Inventory carrying cost is the total expense of holding stock, expressed as a share of its value. The formula: carrying cost rate = total annual holding costs ÷ average inventory value × 100. Benchmarks cited by supply chain bodies such as ASCM and the Institute for Supply Management put typical rates at 20 to 30 percent per year, a range still widely quoted as of 2026.
That percentage sounds abstract until you apply it to your own balance sheet. A brand holding $100,000 of average inventory at a 25 percent carrying rate spends roughly $25,000 a year just keeping that stock on shelves, before a single unit sells. The rest of this guide breaks that number into its components, works the math for a marketplace seller, and shows where the standard 20 to 30 percent rule quietly understates reality for anyone paying Amazon's metered storage fees.
What goes into the carrying cost formula
The numerator, total holding costs, is where most calculations go wrong. Four categories belong in it. Capital cost is the return your cash could earn elsewhere, or the interest you pay if the inventory is financed. Storage cost covers warehouse rent, 3PL fees, or FBA storage charges. Service cost includes insurance, inventory software, cycle counts, and handling labor. Risk cost captures shrinkage, damage, markdowns, and obsolescence.
The denominator is average inventory value at cost, not retail. Use your cost of goods sold records, including freight and duties, and average the inventory balance across the year rather than taking a single point-in-time snapshot. A year-end count taken right after Q4 sell-through will flatter the ratio badly.
Carrying cost sits inside the broader discipline of tracking landed costs, margins, and inventory value accurately. Our complete guide to ecommerce accounting covers how those records fit together, and why per-SKU cost accuracy is the prerequisite for every formula in this article.
One naming note: inventory holding cost and inventory carrying cost are the same metric. Textbooks tend to say holding cost, operators tend to say carrying cost, and both describe the cost of carrying inventory as a ratio of holding expenses to average inventory value.
Worked example: $100,000 of average inventory
Here is the calculation itemized for an ecommerce brand holding $100,000 of average inventory at cost, occupying roughly 800 cubic feet in FBA. All figures are illustrative and use Amazon's published US standard-size storage rates as of the 2024–25 fee schedule.
| Component | Annual cost | Basis (illustrative) |
|---|---|---|
| Capital cost | $10,000 | 10% opportunity cost on cash tied up in stock |
| Storage (FBA, metered) | $11,376 | 800 cu ft at $0.78/cu ft for 9 months, $2.40/cu ft Oct to Dec |
| Service (insurance, software, handling) | $2,500 | Roughly 2.5% of inventory value |
| Risk (shrinkage, markdowns, obsolescence) | $5,000 | 5% of value; higher for seasonal or trend goods |
| Total carrying cost | $28,876 | 28.9% of average inventory value per year |
Two things stand out. First, the total lands near the top of the 20 to 30 percent consensus range, not the midpoint, and that is with a healthy catalog and no aged-inventory penalties. Second, capital cost is the largest line for most non-FBA brands, and it scales with how you fund inventory. A brand financing purchase orders at 12 to 15 percent APR should plug its real borrowing rate into the capital line, which alone can push the total past 30 percent. Our comparison of inventory financing and PO financing covers what those facilities actually cost.
The risk line deserves its own scrutiny. Markdowns and obsolescence are lumpy: zero for months, then a painful write-down when a SKU dies. If part of your stock has already stopped selling, the carrying math changes character entirely, which is why we treat dead stock as its own decision problem with its own exit ladder.
The FBA storage line is a fee schedule, not a percentage
Generic carrying cost articles treat storage as a fixed slice of warehouse rent. For an FBA seller, storage is a metered utility with a seasonal spike and a penalty cliff, and the standard 20 to 30 percent shorthand hides all three features.
Under Amazon's published US fee schedule (the 2024–25 rates; verify against the current schedule before you model), standard-size inventory costs $0.78 per cubic foot per month from January through September, then jumps to $2.40 per cubic foot from October through December. That is roughly a threefold increase exactly when most brands hold peak stock. A seller who carries the same 800 cubic feet all year pays about half of the entire annual storage bill in a single quarter.
Then comes the cliff. Once units have been in fulfillment centers for more than 180 days, an aged inventory surcharge applies on top of monthly storage, escalating through brackets until inventory past 365 days is charged $6.90 per cubic foot or $0.15 per unit, whichever is greater. For a slow SKU, the surcharge alone can exceed the product's remaining margin. This is why FBA sellers should compute carrying cost per SKU and per age bucket, not as one blended percentage. Our ecommerce fee calculators can help you model how storage and fulfillment fees stack against a unit's value.
The practical rule: for FBA inventory, replace the storage slice of the generic formula with your actual fee schedule math. The other three components still apply. Amazon charges you nothing for your cost of capital, but your balance sheet does.
3PL, FBA, or your own warehouse: three cost structures
The carrying cost formula is universal, but the shape of the storage component depends on where the inventory sits. Each structure fails in a different way.
| Structure | How storage is billed | Seasonal behavior | The hidden trap |
|---|---|---|---|
| FBA | Per cubic foot, metered monthly | Rates roughly triple in Q4 | Aged inventory surcharge past 180 days |
| 3PL | Per pallet or bin, plus handling fees | Mostly flat, contract-dependent | Handling charges on every touch add up |
| Own warehouse | Fixed rent, utilities, labor | Flat regardless of volume | Feels free at the margin, until you hit capacity |
A 3PL converts storage into a predictable per-pallet line, which makes carrying cost easy to compute but adds pick, pack, and receiving fees that belong in the service component. An owned or leased warehouse inverts the problem: because rent is fixed, the marginal cost of holding one more pallet looks like zero, which tempts operators into over-ordering. The capital and risk components never stop accruing, whichever building the boxes sit in.
Hybrid setups are increasingly common: bulk stock at a 3PL or in Amazon's upstream storage, with a lean forward position in FBA sized to 30 to 60 days of sales. That structure exists precisely to keep units out of the Q4 rate spike and away from the 180-day surcharge clock. One caution before running FBA too lean: Amazon's low-inventory-level fee, introduced in 2024, adds a per-unit charge on standard-size SKUs holding less than roughly 28 days of historical supply. Size the forward position above that floor, or the savings on storage leak back out as low-inventory fees.
How carrying cost drives EOQ and reorder decisions
Carrying cost is not a reporting metric. It is an input to every ordering decision, most directly through the economic order quantity formula. EOQ balances ordering costs against holding costs, and the holding cost per unit is your unit cost multiplied by your carrying rate.
The sensitivity is larger than most operators expect. Take a SKU with a $10 unit cost, 12,000 units of annual demand, and $50 of fixed cost per order. At a true carrying rate of 29 percent, holding cost is $2.90 per unit per year and EOQ works out to about 643 units per order. Assume a casual 15 percent instead, and EOQ inflates to roughly 894 units, a 39 percent larger order. Underestimating carrying cost systematically pushes you toward bigger, slower orders, which then raise your real carrying cost further. The error compounds in one direction.
The same input shapes safety stock. Every extra unit of buffer is insurance you pay for at your carrying rate, so the right buffer depends on what holding actually costs you versus what running out costs you. Improving inventory turnover and shortening days inventory outstanding are, in cash terms, direct reductions of the carrying cost base.
Carrying cost versus stockout cost
Carrying cost has an opposite number: the opportunity cost of running out. A stockout costs the lost sales, the ad spend that drove traffic to an empty listing, the marketplace ranking that decays during downtime, and the repeat customers who found an alternative. Those costs are spiky and mostly invisible in the ledger, while carrying costs are steady and visible, which biases cautious operators toward overstocking.
The honest comparison is made SKU by SKU. A hero product with strong margin and durable demand justifies carrying extra weeks of cover, because its stockout cost dwarfs its carrying cost. A marginal SKU with thin contribution and fading velocity deserves the opposite treatment, because every week of cover is expensive insurance on a shrinking asset. The reorder point formula is where the two costs meet in practice: it sets the trigger level that balances arriving too late against ordering too early.
Framed this way, carrying cost is really a working capital question. Cash sitting in stock is cash unavailable for the next purchase order or the next marketing push, and the longer the cash conversion cycle runs, the more of your growth is funded by inventory you have not sold yet.
Where Futureproof fits
Every input to the carrying cost formula lives in your books: landed cost per SKU, storage and fulfillment fees, insurance, and write-downs. Futureproof's AI bookkeeper, Vic, keeps those records accurate at the SKU level, so the fees Amazon deducts and the capital tied up in stock show up as real numbers instead of estimates. Margo, our FP&A agent, folds inventory into cash flow forecasts, so you can see what an extra container or a leaner reorder does to your cash position before you commit.
Shopify and Amazon integrations are now in beta. If you want carrying cost computed from your actual books rather than a rule of thumb, join the ecommerce waitlist.
Frequently asked questions
What is a good inventory carrying cost percentage?
Commonly cited benchmarks put typical carrying costs at 20 to 30 percent of average inventory value per year. Below 20 percent generally indicates fast turnover and cheap capital. Above 30 percent usually signals slow-moving stock, expensive financing, or aged FBA inventory accruing surcharges, and it is worth itemizing the components to find which one is inflated.
How do you calculate inventory carrying cost?
Add up a year of holding expenses across four categories: capital cost, storage, service costs like insurance and handling, and risk costs like shrinkage and obsolescence. Divide the total by your average inventory value at cost and multiply by 100. The result is your carrying cost rate, which you can then apply per SKU for ordering decisions.
Is inventory holding cost the same as carrying cost?
Yes. Holding cost and carrying cost describe the same metric: the total expense of keeping inventory, expressed as a percentage of its value. Academic and EOQ literature usually says holding cost, while operators and 3PLs usually say carrying cost. The formula and the components are identical.
Why is carrying cost higher for Amazon FBA sellers?
FBA storage is metered per cubic foot, and rates roughly triple from October through December, exactly when sellers hold the most stock. Inventory stored past 180 days also triggers an aged inventory surcharge that escalates with age. A seller with slow-turning stock can see the storage component alone exceed what generic 20 to 30 percent benchmarks assume for all four components combined.
How can ecommerce brands reduce carrying costs?
Order closer to demand using EOQ and reorder point logic, clear dead stock before surcharges accrue, negotiate 3PL storage for bulk units rather than warehousing everything in FBA, and shorten the cash conversion cycle so less capital sits in stock. Measuring the components per SKU is the first step, because a blended percentage hides which lever actually matters.
