Lease-Accounting Data and Calculation Control Starts Before the Calculation

Lease-Accounting Data and Calculation Control Starts Before the Calculation

At period close, the controller may receive a clean lease-liability schedule while still lacking confidence in the inputs behind it. Was an amendment captured? Did an extension option change the lease term? Was the approved discount rate applied to the correct entity? Lease-accounting data and calculation control therefore cannot begin with recalculating the schedule. It must begin by controlling how contract evidence becomes an approved accounting input.

A lease calculation is only controlled when every judgment-sensitive input can be traced to its source, policy and approval.

The hidden risk is input authority

The systems burden is persistent, not a one-time implementation issue. In its 2025 post-implementation review, FASB concluded that lessees’ initial and ongoing Topic 842 costs were significantly higher than expected, primarily because existing systems and processes often could not support balance-sheet accounting for operating leases. Organizations devoted significant effort to identifying complete lease populations, validating provisions and moving lease information into accounting records. FASB also reported more than 370 technical inquiries and eight clarifying or targeted Accounting Standards Updates, underscoring the continuing interpretation burden around recognition, measurement and related issues. Post-Implementation Review: Leases (Topic 842)

The operating failure occurs when the lease subledger or spreadsheet is treated as authoritative even though essential inputs originated elsewhere. Structured payment schedules and entity codes must be reconciled with semi-structured contracts and amendments, plus unstructured explanations of options, modifications and accounting judgments. This is a mixed-data problem because the calculation is unsafe until those sources agree on lease term, payment stream, modification date, discount rate, classification and organizational ownership. Missing context can propagate into right-of-use assets, liabilities, expenses, journal entries and disclosures.

The consequence can be material. Sonder Holdings reported errors involving lease-term and modification inputs, computations, valuation, payments and previously unrecorded right-of-use asset impairments. It restated prior financial statements and recorded $82.1 million of previously unrecognized 2022 impairment losses. Sonder also disclosed a material weakness in capturing and recording lease agreements timely and accurately. This is a company-specific example from a lease-intensive hospitality business—not an industry benchmark—but it demonstrates why calculation review alone is insufficient. Sonder Holdings 2023 Annual Report

Redesign lease-accounting data and calculation control

The controller needs a workflow that establishes input lineage before posting. A practical target state has four stages:

1. Capture source evidence. Ingest executed contracts, amendments, termination notices and approved rate tables; associate them with legal entity, location, counterparty and lease identifier.
2. Validate and normalize inputs. Compare dates, payment schedules, options and modifications across documents and existing lease records. Flag missing fields, duplicate leases, inconsistent identifiers and conflicts between source versions.
3. Apply policy and calculate. Map approved accounting rules to validated inputs, calculate the schedule, and route judgment-sensitive or low-confidence cases to an exception queue with the relevant contract clauses and prior decisions.
4. Approve and deliver. Record the reviewer, evidence, rationale and policy version; then send approved schedules and journal-ready output to the lease subledger, ERP or general ledger while retaining reconciliation history.

Automation should stop when source documents conflict on a term that affects recognition or measurement, or when a modification requires policy judgment. The accounting policy owner—not the processor—should receive the conflicting clauses, proposed input, calculation impact and prior approval history. The only permitted next action should be approval, correction or return for additional evidence; no posting should occur while the exception remains unresolved.

StarDox Intelligence can serve as the enterprise automation and decision-intelligence layer across this workflow by applying policy, evidence and risk rules, prioritizing exceptions, and preserving explainable approvals and decisions. Its role is not to replace accounting policy, professional judgment or the system of record.

Three actions for the controller

Progress should be measured through control performance, not document-processing volume alone. Useful baselines include first-time-right processing, exception rate, approval cycle time, unresolved exceptions at close and the percentage of posted schedules with complete source-to-approval traceability.

1. Map source authority for every material lease input, including which document, policy owner or system prevails when records conflict.
2. Baseline exception flow by measuring how many leases require manual intervention, why they fail and how long each approval stage takes.
3. Define posting boundaries for missing amendments, conflicting terms, low-confidence extraction and judgment-sensitive modifications.

Together, these three actions reveal whether the principal constraint is capture, policy application, exception handling or downstream delivery.

Sources

• Post-Implementation Review: Leases (Topic 842)
• Sonder Holdings 2023 Annual Report