Safety Stock Calculator

The Safety Stock Calculator sizes the inventory buffer that protects a service level, and the reorder point that goes with it. You give it the mean and variability of daily demand, the mean and variability of the replenishment lead time, the service level you want and your review policy, and it returns the safety stock, the reorder point, the fill rate that results, and what the buffer costs to hold for a year.

It also reports what the safety stock would have been if lead-time variability were ignored, and how much that simpler figure understates the answer.

Demand is in units per day, lead times in days, service level in percent, and costs in your currency.

Use it to set buffers on an item with reasonably stable, normally distributed demand. Both variabilities matter, and the second one is the one most calculations leave out.

Purpose and model boundary

This model estimates safety stock and a reorder point under normally distributed demand with both daily-demand variability and lead-time variability. It supports continuous-review and periodic-review policies, reports a demand-only comparison, and connects cycle service level to expected shortage and fill rate.

It is a single-item statistical approximation. It does not forecast demand, optimize service level, select an order quantity, simulate stockouts, or represent a multi-echelon network.

Inputs and units

Governing relationships

The exposure period is:

E = mu_L for continuous review

E = mu_L + R for periodic review

The service factor z is the inverse standard normal value for p / 100, evaluated by the workbook's rational approximation. Assuming independent demand and lead-time variability, the combined variance and standard deviation are:

Variance = E × sigma_d^2 + mu_d^2 × sigma_L^2

sigma_combined = sqrt(Variance)

The principal inventory relationships are:

Safety stock = z × sigma_combined

Expected demand during exposure = mu_d × E

Reorder point = expected demand during exposure + safety stock

Cycle stock = Q / 2

Average inventory = cycle stock + safety stock

Days of supply = average inventory / mu_d

Annual holding cost of safety stock = safety stock × c × h / 100

The demand-only comparison removes the lead-time-variance term:

Safety stock ignoring lead-time variability = z × sqrt(E × sigma_d^2)

Understatement = full safety stock / demand-only safety stock - 1

For fill rate, the workbook uses the standard normal loss function. With standard normal density phi(z) and upper-tail probability 1 - Phi(z):

Expected shortage per cycle = sigma_combined × (phi(z) - z × (1 - Phi(z)))

Fill rate = median(0, 1, 1 - expected shortage per cycle / Q)

Calculation sequence

1. Resolve review policy and include the review period only for periodic review.
2. Convert the entered cycle service level to an inverse-normal service factor.
3. Combine demand and lead-time variance, then multiply by the service factor.
4. Add expected exposure-period demand to safety stock to produce the reorder point.
5. Derive cycle stock, average inventory, days of supply, and annual safety-stock carrying cost.
6. Calculate the demand-only comparison, expected shortage, and fill rate.
7. Evaluate Model_Status in the order below.

Outputs and interpretation

Reorder_Point is the primary replenishment trigger. Safety_Stock is the variability buffer embedded in that trigger. Cycle_Service_Level is an input probability of no stockout in a cycle; Fill_Rate is a different output measuring the share of demand filled from stock. Understatement_If_It_Is_Ignored quantifies the difference between the full formula and the demand-only approximation when that ratio is defined.

Validation and status logic

The exact returned warning text is preserved above. The workbook condition for its lead-time warning is an understatement ratio greater than 50%.

Assumptions and limitations

• Demand and lead time are approximated as normal and independent. Correlation, seasonality, trends, intermittency, promotions, supplier disruptions, and non-normal tails are not modeled.
• Means and standard deviations must use a consistent daily basis and a representative observation window.
• Continuous review ignores the review-period input. Periodic review extends exposure by the entered review period.
• The service factor is computed by a numerical inverse-normal approximation, not an empirical demand distribution.
• The model allows service levels below 50%; those produce a negative z and can produce negative safety stock. Such a policy is mathematically representable but usually requires explicit business justification.
• Order quantity affects cycle stock and fill rate but not the safety-stock formula. The model does not optimize Q or balance shortage and holding costs.
• The reorder point is not rounded to pack size, pallet multiple, or system-specific integer policy.

Restrictions and non-computing states

This calculator accepts only the two declared review policies, rejects negative numeric inputs, and constrains cycle service level to 0 through 100 percent. The workbook is stricter at its decision boundary: mean demand, mean lead time, and order quantity must be positive, while service level must be strictly between 0 and 100 percent. The review-period input is ignored under continuous review. If demand-only safety stock is zero, the understatement ratio uses error handling and returns zero rather than infinity; interpret the component outputs directly in that case.

Errors and warnings

Values outside the published limits may be rejected before workbook execution. NOT VALID means the demand, lead-time, quantity, or probability basis cannot be used. CHECK retains the result but flags an extreme service target or a large effect from lead-time variability. A calculation-service or network failure is not an inventory-status message.

The workbook derives its relations rather than reproducing any table, chart or figure from a standard or publication. The combined standard deviation, the service factor and the reorder point are the standard formulations.

• Massachusetts Institute of Technology. King, Crash Course in Safety Stock, the derivation covering both demand and lead-time uncertainty. https://web.mit.edu/course/2/2.810/www/files/readings/King_SafetyStock.pdf
• North Carolina State University. Safety Stock Analysis: Inventory Management Models, a tutorial treatment. https://scm.ncsu.edu/scm-articles/article/safety-stock-analysis-inventory-management-models-a-tutorial
• Wikipedia. Safety stock. https://en.wikipedia.org/wiki/Safety_stock
• Wikipedia. Reorder point. https://en.wikipedia.org/wiki/Reorder_point

Demand and lead-time statistics must come from your own history. The defaults here are illustrative and the quality of the estimate matters far more than the precision of the arithmetic.

Additional source notes migrated from Methodology

The workbook applies the independent demand/lead-time variance formula, the inverse standard normal service factor, and the standard normal loss function. The delivered reviewer packet cites Zoho Inventory tools and Omni Calculator for limited comparison; demand-only tools should be compared only with the workbook's demand-only output.