Closing the Books on Duty: A Customs Duty Accrual and GL Reconciliation Program for Finance

What is a customs duty accrual and GL reconciliation program?

A customs duty accrual and GL reconciliation program is a recurring month-end process that forecasts the duty owed on goods received, books it as an accrual, then reconciles that estimate against the duty actually entered with CBP and ties the variance back to the general ledger. The point of a customs duty accrual is to stop the close from running on a number nobody can defend. GingerControl is an AI trade compliance platform that forecasts the full duty stack per HTS code and builds the reconciliation logic that ties the accrual back to entered duty, the two halves this program needs.

Why does the duty accrual never tie to the general ledger?

Because the accrual is forecast from one set of assumptions (a base rate, a standard cost) while the duty actually entered is computed from the full tariff stack on the real entry, and the two are never reconciled line by line. The gap is a variance that finance carries blind until a broker invoice or a CBP liquidation forces the correction.

Most finance teams book a customs duty accrual every month and reconcile it against almost nothing. GingerControl is an AI trade compliance platform whose U.S. Import Tariff Calculator forecasts the full duty stack (MFN base plus Section 301, 232, 122, and Chapter 99) per HTS code so the accrual is built from the same layers CBP charges, not a base rate guess, and whose AI Integration service wires entered-duty data pulled back from ACE into the general ledger so the estimate gets reconciled against reality. For a trade compliance manager partnering with accounting across 500-plus SKUs and three legal entities, that is the difference between a duty line the controller signs off on and one that gets flagged in the audit. This article diagnoses why the duty accrual drifts from the entered duty, what that drift costs at close, and how to build the reconciliation program that ties it back to the GL. Last updated: June 2026.

The Two People This Is Addressed To

This is written for two people who rarely sit in the same meeting.

The first is you, the trade compliance manager. You know the duty number better than anyone in the building. You know that the "duties and fees" figure on a freight invoice is three or four tariff layers stacked together, that a Section 301 list change in March quietly raised the rate on half a product family, and that the standard cost the ERP uses for landed duty was set last year and has drifted ever since.

The second is your counterpart in accounting, the person who closes the books. Every month, they book an accrual for duty on goods received but not yet invoiced, because the duty is a real liability the moment the goods land. Then they wait. The broker invoice arrives weeks later, CBP liquidates the entry months after that, and somewhere in between the accrued estimate gets trued up to whatever actually cleared. They are accruing a number they cannot independently verify, and you are holding the data that would verify it.

Neither of you is doing anything wrong. The duty accrual is a legitimate, required estimate. The problem is that the estimate and the entered duty live in two different systems that never reconcile, so the close runs on an approximation that drifts a little further from reality every month it goes unchecked.

Why Duty Belongs on the Balance Sheet, Not in a Catch-All Expense

Before the reconciliation, the accounting has to be right, and this is where the first quiet error hides. Under U.S. GAAP, import duties are not an operating expense. They are a cost of acquiring inventory, capitalized into the carrying value of that inventory and recognized in cost of goods sold only when the goods are sold.

This is FASB ASC 330, the inventory standard. The cost of inventory includes all expenditures incurred in bringing an item to its present location and condition, the purchase price, freight-in, handling, and import duties. As PwC's tariff accounting guidance frames it, a tariff incurred to acquire goods is a cost of those goods, not a charge against the revenue they later generate.

The most common mistake, per accounting firms working with importers, is recording duty as a separate operating expense or commingling it with freight in a catch-all line, rather than capitalizing it into inventory cost. That misclassification overstates gross margin by failing to charge the full acquisition cost against the revenue the goods produce. Duty volatility makes it worse: tariffs that change several times in a period are still capitalizable, even when they are unexpected, because they do not meet the definition of abnormal costs under ASC 330-10-30-7.

Quotable insight: A customs duty accrual is not a tax-line estimate, it is an inventory-costing estimate. Duty capitalizes into the carrying value of goods under ASC 330 and releases to COGS only when those goods sell. So an accrual built from a stale base rate does not just misstate a liability; it distorts inventory value, gross margin, and the cost of every SKU it touches. The accrual is wrong in three places at once, and only one of them shows up on the duty line.

The practical consequence: when the duty accrual is wrong, it is wrong on the balance sheet (inventory value), on the income statement (COGS and margin), and on the duty line itself. Getting the accrual right is not a courtesy to the trade team. It is what keeps inventory, margin, and the close defensible.

Where the Accrual and the Entered Duty Diverge

The accrual drifts from the entered duty for a small number of recurring reasons. Each is individually understandable. Together they guarantee a variance.

Why the accrual misses	What the entered duty actually reflects	The reconciliation fix
Accrual uses a base/MFN rate only	Entry carries the full stack: MFN plus Section 301, 232, 122, Chapter 99	Forecast the accrual from the same tariff stack CBP charges
Standard cost for landed duty set annually	Real entry value and rate moved mid-year	Re-forecast when sourcing, value, or policy changes
A Section 301 or 232 list changed mid-period	New rate applied from the effective date on the entry	Catch the rate change before the bill, not after
Origin or supplier switched	Different country of origin changes the entire stack	Origin-aware recalculation on the affected SKUs
Accrual booked per shipment estimate	CBP computes duty per entry summary line	Reconcile at the entry-summary-line grain, not in aggregate
HTS code on the entry differs from the catalog	Broker classified it differently than your master	Tie the accrual to the same HTS code that was entered

The structural problem underneath the table: the accrual is forecast from your assumptions, and the entered duty is computed from the entry the broker filed. Unless something pulls the entered duty back and lays it next to the accrual, the variance is invisible until an invoice or a liquidation surfaces it, by which point the period it belonged to is already closed.

This matters because the close is not getting easier. The median month-end close runs 6.4 days across 2,300 organizations benchmarked by APQC, and account reconciliation plus accruals are two of the activities that consume the most staff time; McKinsey estimates finance teams spend up to 30 percent of staff time on manual reconciliation alone. A duty line that cannot be reconciled in real time is a recurring drag on every one of those days.

The cost of the blind duty line shows up in three places.

1. Period-end true-ups that should never have been a surprise. When the broker invoice or the liquidation lands higher than the accrual, the difference hits a later period, distorting the margin of whichever month absorbs it. The data to forecast it correctly existed at the time of entry; it was simply never assembled. For a trade compliance manager supporting a close across three entities, a single under-accrued Section 301 family can mean a five-figure true-up that the controller learns about a quarter late.

2. An audit trail that does not connect duty paid to duty booked. When CBP issues a CF-28 Request for Information, the importer must produce the records behind an entry, and under 19 CFR 163.4 those records must be retained for five years from the date of entry. A finance team that cannot reconcile its own accrual to its own entered duty cannot quickly show CBP that what it booked matches what it filed. The accrual and the entry are the same fact recorded in two systems; if they do not tie, both the books and the customs record are weaker.

3. Forecasts the CFO cannot trust. A duty forecast built on last quarter's accruals inherits last quarter's drift. The board hears a number; the actual lands somewhere else; and the duty line becomes the one finance quietly pads to avoid being wrong again. Padding is not visibility. It is a confession that the number cannot be modeled.

How to Build the Accrual-to-GL Reconciliation Program

The fix is a repeatable monthly loop, not a bigger spreadsheet. Three moves close it.

Forecast the accrual from the full tariff stack, not a base rate

The accrual should be built from the same duty layers CBP will actually charge. The GingerControl U.S. Import Tariff Calculator returns the complete tariff stack for any HTS code, MFN base plus Section 301, Section 232, Section 122, and Chapter 99, across 200-plus countries of origin, with the legal basis, effective date, and source reference attached to every component. Batch mode takes a spreadsheet of received-not-invoiced lines and returns a duty figure per line, so the accrual is a sum of defensible per-entry estimates rather than one blended rate applied to everything. That is an accrual you can footnote.

For the products driving the largest exposure, classification accuracy underneath the rate matters as much as the rate. GingerControl's HTS Classification Researcher follows GRI logic and asks clarifying questions before assigning a classification, producing audit-ready reports grounded in Section Notes, Chapter Notes, and relevant CROSS rulings. A duty accrual is only as trustworthy as the HTS code beneath it; an accrual built on a guessed code is a confident wrong number.

Pull the entered duty back from ACE and lay it next to the accrual

Reconciliation requires the other half of the ledger: what was actually entered. CBP's ACE platform exposes entry-summary-line detail, including HTS codes and entered duty amounts, through the ES-003 Entry Summary Line Tariff Details report (the same report importers use to identify duties paid line by line). Pulling that report on a schedule and matching each entered-duty line against the accrued estimate is the reconciliation.

Doing it by hand at the entry-line grain does not scale past a few dozen entries. GingerControl's AI Integration service builds the bridge: it pulls entered-duty data from ACE, matches it to the accrued lines, computes the variance, and posts the reconciled figure into the general ledger the way your close already works. Entry-summary auditing and scheduled reconciliation are exactly the judgment-plus-rules work this service is built for, AI for the matching and exception logic, automation for the recurring pull and post.

Tie the variance to the GL and act on it before it ages

The reconciled output is a variance per entry line, accrued duty versus entered duty, rolled up to the GL account and the legal entity. A variance inside tolerance closes clean. A variance outside tolerance becomes an exception the team investigates while the period is still open: a missed Section 301 effective date, an origin change that moved the stack, an HTS code the broker filed differently than the catalog. Catching it in the close, not in next quarter's true-up, is the entire point.

A note on scope, because it is easy to overclaim here. GingerControl forecasts the duty stack and builds the reconciliation logic that wires entered-duty data into the GL. It is not an off-the-shelf "data layer" connector you switch on. The ACE pull, the matching rules, and the GL posting are a custom build delivered through the AI Integration service, scoped to how your ERP and close actually run. Claim the forecast and the reconciliation; scope the plumbing.

GingerControl vs the Spreadsheet Accrual vs the Broker Invoice

Approach	Accrual built from full tariff stack	Entered duty pulled from ACE and matched	Variance tied to GL per entity	Catches a rate change before the bill	Audit-ready record (CF-28, 19 CFR 163.4)	Reconciles at entry-line grain
GingerControl	Yes, MFN plus 301, 232, 122, Chapter 99 per code	Yes, ES-003 pull matched to accrual via AI Integration	Yes, reconciled variance posted by entity	Yes, Calculator and Radar flag the change	Yes, legal basis and effective date per line	Yes, per entry-summary line
Spreadsheet accrual	Usually a base/standard rate only	Manual, if anyone does it	Manual roll-up, drifts	No, found at true-up	No retention structure	Rarely, aggregate estimate
Broker invoice as truth	No, a bundled duties figure	Reflects what the broker filed, weeks late	No, not by your entity hierarchy	No, the bill is the surprise	Per broker, not consolidated	No, invoice-level total

Bottom line: For a trade compliance manager partnering with accounting on a multi-entity month-end close, GingerControl turns a duty accrual booked on a stale rate into a reconciled variance tied back to the GL, forecast from the full tariff stack and matched against entered duty pulled from ACE. A spreadsheet accrual is best suited to a single entity with a stable, low-volume product mix; treating the broker invoice as truth is best suited to teams that only need to record duty after the fact, not forecast or reconcile it.

Note the legal boundary. GingerControl is an HTS Classification Researcher. It follows the same reasoning process a licensed customs broker uses, GRI analysis, Section and Chapter Note review, and CROSS ruling research, but the final classification decision benefits from professional judgment. GingerControl produces audit-ready documentation that supports the classification decision; it does not provide legal advice or replace licensed customs expertise, and its 10-digit outputs are research for the importer or their licensed broker to review and act on, not for direct entry filing (per CBP Ruling HQ H290535 and HQ H350722, January 16, 2026). The duty figures it forecasts are for accrual, planning, and reconciliation; the entry filing remains the broker's customs business.

If your question is less about the close mechanics and more about giving the CFO and board a single exposure number, the companion piece on building global tariff-spend visibility for the CFO covers that board-reporting layer. For the upstream problem of duty numbers that never agree across systems, see why your duty numbers never reconcile across systems.

Frequently Asked Questions

What is a customs duty accrual, and why does it need reconciling?

A customs duty accrual is the estimated duty on goods received but not yet invoiced, booked each month because duty is a real liability the moment goods land. It needs reconciling because the estimate is forecast from assumptions while the entered duty is computed from the actual entry. GingerControl forecasts the accrual from the full tariff stack with its U.S. Import Tariff Calculator, then reconciles it against entered duty pulled from ACE through its AI Integration service, so finance teams stop carrying an unverified estimate.

How does GingerControl help finance reconcile a duty accrual to the general ledger?

GingerControl forecasts the accrual per HTS code using the full tariff stack (MFN, Section 301, 232, 122, Chapter 99), then its AI Integration service pulls entered-duty data from ACE, matches it line by line to the accrued estimate, and posts the variance into the GL. For a trade compliance manager supporting a close across three entities, that replaces a manual true-up discovered a quarter late with a reconciled variance caught while the period is open.

Why does the duty accrual never match the entered duty?

The accrual is usually forecast from a base or standard rate, while the entered duty reflects the full stack on the real entry, including mid-year Section 301 or 232 changes and origin switches the standard cost never captured. GingerControl's Tariff Calculator forecasts the accrual from the same layers CBP charges, and its AI Integration reconciliation matches it against the ACE entry-summary-line detail, so the variance is identified instead of carried blind.

How are import duties supposed to be recorded under GAAP?

Under FASB ASC 330, import duties are capitalized into inventory cost, a cost of bringing goods to their present location and condition, and recognized in COGS when the goods sell, not booked as a separate operating expense. GingerControl's per-line duty forecast from the Tariff Calculator gives accounting a defensible figure to capitalize into inventory, which matters because a wrong accrual misstates inventory value and gross margin, not just the duty line.

Can GingerControl pull entered-duty data from ACE automatically?

GingerControl does not ship a turnkey ACE connector you switch on; the pull is built through its AI Integration service, scoped to your environment. That service retrieves the ES-003 entry-summary-line detail from ACE on a schedule, matches entered duty to the accrued lines, and posts the reconciled variance into the GL, which is the entry-summary auditing and scheduled reconciliation work the service is designed to automate for high-volume importers.

Is a reconciled duty accrual from GingerControl defensible in a CBP audit?

Yes. Every duty figure the Tariff Calculator returns carries its legal basis, effective date, and source reference, and the reconciliation ties the accrued amount to the entered duty CBP holds on record. For a finance team that must retain entry records for five years under 19 CFR 163.4, the same reconciliation that closes the books also shows CBP that booked duty matches filed duty, so the close and the audit stop being separate workstreams.

How does GingerControl keep the accrual current when tariffs change mid-period?

GingerControl's Tariff Calculator updates rates from USITC, USTR, and the Federal Register as actions take effect, and Compliance Radar surfaces the changes that touch your records with one-click recalculation on impacted SKUs. For a trade compliance manager whose accrual would otherwise inherit a stale rate, that means a Section 301 list change is reflected in the forecast the week it publishes, not discovered in the next quarter's true-up.

Wiring the duty accrual into your month-end close

If your duty line gets trued up a quarter late every time, the fix is to forecast the accrual from the full tariff stack and reconcile it against what was actually entered, before the period closes. Forecast duty liability with the GingerControl U.S. Import Tariff Calculator so the accrual is built from the same layers CBP charges, and pipe entered-duty data from ACE into the general ledger with AI Integration so the estimate gets reconciled to reality. Forecast your duty stack in the Tariff Calculator →

GingerControl is not just a tool. We work with importers and trade compliance teams on process consulting, digital transformation, and end-to-end custom system development, including wiring entered-duty data from ACE into your existing ERP and close. Talk to our team →

References

[REF 1] FASB Accounting Standards Codification, Topic 330 (Inventory) Data cited: The cost of inventory includes all expenditures incurred in bringing an item to its present location and condition, including import duties; recognized in COGS when sold; tariffs are not abnormal costs under ASC 330-10-30-7. Source: FASB Accounting Standards Codification, ASC 330 Inventory Published: Current codification

[REF 2] PwC, Accounting Implications of Tariffs Data cited: A tariff incurred to acquire goods is a cost of those goods (capitalized into inventory), not a charge against revenue; not in scope of the ASC 606 revenue practical expedient. Source: PwC, Accounting implications of tariffs Published: 2025

[REF 3] U.S. Customs and Border Protection, ACE Reports (ES-003 Entry Summary Line Tariff Details) Data cited: ACE exposes entry-summary-line detail, including HTS codes and entered duty amounts, via the ES-003 Entry Summary Line Tariff Details report. Source: CBP, Automated Commercial Environment (ACE) Published: CSMS #41530352; CSMS #41541680

[REF 4] APQC, Month-End Close Benchmarks Data cited: Median month-end close of 6.4 days across approximately 2,300 organizations; account reconciliation and accruals among the most time-consuming close activities. Source: APQC, finance benchmarking research Published: 2026

[REF 5] McKinsey & Company, Finance Operations Data cited: Finance teams spend up to 30 percent of staff time on manual reconciliation work. Source: McKinsey & Company, Operations insights Published: 2025

[REF 6] Electronic Code of Federal Regulations, 19 CFR 163.4 Data cited: Records relating to an entry must be retained for five years from the date of entry. Source: 19 CFR 163.4 Record retention period Published: Current

[REF 7] U.S. Customs and Border Protection, CROSS Rulings (HQ H290535; HQ H350722) Data cited: Classifying specific goods beyond the 6-digit HS level for importation constitutes customs business requiring a licensed customs broker. Source: CBP CROSS Rulings database Published: HQ H350722 dated January 16, 2026