LIFO ending inventory: 3 figures to check at each count

Key takeaways

• Check units on hand against your purchase and sales records.
• Match the unit cost of each layer still in stock to its invoice.
• Tie the ending inventory value out against goods available.

LIFO, which stands for last in, first out, assumes the newest purchases are sold first. Under periodic LIFO, what remains at the end of the period is therefore costed from the oldest layers. The idea is simple to state. The count is where it goes wrong. A miscounted unit or a layer booked at the wrong cost does not stay on one line. It moves dollars between cost of goods sold and ending inventory, and the error can sit in the books until someone checks the numbers against each other.

OpenStax, in its discussion of inventory valuation, cites an AICPA estimate that roughly 35–40% of all US companies use LIFO. It also notes that the method is more common in some industries, such as oil and gas ( OpenStax, section 10.1). An independent shop is probably nowhere near an oil refinery. The logic is the same for a hardware store that bought the same fastener at several different prices during the year. If you value stock under LIFO, the ending count deserves careful checking because quantity and cost errors both affect its value.

You need a count sheet, your purchase history, and a calculator. You do not need special software. Run the steps in the order given, because skipping ahead is how a count can look right on paper and still fail later.

Roles matter as much as tools. The owner or manager should own the purchase history and the ending figure. Count staff should own the physical count and nothing else. The person who checks the sheet should not be the one who counted, and ideally should not be the one who posts the sales. Splitting the work does not require extra staff in a small shop. It only requires each job to be written on the sheet with a name beside it.

1. Pull the purchase layers before anyone counts

Before the count starts, export or print the purchase history for the period. For each purchase, record the date, the number of units, and the unit cost from the vendor invoice. Add the beginning inventory from the last count as its own layer, with its own unit cost. These rows are your layers. Under periodic LIFO, the ending count will be built from them, starting with the oldest.

Be careful about the source of each unit cost. The invoice is the record that shows what you paid. A current price list, a vendor quote, or last month's average cost may help with planning, but none of them proves what a particular batch cost you. If an invoice is missing, find it before the count. A layer with a guessed cost is one of the easiest ways for a LIFO ending figure to drift without anyone noticing.

Next, work out the units available for the period. Add the beginning units to everything you purchased, then subtract the units sold according to your sales records. The result is the number of units you expect to find on hand. Write that number on the count sheet before anyone starts counting. Keep it off the copy the count staff use, because a known target can pull a count toward that figure.

The OpenStax periodic example works the same way. Over the whole period, 300 of the 585 units available were sold, and if the latest acquisitions are treated as sold first, the units that remain under LIFO are the ones purchased first ( OpenStax, section 10.2). The layer section below shows how the remaining units are costed in that same example.

One more decision belongs on the sheet. Write down whether your system costs each sale when it happens or only at the end of the period. The two approaches can draw from different layers for the same sale, so a sheet that does not say which one you use will be hard to check later. The oldest-layer assignment below applies to periodic LIFO.

Finally, look at the purchase history for anything odd: a delivery with no invoice, two delivery records with the same number, a return booked as a purchase, or a purchase dated after the count date. Any of these can throw off the layers, so resolve them now rather than during the count, when the pressure to finish is highest.

2. Figure one: units on hand

Count every item and write the physical number on the sheet. Count by location, not from memory, and work through the shelves, the back room, and any overflow space in a fixed order so nothing is counted twice or skipped. Recount any item where two people disagree. For high-value items, use a second counter who does not see the expected figure. Their count is a real check only if they could not have adjusted their number to match yours.

Then compare the physical count to the expected figure from step 1. There are three outcomes:

• The numbers match. Move on to figure two.
• The count is higher than expected. Look for deliveries that were never entered or stock sitting in the wrong bin. Sometimes a return went back on the shelf without being logged.
• The count is lower than expected. Look for unrecorded sales, damage, theft, transfers to another location, or a sale keyed against the wrong item. A lower count is often explained away fastest, so give it the same care as a higher one.

Do not adjust the count to match the expected figure. A difference is information. Record the physical number as units on hand, write the variance beside it, and give a reason, even if the reason is only "not yet found." Forcing the count to agree hides the problem and sends a wrong quantity into the cost calculation, where it changes the ending value. For a related counting process, see physical inventory reconciliation.

There is one more check before you move on. Units on hand cannot be more than units available. If they are, check for a missing delivery in the purchase history or double-counted units before using the layers you built in step 1. Fix the history or the count before you cost anything.

Handle returns and damaged stock with care during the count. A returned unit that is back on the shelf but was never reversed in the sales record will make the count look higher than the books. A damaged unit that was thrown out but never written off on paper will make the count look lower than the books. Walk the shelves with the written returns and write-offs for the period in hand, so you can tell physical stock from paperwork. Anything you cannot place should be listed as an open item rather than forced into a category. Review refund checks before restocking when investigating returned stock.

Keep the count sheet simple: each item, its location, its physical count, and a space for notes. Do not add value or cost columns during the physical count. Mixing money into the counting step is how a counter starts second-guessing a quantity because it looks too high or too low for its price.

3. Figure two: the unit cost of each layer

This is the figure that makes LIFO different from the other common inventory methods. Under periodic LIFO, the units still on hand are assumed to come from the oldest purchases. Assign the units on hand to layers starting from the earliest layer and moving forward until the count is used up. The newest layers should hold the least stock left, or none at all.

The OpenStax example shows how this looks in practice. Following that logic, ending inventory included 150 units purchased at $21 and 135 units purchased at $27 each, for a total LIFO periodic ending inventory value of $6,795 ( OpenStax, section 10.2). The specific prices matter less than the order. The earliest layers hold the units that remain, and the newer layers assigned to sales are used up or partly used.

Work through the layers in that order on the sheet. Start with the oldest layer and write how many of its units are still on hand. When a layer is fully used, move to the next one. Stop when the total assigned equals your units on hand from figure one. Then check the work with four tests:

• Each layer's unit cost matches its vendor invoice. Check the invoice itself, not a price list or last month's figure.
• No layer shows more remaining units than were bought in it. If one does, a unit was counted twice, or a sale was costed against the wrong layer.
• The layers used run in order from the oldest. If a newer layer still has stock while an older layer has no units assigned, the assignment was done backward, and the value will be wrong even if the total looks reasonable.
• The units across all layers add up to units on hand exactly. Close is not good enough. A difference of even one unit means a layer or the count needs another look.

Record each layer on its own line with three items: the units from that layer, the unit cost, and the extended value. Do not round the unit cost to make the arithmetic easier. If the invoice shows cents, keep them. Rounding at this stage moves the ending value by a small amount that carries through every layer, and small differences are hard to trace once they are absorbed into a total.

Some counts stop partway through a layer. Part of that layer is sold and part remains. That is normal, and it is where the most careful work goes. Write the split as two lines from the same invoice, with the units used and the units remaining, so the reviewer can see both halves and the cost that applies to each. Do not round the split to a neat number to make the sheet look tidier. A boundary that was rounded to fit is exactly the kind of entry that fails the tie-out later.

In a simple case, only one layer is partly used at the end of the period, usually at the boundary between what sold and what remains. If several layers look partly used, check the sales records for the period, because the sold units may have been drawn from the wrong layers when they were recorded.

4. Figure three: the ending inventory value and the tie-out

Ending inventory value is each layer's units multiplied by its unit cost, with the results added together. Check each multiplication by hand, even if you use a spreadsheet. A formula that points at the wrong cell produces a confident number that is simply wrong.

The third check ties the value to the rest of the period. Goods available for sale is the beginning inventory plus purchases at cost. Subtract the ending inventory from it and you get cost of goods sold. The OpenStax example does exactly this: subtracting the $6,795 ending inventory from the $16,155 total of goods available leaves $9,360 in cost of goods sold for the period ( OpenStax, section 10.2).

Use that relationship as your control. Goods available should equal ending inventory plus cost of goods sold. If your sales records report a cost of goods sold figure that differs from goods available less your ending value, something is wrong. It might be in the count value, the layers, or how the cost of sales was booked. Find which one before anything is posted. Both sides come from the same records, so the gap should point to a specific entry you can trace.

The ending value can look reasonable on its own. A total that seems plausible for the size of the shop does not prove the layers were assigned correctly, because a layer booked at the wrong cost can still produce a believable figure. A tie-out checks agreement between the totals; it does not replace checking each layer against its invoice.

If the tie-out fails, resist the urge to plug the difference into cost of goods sold or a miscellaneous adjustment account. A plug hides the error from the next person who reads the books. Go back through the sheet in order: units first, then layers, then arithmetic. The usual suspects are a delivery entered at the wrong quantity or cost, or a sale that drew from a layer it should not have.

Once the tie-out works, read the figures once more as a reviewer would. Does the ending value move in the direction you expected compared with the last count? Does cost of goods sold line up with the sales volume for the period? A figure can pass every arithmetic check and still be unusual in a way that deserves a question. Asking it before posting costs very little. Answering it after the books are closed takes far longer.

Illustrative example: follow the oldest layer

Suppose the oldest layer has 10 units at $5 each and the newer layer has 10 at $3 each. With 10 units remaining, periodic LIFO assigns $50 to ending inventory. Goods available cost $80, leaving $30 as cost of goods sold.

5. Tie the three figures together and record the count

Check the three figures in order. Units come first because they decide which layers exist. Layers come second because they decide the cost. Value comes last because it depends on both. If the value fails, go back to the layers. If the layers fail, go back to the units. Do not post an adjustment until the three agree or every remaining difference has a written reason.

Then record the count so someone else can follow it without asking you. For each count, the sheet should show:

• the count date, the counter, and the checker,
• expected units, counted units, and the variance with its reason,
• each layer with its units, unit cost, and extended value,
• the ending inventory value, goods available, and cost of goods sold, and
• a sign-off line for the owner or manager.

File the signed sheet with the purchase history from step 1. When a lender, an outside accountant, or a tax preparer asks how the ending figure was built, the sheet should answer the question without a phone call. The same file also works as a baseline for the next count. If the same item shows a variance every time, you have probably found a receiving or sales-recording problem, and fixing it at the source is cheaper than correcting it at every count.

Decide in advance how often you will run these checks, and write the schedule down. Some shops count at the end of each period. Others count more often and rely on sales records in between. Either way, a count run at a regular interval gives you a fixed point to compare against. A count that happens only when someone remembers it gives you a number with no history behind it.

Keep the process the same from one count to the next. A sheet that changes format every quarter makes the checks harder to compare, and comparison across counts is where a slow error shows up first. If a new person does the count, walk them through the sheet once beforehand, so they know what each column is for and why the expected figure stays hidden from the counting step.

Finally, LIFO is a method with its own rules, and it has to be applied consistently to mean anything from one period to the next. The OpenStax chapter gives the background on how the basic inventory methods differ and what cost flow assumption each one makes ( OpenStax, section 10.1). For the specific rules that apply to your business, talk with your accountant. They can tell you what your records need to show and whether your current setup supports the checks in this post.

Today, choose one item and gather its invoices, count it, assign its layers, and calculate its ending value.

Related guides: Units sold from a stock count: 3 figures to check · Zoho Inventory items: 5 checks before you save.

Common questions

Which costs belong on the sheet?

Use the invoice cost for each purchase layer, rather than a current quote or price list.

What if the totals disagree?

Check units, layer assignments, and arithmetic before posting an adjustment.