substance-use-disorder programs, 42 CFR Part 2 confidentiality protections apply on top of HIPAA, with stricter consent requirements for disclosure of patient identifying information related to SUD treatment; the billing workflow has to segregate SUD encounter data and apply the Part 2 consent rules before any release to a payer or third-party administrator. Both of those layers (340B compliance and 42 CFR Part 2) sit outside the strict NCR math but ride alongside it on every engagement and the framework treats them as gating controls on the cash story.

01
Pull twelve months of remits across all payers.
Group by payer. Identify PPS, wrap, and FFS payments.
02
Build the expected reimbursement model.
PPS rate by site. MCO fee schedules. Sliding fee schedule.
03
Compute NCR by payer for prior twelve months.
Cash collected over expected reimbursement. Lock the baseline.
04
Build the T1015 denial taxonomy.
Ten root cause categories. Rank by dollar impact on the panel.
05
Reconcile wrap dollars against UDS Table 9.
Identify the encounter pool with unposted or under-posted wrap.
06
Move board reporting onto NCR.
Retire GCR from the board pack. Add the wrap reconciliation page.
07
Segment BH and SUD inside the NCR.
BH encounter pool. BH-specific payer mix. BH NCR as a sub-metric.
08
Stand up the monthly NCR review ritual.
CFO and senior partner. NCR by payer. Wrap recovery filings.

GlossaryThe vocabulary of FQHC finance.

FQHC
Federally Qualified Health Center. A community health center that bills PPS.
PPS
Prospective Payment System. The flat per-encounter rate FQHCs are paid.
T1015
The CMS encounter code FQHCs bill to claim the encounter rate.
Wrap
State payment that brings an MCO-paid visit up to the PPS rate.
UDS
Uniform Data System. Annual report every FQHC files with HRSA.
Table 9
The UDS table that anchors encounter counts by payer.

About the authorsWho wrote this paper.

Aparna Suresh
Senior partner · BACB co-author · ASP-RCM
Twenty-plus years across FQHC, hospital, and specialty RCM. Founded ASP-RCM in 2019 and built the FQHC senior partner team behind this framework.
ASP-RCM FQHC team
Finance · UDS reporting · Wrap recovery
The cross-functional team responsible for the NCR dashboards, the T1015 taxonomy, and the wrap reconciliation playbook anchored in this paper.

Common questionsFrequently asked: FQHC NCR.

Why is GCR the wrong lens for an FQHC?
GCR divides cash collected by gross charges. For an FQHC paid on PPS, the gross charge has almost no relationship to the cash the encounter generates. PPS pays a flat encounter rate. The gross-to-net ratio is distorted by the encounter rate floor, not by collection performance. NCR is the right lens because it divides cash collected by expected reimbursement on the actual payer mix, and the expected number for an FQHC encounter is the PPS rate plus the wrap.
What is the T1015 code and why does it matter so much?
T1015 is the FQHC encounter code used to bill the encounter rate. It is the single highest-frequency line on an FQHC remit. When it denies, the encounter denies. T1015 denials concentrate in roughly ten root cause categories, and a small number of those categories explain the majority of the denial dollars. Knowing the T1015 taxonomy is the first step in any FQHC AR program.
What is the PPS wrap?
For FQHCs serving Medicaid managed care members, the MCO pays a contracted fee schedule for the visit. The state then pays the FQHC the difference between that MCO payment and the PPS rate. That second payment is the wrap. Reconciling the wrap against UDS Table 9 is the discipline that determines whether the FQHC was paid the full PPS rate for every encounter.
How does UDS Table 9 fit?
UDS Table 9 is the annual encounter and payer-mix report every FQHC files. The wrap reconciliation is anchored to Table 9 because Table 9 is the single internal source of truth for encounter counts by payer. If the wrap dollars do not reconcile to Table 9 encounter counts, the gap is the audit finding.
How does the framework handle SUD and behavioral health inside an FQHC?
FQHCs with behavioral health programs face an additional layer because BH encounters carry distinct payer rules. The framework handles SUD and BH inside the same NCR lens by segmenting the BH encounter pool, recalculating expected reimbursement on the BH-specific payer mix, and treating the BH NCR as a sub-metric.
Does this framework apply to look-alikes and non-FQHC community health centers?
Partially. FQHC look-alikes that bill PPS get most of the benefit. Community health centers that do not bill PPS revert to a fee-schedule lens where the NCR vs GCR distinction is less material. The denial taxonomy still applies.
What is the worked example?
An anonymized seven-site FQHC across one state. Pre-engagement GCR sat at 47 percent. NCR sat at 76 percent. T1015 denials concentrated in two categories. After the framework ran for six months, NCR moved to 92 percent. Wrap dollars increased by roughly $1.8M annualized.
Does ASP-RCM replace the FQHC billing team?
No. The framework runs as a working cadence with the in-house billing team. ASP-RCM provides the dashboards, the denial taxonomy, the wrap reconciliation playbook, and the senior partner oversight. The billing team owns the work.

Want this framework applied to your FQHC?

Send twelve months of remits and your last UDS Table 9. Inside 30 days, a written NCR baseline, a T1015 denial taxonomy with dollar impact, and a wrap reconciliation gap report. Yours to keep.