You're staring at a spreadsheet. Somewhere in that grid is the story of Dunbar Incorporated's inventory — what they bought, what they sold, what's still sitting on the shelf, and what it's all worth. The problem? Because of that, the story doesn't tell itself. Consider this: column after column of item numbers, unit costs, quantities on hand, dates received, dates sold. You have to extract it.
Inventory records for Dunbar Incorporated revealed the following: a mismatch between the physical count and the perpetual ledger. A batch of raw materials received in December but recorded in January. A gap between FIFO and LIFO valuations that changes net income by six figures. And a write-down that nobody documented until the auditors asked.
You'll probably want to bookmark this section.
Sound familiar? If you've ever closed a month-end with inventory on the books, it probably does.
What Inventory Records Actually Tell You
At its core, an inventory record is a promise. It says: we have this much stuff, it cost this much, and here's the proof.* But the promise gets tested every time goods move — received, transferred, consumed, sold, returned, damaged, stolen, or simply miscounted.
You'll probably want to bookmark this section.
Dunbar's records, like most companies', live in three places at once:
- The perpetual system (ERP, WMS, or that one spreadsheet someone built in 2014)
- The general ledger (where inventory hits the balance sheet)
- The physical reality (what's actually on the shelves, in the yard, on the truck, at the vendor's dock)
When all three agree, you sleep well. When they don't, you get the Dunbar situation: a $427,000 variance that took three people two weeks to untangle.
The Data Points That Matter
Every inventory record should capture, at minimum:
- Item identifier (SKU, part number, serial/lot for traceability)
- Quantity on hand (by location, by status — available, allocated, in transit, quarantine)
- Unit cost (which cost? standard, moving average, FIFO layer, LIFO layer, specific ID)
- Date acquired (for aging, obsolescence, LIFO conformity)
- Valuation method tag (so you know which layer you're pulling from)
- Last count date and count variance (cycle count history)
Missing any of these? You're not alone. Most mid-sized companies miss at least two.
Why the Dunbar Case Matters Beyond One Company
You might wonder: why does a single company's inventory mess warrant an article? Because the patterns are universal. That said, the specific numbers change — $427K here, $1. 2M there — but the root causes repeat across industries, ERPs, and decades Small thing, real impact..
The Income Statement Impact
Inventory isn't just a balance sheet line. Understood by the board? In practice, it becomes* cost of goods sold. Here's the thing — disclosed? Now, dunbar's switch from LIFO to FIFO for a single product line (done quietly in Q3) inflated gross profit by $183,000. That said, yes. That's why legal? Barely. Plus, every valuation choice — FIFO, LIFO, weighted average, specific identification — flows straight to gross margin. Not until the audit.
Not obvious, but once you see it — you'll see it everywhere.
The Tax Angle
LIFO saves tax in inflationary periods. Practically speaking, the deferred tax liability was understated. That said, dunbar's records showed a "LIFO reserve" footnote that hadn't been updated in two years. But it requires conformity — if you use LIFO for tax, you generally must use it for financial reporting. The IRS doesn't like that.
The Operational Signal
Inventory records are also a diagnostic tool. Because of that, frequent stockouts on A-items? In real terms, dunbar had 47 SKUs with negative on-hand quantities at year-end. Still, your reorder points are wrong. On top of that, your purchasing is off. And slow-moving SKUs? Your receiving or shipping process is broken. Because of that, forty-seven. Negative quantities in the system? That's not a rounding error — that's a process failure Still holds up..
How to Read Inventory Records Like a Forensic Accountant
You don't need a CPA license to spot trouble. You need a checklist and the willingness to dig The details matter here..
Step 1: Reconcile the Three Worlds
Start with the perpetual-to-GL reconciliation. Run the inventory subledger detail. Sum it. Compare to the GL control account. Any difference? Think about it: that's your first red flag. Dunbar's was $38,000 — small enough to ignore, large enough to matter Not complicated — just consistent. No workaround needed..
Then do the physical-to-perpetual reconciliation. And this is where cycle counts live. The year-end physical came in $427K below perpetual. Here's the thing — if you only count once a year, you're guessing for 364 days. The variance account absorbed it. Dunbar counted annually. Nobody investigated why until March Less friction, more output..
Step 2: Test the Valuation Layers
If you use FIFO or LIFO, your system maintains cost layers. So do the consumed layers match the method? Consider this: pull a high-volume SKU. Also, the cost flow didn't match the physical flow. Each receipt creates a layer. Dunbar's system was set to FIFO but the warehouse picked oldest-expiration-date first (FEFO). Each issue consumes layers in sequence. Trace five recent issues. The valuation was technically wrong — and materially so for perishables Worth knowing..
Step 3: Check the Cutoff
This is where most errors hide. Cutoff means: did the transaction record in the right period?
- Goods received Dec 31, invoice dated Jan 3, recorded Jan 5 → cutoff failure
- Goods shipped Dec 30, FOB shipping point, revenue recorded Jan 2 → cutoff failure
- Goods on consignment at customer site, included in your count → not your inventory
Dunbar had all three. The consignment goods alone were $91K.
Step 4: Analyze the Reserves
Obsolescence reserve. Lower of cost or market (LCM) reserve. Shrinkage reserve. These are estimates — which means they're judgments. And judgments can be managed.
- When was the reserve last recalculated?
- What assumptions changed?
- Does the reserve policy match actual write-off history?
Dunbar's obsolescence reserve used a flat 3% of ending inventory. In practice, actual write-offs over five years averaged 7%. So the reserve was understated by roughly $210K. Management called it "conservative." The auditors called it a material misstatement.
Common Mistakes That Keep Showing Up
1. Treating the ERP as Truth
The system shows 1,247 units. Now, *Wrong. The system must be right — it's the system. ** The system is only as good as the transactions fed into it. The shelf shows 1,182. Missed receipts, unrecorded scrap, mis-picks, returns not processed — every gap between process and system widens the variance.
2. Ignoring In-Transit and Consignment
Goods on a truck. Day to day, goods at a vendor for processing. Goods at a customer on consignment.
it? Because of that, if the title, risk, or reward hasn't transferred, **it's not yours. ** Dunbar had $91K in consignment goods sitting in their warehouse because the warehouse team didn't flag the consignment labels. The system booked them as owned. The balance sheet inflated. The auditors found it only because they pulled the consignment agreements from the file — a document that had been filed and never read by anyone on the audit team until year three Which is the point..
2. Skipping the Count Procedure Review
You can't audit what you don't understand. - **What happens with exceptions?In practice, no tag, no count. - **Who supervises?Consider this: ** Trained staff? Temporary hires? In real terms, before the count, ask:
- **Who counts? - What's the tagging process? Every count line needs a tag number, a counter's initials, and a supervisor's verification. Third-party counters? ** Is the supervisor independent of the warehouse team? ** Are variances flagged on the spot or reconciled the next morning?
Dunbar used a printout of the item list and clipboards. But counters went aisle by aisle. Consider this: tags were collected at the end of the day. Now, by then, the counters had gone home and the supervisor had left for a meeting. The tags were reconciled by someone who hadn't been present during the count. That's not a procedure — that's a suggestion.
3. Overlooking Work-in-Process (WIP)
If Dunbar had been a manufacturer, the WIP would have been the real problem. Still, raw materials enter a production line. Day to day, labor and overhead get applied over days or weeks. At what percentage is the unit "complete"? That percentage determines the inventory value.
- Is the completion percentage based on actual time studies or management estimates?
- Are overhead allocation rates current, or frozen from a prior year?
- Are partially completed units counted at all, or excluded from the physical count entirely?
Dunbar didn't manufacture, but they did assemble kits for custom orders. Because of that, the kitting area was treated as "finished goods" because it was next to the shipping dock. Because of that, in reality, some kits were 60% assembled with final quality checks pending. Those units should have been in WIP. The misclassification inflated finished goods and understated work-in-process.
4. Failing to Test the Count Itself
The inventory count is the foundation. If the foundation is off, everything built on it is wrong. Auditors test counts by:
- Observing the count process (not just reviewing the count sheet afterward)
- Performing their own test counts on a statistically valid sample
- Tracing counted items back to the system and forward to the count sheet
- Investigating large or unusual variances immediately, not after the fact
At Dunbar, the audit team reviewed the count sheets the Monday after the count. They didn't observe the count. They didn't test-count a single aisle. They compared the count sheet to the system and noted a 2.Still, 1% variance — within the 3% threshold they'd set. They signed off.
The 2.And 1% variance was real. But it was the wrong* 2.1%. The test would have caught it if it had been done during the count, when the counters could explain why Bin A-14 had 47 units instead of the 62 the system showed — because 15 units were on a pallet in the wrong aisle, tagged to the wrong bin number. That single correction would have reduced the variance to 1.6%, which is still within threshold but materially different in dollar terms for a company of Dunbar's size.
Honestly, this part trips people up more than it should.
The Real Cost of Getting It Wrong
Inventory is often the largest current asset on a company's balance sheet. For Dunbar, it represented 34% of total assets. A misstatement in inventory doesn't just affect one line item — it cascades.
- Overstated inventory → overstated assets → overstated equity → inflated financial ratios
- Understated cost of goods sold → overstated net income → misstated tax provision
- Misclassified inventory (
misclassified inventory $\rightarrow$ distorted gross margins $\rightarrow$ misleading management reports used for strategic decisions
When the inventory numbers are wrong, the entire narrative of the company's health becomes a fiction. And a company might appear to be growing and profitable because it is "parking" its losses in unsold stock, or it might appear to be in a liquidity crisis simply because it failed to account for goods currently sitting on a loading dock. For stakeholders—lenders, investors, and tax authorities—this isn't just a bookkeeping error; it is a fundamental failure of transparency That's the whole idea..
The Ripple Effect on Decision-Making
Beyond the balance sheet, flawed inventory data poisons the operational side of the business. If the system reports higher stock levels than what is physically available, the procurement team may delay essential orders, leading to stockouts and lost sales. Conversely, if inventory is understated, the company may over-order, tying up precious cash in excess raw materials that sit in a warehouse, gathering dust and risking obsolescence.
In the case of Dunbar, the 2.Think about it: 1% variance was treated as "noise. " But for a company operating on thin margins, that "noise" is actually the signal that the business is losing control of its most valuable resource.
Conclusion
Inventory management is not a back-office clerical task; it is a high-stakes exercise in precision. As we have seen, the errors rarely stem from simple math mistakes. Instead, they arise from the "gray areas"—the subjective estimates of work-in-process, the misclassification of goods based on location rather than status, and the failure to rigorously validate physical counts.
For auditors and management alike, the lesson is clear: trust, but verify. A threshold for materiality is a useful tool, but it should never be used as a license for complacency. To ensure financial integrity, a company must move beyond "paper audits" and engage with the physical reality of the warehouse floor. Only then can the numbers on the balance sheet truly reflect the reality of the business Most people skip this — try not to..
