Inventory observation: bridge count-date evidence to year-end
A count observed before year-end does not automatically establish the year-end inventory quantity. The bridge depends on the intervening movements and their cutoff. A useful workpaper starts with a specific item and location, tests both incoming and outgoing transactions, and retains discrepancies until evidence explains them.
This guide concerns an issuer audit under PCAOB standards, using AS 2510 as displayed on October 25, 2026. The fictional count and records illustrate an evidence method, not a claim that a public investor can conduct or reproduce the issuer’s inventory observation. The case concerns quantities; ownership and valuation remain separate questions.
Establish what the count actually measures
AS 2510 .09–.12 discuss observation, the reliability of counting methods and intervening-transaction testing. The standard expressly limits its subject to observation rather than every procedure needed for inventories.
Record item code, unit of measure, location, count date and time, movement-freeze arrangements, and count-sheet identifier. “A warehouse was visited” does not establish that the selected item was counted, that all locations were covered, or that the recorded condition was acceptable.
A quantity can be counted accurately in pallets while the ledger uses individual units. Retain the conversion factor and inspect whether pallet contents vary. A spreadsheet error in unit conversion can look like a physical shortage even when the goods and counting observations agree.
Also distinguish tracing recorded items to the floor from tracing floor items to records. The former and latter address different risks of missing or nonexistent entries. A selected count test is not automatically a census of every inventory position.
Work a forward bridge
Assume one fictional item measured in individual units at one location. Its December 20 physical count is 1,000. From after that count through December 31, independently tested receipts add 200 and tested dispatches remove 150. Assume no production conversion, scrap, returns, inter-warehouse transfers, third-party goods, or other movements in this simplified case.
Expected year-end quantity = 1,000 + 200 − 150 = 1,050.
| Bridge component | Units | Evidence to retain |
|---|---|---|
| December 20 count | 1,000 | Signed count record and relevant observation |
| Subsequent receipts | +200 | Dated receipt and underlying movement evidence |
| Subsequent dispatches | −150 | Dispatch record and relevant supporting documents |
| Expected December 31 quantity | 1,050 | Recomputed bridge |
| December 31 book quantity | 1,080 | Same item, location and unit |
| Unexplained book excess | 30 | Investigation required |
The 30 is an unexplained quantity difference, not yet a theft finding, a valuation loss or an accounting correction. Its nature follows from the investigation.
Suppose the true 200 receipts include a single receipt of 30 entered twice in the ledger. The books therefore show receipts of 230. The incorrect roll-forward is 1,000 + 230 − 150 = 1,080. Removing the documented duplicate of 30 returns the book quantity to 1,050.
That explanation requires the original receiving evidence, unique document numbers, posting dates and confirmation that there was only one physical receipt. A repeated reference number alone could reflect a valid split shipment. Do not book a correction merely to force agreement with the theoretical figure.
Test cutoff rather than relying on timestamps alone
The count-to-year-end movement window must begin after the count cutoff. A December 20 receipt already included physically at the count cannot be added again because it was posted later. A dispatch physically completed before the count cannot also be deducted as an intervening movement.
Create a chronology using actual movement time, document date and posting time as separate fields. Those dates can differ legitimately. Resolve which event determines the quantity bridge and separately assess the contractual recognition consequences.
For instance, a posting on December 31 for goods physically received January 2 is not a December physical receipt. Whether those goods should nevertheless be recognized as goods in transit depends on rights and applicable accounting evidence. Do not solve that ownership question by placing nonexistent warehouse units into the physical bridge.
Investigate transactions around both ends of the window, not just the last day of the year. A clean December 31 cutoff will not repair a double-counted December 20 movement.
Reverse the bridge for a later count
In a separate hypothetical case, the count occurs on January 10 and establishes 1,120 units. Tested January 1–10 receipts are 170 and dispatches are 100. There are no other movements.
December 31 quantity = 1,120 − 170 + 100 = 1,050.
Receipts after year-end are removed because they were not present at year-end; later dispatches are added back because they reduced the later count. Reversing only receipts would produce 950 and ignore the 100 units that left after year-end.
AS 2510 .12–.13 provide context for physical evidence and intervening records, including circumstances involving prior inventories. The arithmetic does not by itself establish that those records or counting procedures are reliable.
Save forward and reverse bridges as different dated models. A later count does not magically observe the earlier date. The evidence must support the intervening population, including potentially unrecorded movements.
Ownership and condition are different mechanisms
Now alter the forward case’s assumptions: the December 20 physical 1,000 includes 80 units owned by an unrelated party and held in custody. Assume the remaining movements relate only to the issuer’s goods. The physical count remains 1,000, but the issuer-owned starting quantity is 920 and its simple owned-quantity bridge is 920 + 200 − 150 = 970.
The 80 did not disappear. It never belonged in that assumed asset population. Custody contracts and ownership documentation, not more accurate counting, resolve this difference.
Likewise, a counted unit can be damaged or obsolete. Accurate quantity does not establish recoverable carrying value. Separate physical condition observations, cost records and applicable valuation evidence. The inventory’s monetary balance cannot be verified by multiplying every counted unit by an unsupported selling price.
For outside warehouses, AS 2510 .14 discusses direct custodian confirmation and additional procedures in relevant circumstances. A warehouse receipt or confirmation addresses particular evidence questions; it does not automatically remove rights, pledge or reliability issues.
Execute a movement-level checklist
| Step | Input and procedure | Stop condition |
|---|---|---|
| Define scope | Item, site, units and count timestamp | Records refer to another population |
| Establish count | Count sheets and observation evidence | Count reliability unresolved |
| Identify movements | Complete receipt and dispatch sequence | Missing sequence or unexplained gap |
| Test both endpoints | Physical, document and posting dates | Duplicate or cross-period item unresolved |
| Recompute | Opening count plus all applicable movements | Unsupported adjustment needed to agree |
| Separate assertions | Ownership, condition and valuation files | Quantity used as universal proof |
Public filings may reveal inventory balances, locations, auditor descriptions, custody arrangements or known count limitations. They rarely provide complete count sheets, gate logs and movement ledgers. An investor can assess disclosed scope and identify missing evidence without claiming to have verified 1,050 units at a real issuer.
When sources disagree, first align units, location and count scope, then inspect duplicate receipts, unposted dispatches, transfers and cutoff. Keep a separate unresolved-difference row rather than altering the count to equal the ledger.
Update the bridge when additional movement records, a revised count, a custody contract or a corrected posting becomes available. Retain the original discrepancy and the exact document that resolved it. A balanced quantity bridge is a completed calculation only to the extent its inputs and assertion-specific evidence have also been established.
Separate a warehouse count from goods in transit
In another alternative, the ledger balance of 1,080 includes 30 units in transit outside the counted warehouse. Assume independent contracts and transport evidence establish the issuer’s rights at year-end and that those units are not among the 1,050 in the warehouse. Then 1,050 warehouse units plus 30 transit units can reconcile a broader inventory population of 1,080.
This is not the duplicate-receipt correction in the main example. Removing 30 solely to force the ledger to match the warehouse would delete a different supported position. Keep the main case’s same-location assumption and the alternative’s broader population explicit.
Use separate location codes for owned in-transit goods, warehouse goods and third-party custody. Reconcile transfers between codes so the same shipment is not counted at both departure and destination. Missing rights evidence should remain unresolved; physical absence alone cannot decide ownership.
FAQ
Does a pre-year-end count prove the year-end amount?
No. The intervening movements and cutoff require evidence.
Do goods physically present always belong to the issuer?
No. Third-party custody goods can exist without being the issuer’s assets.