Inventory carrying cost percentage
Carrying cost percentage i expresses annual holding cost as a fraction of unit cost: H = i * C. SMB default lands in the 20 to 30 percent range for stable goods[1]. Decomposes into capital (5 to 10), warehouse (2 to 5), services (1 to 2), risk (1 to 3), and obsolescence (variable). Mis-estimating i by 3 percentage points moves Q* by about 6 to 7 percent: not catastrophic, because the cost curve is flat near Q*.
Annual cost of holding inventory, expressed as a percent of unit cost.
Component-by-component build-up
- Capital cost: blended WACC or incremental borrowing rate. 2026 SMB range 6 to 10 percent.
- Warehouse: rent + utilities + labour, expressed as percent of average inventory value. SMB range 2 to 5 percent.
- Services: insurance, WMS license, inventory tax where applicable. 1 to 2 percent.
- Risk: shrinkage, damage, spoilage. 1 to 3 percent for staples, higher for perishables.
- Obsolescence: the variable one. 2 percent for industrial commodities, 20 percent+ for short-lifecycle electronics.
Sensitivity to i
Q* depends on 1/sqrt(i). A 5 percentage-point swing in i (from 25 percent to 30 percent or 25 percent to 20 percent) moves Q* by about 10 percent, and moves total annual variable cost by under 1 percent. This is the famous flat-curve robustness property from Silver, Pyke and Peterson[2].
When to use a higher i
Default i should rise above 30 percent when: (a) cost of capital is elevated (startup borrowing at 12 percent+), (b) obsolescence is built in (electronics, fashion), (c) shrinkage is documented above industry average (high-theft retail). Don’t round up i just to avoid carrying inventory; show the components.