Should my small business use weighted-average or FIFO inventory costing?

What the law actually says about inventory cost

The Income Tax Act does not name a costing method. Subsection 10(1) says that property in the inventory of a business is valued at the end of the year at the cost at which the taxpayer acquired it or its fair market value at the end of the year, whichever is lower, or in a prescribed manner [1]. Regulation 1801 supplies the prescribed alternative: all the property in all the inventories of a business may instead be valued at fair market value [2].

So there are two year-end valuation choices, lower of cost and market, or fair market value for everything. Inside the first choice you still need a way to work out what "cost" means when you bought the same product five times at five different prices. That is where the costing method comes in.

The CRA's interpretation bulletin on inventory valuation, now archived but still the clearest statement of the administrative position, lists the methods most commonly used to determine cost: specific item, average cost, and first in, first out. It then states plainly that last in, first out and the base stock method are not accepted for income tax purposes [3]. International accounting standards land in the same place: specific identification for items that are not interchangeable, and FIFO or weighted average for items that are [4].

How weighted average works

Weighted average keeps one unit cost per item. Every time new stock arrives, you add the cost of the new units to the cost of what is already on the shelf and divide by the new total quantity. Sales then come off at that single blended unit cost.

Say you hold ten units that cost four dollars each, and you receive ten more at six dollars each. You now have twenty units at a weighted-average cost of five dollars. Sell eight and cost of sales is forty dollars; the twelve left on the shelf are carried at sixty. Nobody has to remember which box came from which shipment.

The method suits interchangeable goods: fasteners, consumables, parts, packaged products where one unit is indistinguishable from the next. It also smooths price swings, so one expensive emergency order does not distort the margin on the next week of sales.

How FIFO works

First in, first out assumes the oldest units are sold first. Each receipt becomes its own cost layer. Sales consume the oldest layer until it is exhausted and then move to the next. Using the same example, selling eight units would cost them all at four dollars, leaving two units at four and ten at six on the shelf.

FIFO often matches physical reality for perishable or dated goods, where you genuinely do rotate the oldest stock out first. Its drawback is bookkeeping: every item carries a queue of layers, and every sale, return and adjustment has to be applied to the right one. With a few hundred items and frequent receipts, that queue becomes the thing people get wrong.

What the choice does to your profit

When purchase prices rise, FIFO leaves the newest and dearest units in closing inventory, so cost of sales is lower and profit is higher. Weighted average pulls some of the newer cost into cost of sales sooner, so it tends to report slightly less profit in an inflationary year. When prices fall, the effect reverses.

It is worth being clear about the size of this. Over the life of a business, every unit you buy is eventually expensed; the method only moves profit between years. For most small resellers the annual difference is modest compared with the effect of a sloppy count or unrecorded freight. The method is rarely what makes an inventory number wrong.

Consistency is the actual rule

Subsection 10(2) makes your opening inventory for the year equal to the value you used as closing inventory the year before [1]. Subsection 10(2.1) then requires that a valuation method used at the end of one year be used at the end of the following year, unless the taxpayer changes it with the concurrence of the Minister [1].

The CRA's bulletin adds the practical expectations. The method used for tax should normally be the same as the one used for the financial statements. Where accounting principles permit more than one method, the tax method should be the one that gives the truer picture of income, and it must be followed consistently from year to year. A change will only be accepted if the new method is more realistic and gives a truer picture; the request is made in writing, normally before filing the return for the year of the change [3].

That is why the choice deserves a few minutes of thought at the start. Picking weighted average in software while your accountant prepares statements on FIFO, or quietly flipping methods because one year looks better, creates exactly the inconsistency the Act is written to prevent.

Lower of cost and market, item by item

Whichever method you choose, it only produces the cost side of the comparison. At year end the bulletin describes comparing cost and fair market value separately for each item, or each usual class of items where specific items are not readily distinguishable, and carrying the lower figure for each [3]. IAS 2 expresses the same idea as lower of cost and net realisable value [4].

This is where real money moves. Damaged stock, discontinued lines and anything that has sat unsold for two seasons may be worth far less than you paid. Writing those down at year end is legitimate and expected. What the bulletin does not accept is writing inventory down below market simply to preserve a profit margin in the year the goods are finally sold [3].

Where the number lands on the T2

For a corporation, the costing method feeds the General Index of Financial Information. Opening inventory goes on item 8300, purchases on 8320, closing inventory on 8500, and cost of sales on 8518; the year-end inventory balance also appears on the balance sheet as item 1120 [5]. Because this year's opening figure must equal last year's closing figure, an inconsistent method shows up as a break between two returns.

What to cost into each unit, whichever method you use

Both methods are only as good as the cost you feed them. The bulletin describes cost for merchandise purchased for resale as laid-down cost, which includes invoice cost, customs and excise duties, transportation and other acquisition costs, and storage where significant [3]. IAS 2 similarly includes costs of purchase and other costs incurred to bring inventory to its present location and condition [4].

In practice this means inbound freight, brokerage and duty belong in unit cost, not in a general shipping expense account. A weighted average built on invoice prices alone understates every unit and overstates every margin, no matter how carefully the arithmetic is done.

A sensible default for a small business

If your goods are interchangeable, you do not track expiry dates, and nobody on the team wants to manage cost layers, weighted average is the simpler method to run correctly. If your stock is perishable or dated and you physically rotate it, FIFO may reflect reality better and is equally acceptable. Talk to whoever prepares your financial statements before the first year end, choose once, write it down, and keep it.

MapleInventory uses weighted-average landed cost. Each receipt folds its product cost, freight, duty and local stocking costs into a new running unit cost, with recoverable GST/HST kept out of cost by default. The resale inventory report then shows on-hand quantity, the unit cost broken into its parts, and the extended value by category, location or vendor, which is the cost side of the year-end comparison ready to review against market.