The square-root law of inventory pooling is the rule of thumb behind every warehouse-consolidation business case: combine N stocking locations into one and your safety stock falls by roughly the square root of N. Consolidate four warehouses, halve the safety stock. It is elegant, widely quoted, and built on assumptions that real networks routinely violate, so it is wrong more often than its confident use suggests.
What the law actually says
Pooling demand reduces relative variability: combine independent demand streams and the standard deviation of the total grows slower than the total mean. Under its assumptions, pooled safety stock scales with the square root of the number of locations, so consolidating from N sites to one cuts the safety-stock requirement by a factor of about sqrt(N).
A four-warehouse example with correlated demand
Assume a fictional network holds 100 units of safety stock at each of four equal warehouses, or 400 units in total. With independent demand, the same service target and unchanged lead time, pooling into one warehouse gives 400 / sqrt(4) = 200 units.
For equal demand variances and the same correlation r between every pair of locations, the pooled-to-original safety-stock ratio is sqrt((1 + (N - 1) * r) / N). At r = 0.5 and N = 4, it is sqrt(2.5 / 4), about 0.7906. The resulting safety stock is about 316.23 units, rounded up to 317, rather than 200. At r = 1 the ratio is one, so pooling offers no safety-stock reduction under these assumptions.
This isolates the effect of correlation. It does not include a change in transport time, service promise or order quantity. Use the safety-stock and reorder-point example to keep those decisions separate. MIT’s Supply Chain Fundamentals reference explains risk pooling and the square-root relationship in its multiple-locations lesson.
The assumptions it quietly requires
The clean sqrt(N) result holds only when:
- Demand across locations is independent. No correlation.
- Locations are identical. Same mean and variance of demand.
- Demand is normally distributed. So the variance arithmetic is clean.
- Service level and lead time are unchanged by the consolidation.
Real networks break all four.
Where it misleads
- Correlated demand. If locations rise and fall together (a national promotion, a seasonal swing, a region-wide weather event), pooling removes far less variability than sqrt(N) promises, because the demands were not independent to begin with. High positive correlation can shrink the benefit toward zero.
- Unequal locations. A few big sites and many tiny ones do not pool like N equal ones; the law overstates the gain.
- Non-normal, lumpy demand. For intermittent or fat-tailed demand the normal-based arithmetic does not hold, the same trap as non-normal safety stock.
- Longer lead times after consolidation. A central warehouse often sits farther from customers; the longer (and more variable) lead time can claw back the pooling saving.
What to do instead
Do not bank a consolidation case on sqrt(N). Estimate the real pooled variability from your actual, correlated, non-identical demand, by computing the variance of the summed demand including covariance, or by simulating the consolidated network in a digital twin, and size safety stock from that. This is the same reason multi-echelon optimization beats single-echelon rules of thumb: it models the network you actually have, not an idealized one.
The takeaway
The square-root law is a useful intuition, pooling reduces relative variability, but a dangerous number to plan with, because its independence-and-identical-demand assumptions rarely hold. Correlation, unequal sites, non-normal demand, and longer central lead times all erode the benefit. Compute the pooled variance from your real demand before you promise the savings.
Working through this in your warehouse?
The team that wrote this also implements inventory architecture, audits operations, and advises on transformation engagements. AvanSaber’s inventory practice runs case-by-case engagements for mid-market and enterprise inventory teams.
Book a discovery conversation or describe your situation at [email protected].
Learn more about our engagement shapes: Inventory Implementation, Audits, Advisory.