EOQ or Periodic Review? Choosing an Inventory Policy That Survives Real Demand

Every inventory course starts with the economic order quantity, and every practitioner eventually discovers that it assumes away the two things that actually hurt: demand variability and supply variability. EOQ is still worth knowing, because it tells you the shape of the cost curve. It just is not a policy on its own.

What EOQ actually tells you

EOQ = square root of (2 × D × S / H), where D is annual demand in units, S is the fixed cost per order, and H is the holding cost per unit per year. The important property is not the number — it is that the total cost curve around the number is remarkably flat. Order 30 per cent above or below EOQ and your total ordering-plus-holding cost rises by roughly 3 to 4 per cent.

That flatness is a gift. It means you can round the EOQ up to a pallet, a container, or a minimum order quantity without meaningful penalty, and it means arguing about the exact holding cost percentage is usually wasted effort.

Continuous review (s, Q)

You watch stock continuously; when it drops to the reorder point s, you order a fixed quantity Q. Reorder point is expected demand over the lead time plus safety stock:

s = d_bar * L + z * sigma_L

where
  d_bar   = average demand per period
  L       = lead time in periods
  z       = service factor for your target cycle service level
  sigma_L = standard deviation of demand over the lead time

When both demand and lead time vary — which is the normal case for imported raw material — sigma_L is not just the demand sigma scaled by root L. Use:

sigma_L = sqrt( L * sigma_d^2  +  d_bar^2 * sigma_L_time^2 )

Skipping the second term is the single most common reason a factory that "carries three weeks of safety stock" still stocks out. Lead time variance on a 45-day ocean lane with a standard deviation of 8 days dominates the calculation entirely.

Periodic review (R, S)

You check stock every R periods and order up to a target level S. The exposure window is now R + L, not L, so:

S = d_bar * (R + L) + z * sigma_(R+L)

Periodic review carries more safety stock for the same service level. In exchange, it lets you consolidate orders across items from the same supplier, which is often worth far more than the extra stock — particularly when freight is the dominant cost.

Choosing between them

Service level is not fill rate

The z value in every formula above delivers a cycle service level: the probability of not stocking out during a replenishment cycle. Management usually means fill rate: the fraction of demand met from stock. They are different numbers, and fill rate is almost always higher than cycle service level for the same safety stock, because a stockout late in a cycle affects few units.

If you promise 95 per cent and deliver against the wrong definition, you will over-invest in stock by 15 to 25 per cent. Agree the definition in writing before you size anything.

A sanity check that takes ten minutes

Take your top 20 SKUs by value. For each, compute demand-over-lead-time and its standard deviation from the last 12 months of actual issues — not forecasts. Compare the implied safety stock at z = 1.65 with what you actually hold. In most factories the answer is that fast movers are under-protected and slow movers are carrying six months of cover. Fixing that mix usually releases cash without changing service at all.