Late write?downs from fragmented signals; SAP inventory, analytics sell?through, and ICIS prices drove proactive reserves

Overview

A chemical producer repeatedly booked late inventory write?downs because sell?through signals and ERP stock data were reconciled only at close. Finance worked from aging reports while Sales and Supply Chain looked at separate velocity views, so reserves were reactive and adjustments spilled into the next period. Intelligex built a pipeline that unified ERP inventory, sell?through and forecast data from the BI layer, and external market prices, then generated reserve proposals with finance?owned rules and approvals. Reserve decisions became evidence?based, post?close adjustments diminished, and Supply Chain engaged earlier with clear exposure and disposition options—without replacing the ERP, BI stack, or planning tools.

Client Profile

• Industry: Chemicals (intermediates, specialties, and blends)
• Company size (range): Enterprise manufacturing and distribution footprint
• Stage: ERP in place for inventory and costing; sell?through tracked in BI; reserves handled via spreadsheets at close
• Department owner: Finance & Accounting (Cost Accounting and Controllership)
• Other stakeholders: Supply Chain/Planning, Sales, Plant Operations, Quality, Procurement, IT/Integrations, Internal Audit

The Challenge

Excess and obsolete exposure surfaced late. On?hand and in?transit inventory lived in the ERP by plant, batch, and item, but sell?through and forecast changes were analyzed in the BI layer with different product hierarchies. Throughout the month, teams debated whether slow?moving lots reflected seasonality, customer transitions, or true obsolescence. At close, Finance reconstructed a reserve by stitching together aging reports, sales velocity exports, and manual price checks. Adjustments often required rework the next period when late sales or price updates arrived.

Pricing and quality added complexity. Standard cost and recent transactional prices did not consistently reflect current market conditions, and quality holds or specification changes affected salability. When markets moved, net realizable value (NRV) analysis started from scratch. Plant and region teams kept local spreadsheets for special programs and customer commitments, which were rarely reflected in reserve calculations. Audit support required backfilling the links from reserve journals to batch?level evidence and rationale.

Why It Was Happening

Root causes were fragmented data and the absence of a governed reserve model. ERP inventory, BI sell?through, and external price indices used different item, region, and customer references. There was no canonical schema to join on?hand by lot to trailing sell?through, forecast coverage, and market price signals. Reserve policies existed on paper—aging thresholds, NRV constraints, and write?down mechanics—but they were applied ad hoc in spreadsheets with limited lineage. Exceptions lived in email, and approvals were not tied to the journal entries.

Ownership and timing were diffuse. Finance owned policy and journals, Supply Chain owned forecasts and disposition plans, and Sales owned customer commitments. Without a single pipeline to align inputs, apply transparent rules, and route exceptions, reserve setting relied on last?minute reconciliations rather than continuous monitoring.

The Solution

Intelligex implemented a reserve orchestration pipeline that unified inventory, sell?through, and market price data under a canonical model, then generated proposed reserves with finance?owned rules and approvals. ERP inventory by item, plant, and batch flowed into the model; BI provided trailing sales velocity and forecast coverage; external market price feeds informed NRV checks. The engine produced reserve proposals by item and batch with reason codes—slow?moving, forecast shortfall, NRV constraint, quality hold—and routed exceptions to Finance, Supply Chain, or Sales. Controllers approved entries through a maker?checker gate with linkbacks to evidence. Policy alignment followed inventory valuation principles in standards such as IAS 2 and comparable guidance under US GAAP (see ASC 330), while the company’s ERP and BI stack remained the systems of record.

• Integrations: Inventory, cost, and batch attributes from the ERP (for example, SAP); sell?through and forecast signals from the BI platform; market price benchmarks from trusted indices (for example, ICIS); journal creation and postings in the ERP; notifications to collaboration tools.
• Canonical reserve schema: Standard fields for item, plant, batch/lot, unit of measure, on?hand and in?transit, standard and recent prices, trailing sell?through windows, forecast coverage, quality status, market price reference, reason codes, and policy version.
• Reserve policy engine: Finance?owned rules for aging thresholds, slow?moving velocity triggers, NRV constraints, quality holds, and disposition impact; effective dating and rationale recorded for changes.
• NRV and pricing checks: Comparison of standard cost and recent prices to market benchmarks; configurable logic for transport and conversion assumptions where relevant; flags for negative margin risk.
• Exception routing: Items requiring context (customer commitments, planned reformulations, targeted promotions) routed to Sales or Supply Chain; Finance retained final approval with side?by?side evidence.
• Journal generation: Proposed reserve and release entries by entity and plant with batch?level detail; safeguards against double counting; linkbacks to underlying lots and policy rules.
• Dashboards: Exposure by business unit and plant, reserve drivers, forecast coverage gaps, and NRV exceptions; drill?downs to lots, customers, and price references.
• Audit and permissions: Role?based access; immutable logs of inputs, rule evaluations, approvals, and postings; exportable evidence packs with citations to BI views, batch records, and market price sources.

Implementation

• Discovery: Cataloged ERP item and batch structures, costing methods, and quality statuses; inventoried BI metrics for sell?through and forecast; identified external price sources by product family; reviewed reserve policy and audit feedback; sampled late adjustments and their root causes.
• Design: Defined the canonical reserve schema and identity crosswalks across ERP and BI; authored reserve rules for aging, velocity, NRV, and quality holds with effective dating; specified reason codes and routing; designed dashboards and evidence packs; mapped journal posting patterns and controls.
• Build: Implemented connectors for ERP inventory and cost data; integrated BI sell?through and forecast feeds; ingested market price benchmarks; built the policy engine and NRV checks; created the exception queue and maker?checker approvals; developed journal generation and lineage; assembled dashboards and notifications.
• Testing/QA: Ran in shadow mode: generated reserve proposals while the legacy spreadsheet process continued; reconciled outcomes to prior results; tuned identity mappings, velocity windows, and NRV assumptions; piloted exception routing with Finance, Supply Chain, and Sales.
• Rollout: Enabled proposals for selected business units and plants first; retained manual reserve setting as a controlled fallback; expanded coverage after stable cycles; made controller approval mandatory for exceptions once teams were trained.
• Training/hand?off: Delivered sessions for Finance, Supply Chain, and Sales on reading proposals, approving exceptions, and interpreting dashboards; updated SOPs for month?end reserve reviews and mid?month monitoring; transferred ownership of rules, reference mappings, and dashboards to Controllership under change control.
• Human?in?the?loop review: Established recurring reviews to assess rule effectiveness, NRV benchmarks, and exception trends; decisions recorded with rationale and effective dates.

Results

Reserves shifted from retrospective to proactive. Batch?level exposure was visible during the period, proposals included reason codes and evidence, and Supply Chain engaged sooner with disposition plans. Controllers focused on material exceptions rather than on assembling data, and reserve entries tied directly to lots, sell?through signals, and market references. Post?close cleanup declined because decisions were recorded under a consistent rule set with clear lineage.

Audit readiness strengthened. Evidence packs included policy versions, rule evaluations, and links to ERP lots, BI views, and price sources. Discussions centered on policy interpretation and operational plans rather than on data reconstruction. ERP and BI tooling stayed in place; the change was a governed pipeline that aligned inventory reality, market context, and reserve policy.

What Changed for the Team

• Before: Spreadsheets reconciled ERP aging and sell?through at close. After: A pipeline unified inventory, velocity, and price signals with live proposals.
• Before: NRV checks were one?off and manual. After: Market benchmarks informed standardized NRV constraints with citations.
• Before: Exceptions lived in email. After: Maker?checker routed items to Finance, Supply Chain, or Sales with reason codes and evidence.
• Before: Reserve journals lacked batch context. After: Entries linked to lots, plants, and policy rules with audit?ready lineage.
• Before: Supply Chain heard about exposure late. After: Dashboards showed at?risk items early with forecast and disposition context.
• Before: Post?close edits were common. After: Approvals and rules reduced late adjustments and clarified decisions.

Key Takeaways

• Unify signals; combine ERP inventory, sell?through, forecasts, and market prices under a canonical model.
• Encode policy; finance?owned rules for aging, velocity, NRV, and quality should be versioned and auditable.
• Route context; involve Supply Chain and Sales on forecast and disposition questions while Finance retains approval.
• Tie journals to lots; batch?level lineage strengthens reviews and audit support.
• Monitor continuously; dashboards surface exposure early so actions beat the close calendar.
• Integrate, don’t replace; layer a governed pipeline on top of the ERP, BI, and price sources you already use.

FAQ

What tools did this integrate with? The pipeline pulled inventory, cost, and batch data from the ERP (for example, SAP), used the BI platform for sell?through and forecast signals, ingested market benchmarks from trusted sources such as ICIS, and posted reserve entries in the ERP. Policy alignment followed concepts in IAS 2 and US GAAP summaries of ASC 330.

How did you handle quality control and governance? Reserve rules lived under Controllership change control with effective dating and rationale. Every proposal carried the policy version, reason codes, and links to lots, BI views, and price references. Maker?checker approvals were required for exceptions and high?impact items, and all evaluations, approvals, and postings were immutably logged.

How did you roll this out without disruption? The pipeline ran in shadow mode first, generating proposals while Finance continued its spreadsheet process. Results were reconciled to prior periods, and rules and mappings were tuned. Rollout began with selected business units and plants, with manual reserve setting retained as a fallback. Mandatory approvals for exceptions were enabled after training.

How were market prices and NRV evaluated? The engine compared standard cost and recent transactional prices to external benchmarks and configured assumptions for transport or conversion where needed. If NRV implied a lower recoverable value, the proposal reflected that constraint with citations to the benchmark and assumptions. Items near thresholds were routed for Finance review with side?by?side context.

What about lot?level aging and quality holds? Inventory was evaluated at batch/lot level with on?hand, in?transit, and quality status. Lots on hold or subject to specification changes were flagged with quality evidence. Reserve drivers and proposed actions (hold pending rework, disposition plan, or write?down) were visible in dashboards and routed to Supply Chain and Finance for decision.

How did Supply Chain and Sales participate? Exceptions involving forecast coverage, customer commitments, or planned reformulations were routed to the relevant owner. Their inputs—updated forecasts, disposition plans, or customer notes—were attached to the proposal. Finance retained approval and captured the final decision with reason codes.