Glossary

Reconciliation

Reconciliation is the process of comparing two records of the same financial or business activity, identifying differences, and determining whether each difference reflects timing, missing information, an error, or a valid adjustment. The records may agree after supported adjustments, while documented outstanding differences can remain open for follow-up.

Why reconciliation matters

A balance can look reasonable while still containing duplicate entries, omitted transactions, posting errors, or activity recorded in different periods. Reconciliation tests a record against another source that represents the same activity. That comparison helps finance and operations teams find discrepancies before those differences flow into reporting, settlement, cash management, or the accounting close.

A completed reconciliation also creates a clear account of what matched, what changed, and what still needs attention. It does not prove that every underlying transaction is valid. The strength of the control depends on the records compared, the matching criteria, the review performed, and the treatment of exceptions.

Simetrik supports reconciliation by applying configured transaction matching rules and surfacing unmatched records or exceptions for review. Its deterministic control core produces repeatable results from the same inputs and rules, while traceability around configured controls and actions supports the investigation of exceptions.

How reconciliation works

  1. Define the scope. Choose the account, process, entity, period, and two records that should represent the same activity. Examples include a general ledger account and a subledger, an internal cash record and a bank statement, or sales records and payment settlements.
  2. Prepare the records. Bring both records into a comparable form. Check that dates, amounts, currencies, identifiers, signs, and period boundaries are interpreted consistently before matching begins.
  3. Compare and investigate. Match corresponding items and isolate differences. Determine whether each difference comes from timing, an unrecorded item, a duplicate, a classification problem, a data issue, or another explainable cause.
  4. Adjust, document, and review. Correct supported recording errors, record valid adjustments, document outstanding items, and compare the updated balances again. Unresolved exceptions remain visible for follow-up and review.

Bank reconciliation example

Suppose a company’s cash ledger shows $100,000, while its bank statement shows $101,250. The team finds a $1,500 deposit recorded in the ledger that has not yet reached the bank, $3,000 in issued checks that have not yet cleared, and a $250 bank fee missing from the ledger. The adjusted bank balance is $101,250 + $1,500 – $3,000 = $99,750. After recording the bank fee, the adjusted ledger balance is $100,000 – $250 = $99,750. The adjusted balances agree, while the deposit and checks remain documented as timing items until they clear.

What can be reconciled?

Reconciliation is broader than bank reconciliation. Businesses may reconcile accounts receivable to customer activity, accounts payable to vendor records, subledgers to the general ledger, intercompany balances between entities, inventory records to physical or operational counts, and payment or settlement data across systems. In each case, the records must describe comparable activity and use a clearly defined scope.

Common causes of reconciliation differences

  • Timing differences, such as a payment recorded internally before it settles externally.
  • Missing or duplicated transactions in one record.
  • Incorrect amounts, dates, signs, currencies, or account classifications.
  • Fees, credits, reversals, or adjustments recorded by only one party.
  • Different identifiers or data formats that prevent corresponding items from matching.

Manual and automated reconciliation

Manual reconciliation may be practical when the data set is small and the comparison is simple. As volume, sources, entities, or rule complexity increase, software can apply configured matching logic consistently and separate matched items from exceptions. Human review is still needed to assess unusual differences, support adjustments, and decide how unresolved items should be handled. Automation does not remove the need for sound rules, complete records, and accountable review.

Frequently asked questions

What is the difference between transaction matching and reconciliation?

Transaction matching pairs records that appear to represent the same event. Reconciliation is the broader process and control: it defines the scope, compares balances or activity, investigates unmatched items, records supported adjustments, and documents exceptions for follow-up.

What are common types of reconciliation?

Common types include bank, accounts receivable, accounts payable, intercompany, inventory, payment, and subledger to general ledger reconciliation. The useful classification depends on which records and business activity are being compared.

What happens when a reconciliation difference cannot be resolved?

The item should remain documented as an exception, with its amount, period, available evidence, investigation status, and next review step. An unexplained difference should not be treated as matched simply because the overall balance is close.

How often should reconciliation be performed?

Frequency should reflect the account’s risk, transaction volume, reporting cycle, and how quickly differences need attention. High activity cash or settlement records may need more frequent review than a low activity account. The chosen cadence should be documented and applied consistently.

How can Simetrik support a reconciliation workflow?

Simetrik is an AI-native financial control platform that applies deterministic matching rules and surfaces unmatched records or exceptions for review. AI supports data preparation, suggestions, and review within the platform, while the deterministic core governs control execution. Traceability connects configured controls and actions with the records and review decisions.

Can AI be used in a Simetrik reconciliation?

Yes. AI is part of Simetrik’s AI-native platform and supports data preparation, transformations, matching suggestions, analysis, and review. Its deterministic core executes reconciliation controls using the configured rules and inputs. Teams review suggestions and exceptions and retain responsibility for the accounting decisions.

Ready to transform your reconciliation workflows?

This site is registered on wpml.org as a development site. Switch to a production site key to remove this banner.