Inventory valuation: FIFO and weighted average cost

Second week of January, and the year-end report lands on the table at an electrical supplies wholesaler. The count is done, the quantities agree, no box is missing. Yet three different costs are circulating for the same reel of cable: the last purchase price the salesperson used when quoting, the average cost sitting on the item card in the bookkeeping software, and a third figure on the accountant's inventory list. None of them is wrong. Each answers a different question. The trouble is that nobody said out loud, at any point during the year, which question they were asking. The bill arrives on a large tender won in November, quoted at a number below what it costs to buy the same goods again today.

Inventory valuation is treated as a technical footnote and quietly sets pricing, profit and tax at the same time. This article covers why you have to pick a method at all, how FIFO and weighted average actually work, exactly where the two diverge, what each shows when purchase prices are rising, why your price list must never be fed from either, which costs belong inside inventory value, at what level valuation should be held, how count differences and write-downs break the arithmetic, the assumptions your software makes without telling you, and where to start.

1Purchase2Added costs3Stock value4Sale5Cost of sales6Gross margin

Why you have to choose a method at all

The boxes in the warehouse are identical to each other. Their costs are not. If you bought the same item three times in a year at three different prices, there is no physical fact that tells you which cost leaves when one unit is sold. Whichever box you pull off the shelf, the price is not printed on it. What fills that gap is a rule, and the rule is what we call an inventory valuation method.

What the rule determines is not one number but a number that splits in two. The total you paid gets divided between cost of goods sold and the value of what remains in the warehouse. The more you allocate to one side, the less is left for the other, which means the choice of method touches the income statement and the balance sheet in the same move. Teams that never make this connection can run a disciplined stock count and still publish misleading reports. The quantity side of the setup is covered separately in stock tracking and inventory management.

FIFO: the next layer leaves

FIFO holds every purchase as its own layer. When goods are received, the system records a quantity and a unit cost for that quantity; when a sale happens, cost is drawn from the oldest layer, and once that layer is exhausted the next one takes over. A single sale can partly consume more than one layer, and that is perfectly normal. Whatever remains in stock always consists of the most recent purchases.

The most common misreading is treating FIFO as a warehouse instruction. FIFO is a cost assumption, not a physical rule. Which carton the picker grabs has no accounting consequence whatsoever; even if the newest box goes out the door, the cost still leaves from the oldest layer. Of course, goods with expiry dates should physically move oldest-first too, but that is stock discipline, not a requirement of the valuation method. Teams that conflate the two assume their accounts break the moment their shelving does.

Weighted average: recalculated on every receipt

Weighted average has no layers, only a single unit cost. When a new receipt arrives, the total value of existing stock and the total value of the new purchase are added together, divided by the combined quantity, and the item card is updated. Sales are then costed from that one number. The record-keeping load is noticeably lighter, and in exchange you permanently lose the ability to trace which batch was sold at which cost.

A moving average is not a periodic weighted average

Two distinct practices hide behind similar names, and they get mixed up constantly. A moving average recalculates cost at every goods receipt, so a sale made mid-month is costed at whatever the average happened to be at that moment. A periodic weighted average computes a single average across the entire period's purchases and applies it to every sale in that period. Same data, different results. Your bookkeeping software most likely runs a moving average, while your accountant may be closing the period on a periodic calculation. That mismatch is very often the hidden source of the three numbers in the opening scene.

Where exactly do the two methods diverge?

When purchase prices hold steady, the difference between the methods is zero. The gap opens the moment prices move, and it widens with the speed of the movement. The table below shows where the choice genuinely changes an outcome.

Point of comparison	FIFO	Weighted average
Cost of sales while prices rise	Drawn from older, cheaper layers	Sits between old and new purchases
Value of remaining stock on the balance sheet	Stays close to current purchase prices	Lags behind current prices
Reported gross profit	Higher and more volatile	Smoother and steadier
Speed at which a price shock reaches reports	Delayed until old layers run out	Partly reflected on the first receipt
Record-keeping load	Requires layer tracking	One cost field is enough
Items with batch or serial tracking	Fits naturally	Hides the difference between batches

The most practical line in that table is the second from the bottom. In a wholesaler carrying hundreds of items, layer tracking is real work unless the software does it automatically, and hand-kept layer lists degrade within weeks. Conversely, for serialized or batch-coded goods, an average cost makes the genuine difference between two batches invisible and distorts warranty cost calculations along with it. The choice therefore follows the shape of your product range, not personal preference.

Which one tells the truth when prices rise?

In a period of rising purchase prices, FIFO writes the older and cheaper layer into cost of sales. The result is a higher gross profit on paper, while the stock left on the balance sheet looks closer to real replacement cost. Weighted average softens both: profit shines less, and stock value sits a little behind the current market. The trap here is that part of that flattering profit is not spendable, because the money is already needed to buy the same goods again.

When prices fall the relationship reverses and FIFO becomes the method reporting lower profit. No method is inherently conservative or optimistic; the direction of prices sets the direction of the distortion. How to read financial statements under high inflation is covered in inflation accounting. Which method you may use in Türkiye for your own filings, whether it can be changed, and how it flows into your returns depends on your specific situation, so settle that part with your accountant rather than with an article.

Your valuation method explains the past while your price list has to pay for the future, and feeding both from one number makes every price increase one step late.

Never feed your price list from your valuation method

The standard reflex is to change the valuation method when the sales team keeps misreading profitability. That is reaching for the wrong tool. FIFO and weighted average both look backward, and neither tells you what buying the same goods again would cost today. Yet that is precisely the number a pricing decision requires.

The fix is simple and rarely implemented: keep two separate fields on the item card. One is the valuation cost accounting uses; the other is a current replacement cost used for pricing, fed by the latest purchase or the supplier's live quote. The quote screen reads the second field, the financial statements read the first. Companies that separate them stop being late on price increases, because the trigger becomes visible at the moment of purchase. The wider pricing decision sits in pricing strategies, and how to measure the outcome is in gross profit margin.

Which costs belong inside inventory value?

Beneath the method debate sits a layer where mistakes are far more frequent: the cost itself. The stock value of an item is not simply the invoice amount. Some of the spending that makes goods sellable attaches to that value, and some of it does not. The distinction below is enough of a framework for most small and mid-sized companies.

• Purchase price and discounts: The invoice amount is the core; a volume rebate arriving later reduces cost retroactively, and most teams never make that adjustment.
• Freight inbound: Transport that brings goods to your warehouse belongs in stock value; transport from the warehouse to the customer does not, that is a selling expense.
• Duties and non-recoverable taxes: Import taxes you cannot offset stick to the cost of the goods; recoverable ones do not.
• Insurance and handling: Transit insurance and unloading at the destination are among the costs that make goods sellable.
• Currency differences: On foreign-currency purchases, up to which moment the difference belongs in stock is the most contested line of all, so check your own case with your accountant.
• What must stay out: Storage, general administration, financing interest and outbound freight inflate stock value; they are period expenses.
• Shrinkage and returns: Damage found at goods-in reduces quantity but spreads its cost over the surviving units, while the cost at which a customer return re-enters stock must be defined as a rule in advance.

On foreign-currency purchases the sequence of these items deserves particular care, since the date on which the rate is fixed can produce two different stock costs from the same invoice. The recording side of that is covered in foreign-currency invoices and exchange differences, and how cost is allocated down to the item level in the fundamentals of cost accounting.

At what level should valuation be held?

Even a well-chosen method turns meaningless when applied at the wrong level. In a company holding item cards at product level while purchasing at variant level, the average cost is an average of things that are not comparable. If two sizes of the same model arrive at different prices and the system pools them onto one card, the resulting number represents no real transaction at all.

The second question is the warehouse level. When the same item sits in two locations, is it tracked at one cost or does each location carry its own? Whether a transfer between locations moves cost with it gets decided here too; when it does not, one warehouse appears to have received goods for free. For items requiring serial or batch tracking, the soundest approach is to attach cost to the batch and keep it out of the average entirely. How to structure the item card is detailed in product variant management.

Count differences, shrinkage and write-downs

Every valuation method rests on one assumption: the quantities are right. When they are not, the method debate is beside the point, because a wrong quantity generates a wrong cost automatically. Correcting a count difference in units while forgetting the cost side is a common failure too; goods found missing have to be absorbed as an expense somewhere.

The bigger error lives elsewhere: slow-moving stock whose sellability has faded but which is still carried at day-one cost. An item that has not moved in two years sitting at full value flatters the balance sheet and delays the decision. What deserves discussion at that point is not FIFO versus average but when the value gets written down. To see which goods have aged and by how much, look at stock turnover and aging analysis, and to build the counting discipline itself, at stock counts and cycle counting.

Which method is your software actually applying?

This section carries the details that cost more money than the method debate itself. No valuation decision is solid until you have put four questions to your system. First: when a backdated purchase invoice is entered, does the software recalculate the cost of the sales that followed it? Under a moving average the answer is usually yes, which means last month's reported gross margin quietly changes underneath you.

Second: is negative stock allowed? When goods physically arrive but are sold before the purchase invoice is entered, the system has to invent a cost, and entering the invoice afterwards corrupts the average. Third: at what cost does a customer return re-enter stock, the cost it carried when sold or today's average? Fourth: does a transfer between warehouses carry cost with it? Get those four answers in writing. Most teams learn them at year-end, from a report that will not balance.

Where to start and what to watch

Five steps are enough to begin. Choose a method and write down in one sentence which one you chose; while it is unclear who sees which cost on which screen, the argument never ends. Turn the list of what enters cost and what stays out into a short note and hand it to everyone who enters purchase invoices. Close the negative-stock hole. Open a separate current-cost field for pricing. Finally, review slow-moving items once a quarter.

The indicators can stay just as plain. Gross margin swinging unexpectedly between months is usually a data problem rather than a method problem. The count of backdated entries shows how often the past is being rewritten. The count of negative-stock events is the earliest warning that costs are corrupting. Count variance and stock turnover complete the picture. How these come together at the close of the year is collected in year-end close and inventory.

Most valuation arguments start because purchases, stock and sales are recorded in three separate places. In Rocketly, purchase invoices, stock movements, sales records and reporting run on the same data, with the cost fields gathered in one place; open a free account and build your own cost flow.