How are accruals and prepaids handled in a recurring CAS close?
Schedule-driven entries — prepaid amortization, straddling payroll, interest, commissions — are computed from supplied source documents using a method the CPA firm has approved. Judgment estimates and materiality decisions stay with the firm. Each entry is documented so the number can be agreed rather than rebuilt.
Accruals and prepaids are where a recurring close most often stops being recurring. The posting itself is trivial; what is not trivial is that half these entries can be derived by anyone from a document, and the other half rest on an assumption that belongs to the CPA firm. Treating the two as one task is what produces both the slow closes and the entries nobody can explain six months later.
This sits inside CAS production support and depends on the conventions established at onboarding.
Two kinds of entry that look identical in the ledger
A prepaid insurance amortization and an accrual for unbilled consulting both appear as a journal entry with a description and an amount. They are not the same kind of thing:
- Derived from a document. Prepaid amortization, a lease straight-lined over its term, payroll apportioned across a period end, interest between statement dates. The source exists, the method is written down, and two preparers reach the same figure.
- Derived from an assumption. How much vendor work has been performed but not invoiced, what proportion of a bonus pool has been earned, whether a receivable is collectible. The number depends on a position someone has to take.
Preparation work owns the first category end to end. For the second it can assemble the evidence and quantify the exposure, but the method and the assumption come from the firm before anything is posted.
What a prepaid schedule has to tie to
A prepaid schedule that only reconciles to the ledger is doing half its job. It should tie in three directions at once: to the account balance, to the invoice or contract that created it, and to the expense recognized this period. It should also show the remaining term, so the reviewer can see where the balance is going without recomputing it.
The common failure is a schedule inherited from a previous preparer that foots perfectly and no longer traces to anything. It is not wrong on its face, which is exactly why it survives — and it is unsupported, which only becomes visible when someone asks for the original document.
The four sources a close should check every period
Most missed accruals are missed because the close relies on somebody remembering. The same four places produce the overwhelming majority of them, and checking all four each period is cheaper than recalling any of them:
- The payroll register, for the days falling either side of the period end
- Open purchase orders and vendor confirmations, for work performed but not yet billed
- Loan and financing agreements, for interest between statement dates
- Commission and bonus plans, for amounts earned in the period and paid later
The standing entry that quietly stops being true
A recurring accrual posting the same amount every month is the most comfortable entry in a ledger and one of the most dangerous. It began as an estimate derived from something; once it repeats unexamined it becomes an assumption, and the driver underneath it moves — a contract ends, a rate is renegotiated, a team changes size. The entry keeps looking reasonable and stops being right.
The cheap protection is to make a recurring entry carry its own provenance: the source it was derived from, and the date it was last re-derived. A reviewer can then see its age, which is the thing that actually predicts whether it is still correct.
Reversals are procedural, not analytical
An unreversed accrual counts the same cost twice, and it is unusually hard to find afterwards because both the accrual and the later invoice are individually defensible. No amount of reviewing the entries in isolation surfaces it.
What surfaces it is procedure: the reversal is scheduled at the moment the accrual is booked rather than remembered later, and the accrual accounts are checked for residual balances as part of closing the period — the same discipline that applies to clearing accounts in ledger hygiene.
What stays with the firm
Materiality is a professional judgment about the statements as a whole and about who relies on them, so whether an item is large enough to book is not a preparation decision. Neither is the estimate methodology, nor the final sign-off on the period.
What preparation can do is make those judgments cheap to make: surface the item, quantify it, attach its source, and state plainly what is unknown. Where the support has not arrived, the item belongs on the exception list with its amount and what is missing — not booked on a guess, which looks finished, and not skipped, which looks like nothing happened.