Inventory Planning and Control

Introduction and context

Inventory is the stock of materials, components, work in process, finished goods, spare parts, and trading goods held to support future operations or customer demand. In supply chains, inventory performs a useful buffering function: it decouples supply from demand, protects operations against uncertainty, supports economies of scale, and improves responsiveness. It also consumes working capital, occupies space, creates obsolescence risk, and can conceal process problems.

Inventory planning and control therefore involves a central managerial trade-off. Too little inventory causes stockouts, lost sales, production interruptions, and poor customer service. Too much inventory increases holding cost, damage, shrinkage, expiry, obsolescence, and financial exposure. The objective is not simply to minimize inventory, but to achieve an economically appropriate balance between availability, cost, risk, and strategic responsiveness.

This chapter develops the principal models used for that balance: inventory classification, economic order quantity, safety stock, reorder-point systems, periodic review, service-level analysis, multi-echelon inventory, and the newsvendor model. The models are simplified representations of operational reality. Their value depends on appropriate assumptions, reliable data, and disciplined managerial judgment.

Learning objectives

By the end of this chapter, the reader should be able to:

1. Classify inventory according to purpose, value, demand characteristics, and supply risk.
2. Explain the cost and service trade-offs in inventory decisions.
3. Apply EOQ, production-order-quantity, reorder-point, safety-stock, and periodic-review models.
4. Relate service levels to demand variability and inventory investment.
5. Distinguish continuous-review, periodic-review, single-period, and multi-echelon decisions.
6. Interpret model assumptions, limitations, and sensitivity results.
7. Design an integrated inventory policy using data, technology, governance, and managerial judgment.

1. The role and economics of inventory

Inventory may be held for several reasons:

• Cycle stock results from ordering or producing in batches.
• Safety stock protects against uncertainty in demand, lead time, or supply.
• Anticipation stock is accumulated before seasonal peaks, promotions, strikes, or expected price increases.
• Pipeline inventory is in transit between supply-chain stages.
• Decoupling stock separates operations so that a disruption in one process does not immediately stop another.
• Speculative inventory is held because of expected price, exchange-rate, or supply changes.
• MRO inventory consists of maintenance, repair, and operating supplies.

The principal annual inventory-cost components are purchasing or production cost, ordering or setup cost, holding cost, shortage cost, and disposal or obsolescence cost. If annual demand is D units, order quantity is Q units, ordering cost per order is S, unit purchase cost is c, and annual holding-cost rate is i, then annual holding cost per unit is h = ic. Under constant demand and instantaneous replenishment, average cycle stock is Q ÷ 2. The basic annual relevant cost is therefore:

Here, TC(Q) is annual total relevant cost, DC is annual purchase cost, (D ÷ Q)S is annual ordering cost, and (Q ÷ 2)h is annual cycle-stock holding cost. The formula excludes safety stock and shortage cost and assumes that all demand is eventually served.

The model demonstrates why neither extremely small nor extremely large orders are generally optimal. Small orders increase order frequency and ordering cost; large orders increase average inventory and holding cost.

2. Inventory classification and segmentation

A single policy should not be applied to every item. Classification creates differentiated controls.

ABC classification

ABC analysis ranks items by annual dollar usage:

A common, though not universal, classification is:

Percentages are conventions rather than laws. An inexpensive item may nevertheless be operationally critical, while a high-value item may have little service impact.

Additional classifications

• VED: vital, essential, desirable; useful for spare parts and healthcare supplies.
• FSN: fast-moving, slow-moving, non-moving; useful for detecting dormant stock.
• HML: high, medium, low unit price.
• SDE: scarce, difficult, or easy to procure.
• XYZ: commonly based on demand or inventory-value variability.
• Intermittent-demand classification: separates frequent demand from lumpy, sporadic demand.

A robust segmentation combines dimensions. For example, an A–V item is financially important and vital to operations, so it deserves close review, high-quality master data, and explicit contingency plans. A C–D item may be inexpensive and desirable rather than essential, making a simple two-bin system appropriate.

Classification should be dynamic. Changes in demand, product design, supplier performance, criticality, and regulatory requirements can move an item from one segment to another.

3. Economic order quantity

Derivation of the basic EOQ

Under the basic EOQ assumptions—known and constant demand, constant lead time, instantaneous replenishment, no shortages, constant unit price, and independent replenishment cycles—the optimal order quantity minimizes relevant annual cost:

Differentiating with respect to Q and setting the derivative equal to zero gives:

Thus:

Q is the economic order quantity. D must be expressed in units per year, S in currency per order, and h in currency per unit per year. The resulting Q is in units per order.

At the optimum, annual ordering cost equals annual cycle-stock holding cost:

The number of orders per year is D ÷ Q*, and the average time between orders is the operating period divided by that number of orders. EOQ is a cost benchmark, not an instruction to ignore supplier constraints, transport economics, minimum order quantities, or service requirements.

Worked example: EOQ

A distributor sells 24,000 units annually. The ordering cost is $75 per order and the annual holding cost is $3 per unit. Then:

The firm should order approximately 1,095 units under the model assumptions. Annual ordering cost is approximately (24,000 ÷ 1,095) × 75 = $1,644, and annual cycle-stock holding cost is approximately (1,095 ÷ 2) × 3 = $1,643. Small differences result from rounding.

Extensions and limitations

If replenishment occurs gradually while demand continues, the economic production quantity is:

where p is the production rate and d is the demand rate, both in compatible units per time, with p > d. The factor p ÷ (p − d) increases the optimal lot size because maximum inventory is lower than Q when production and consumption overlap.

Quantity discounts require comparison of feasible price-break quantities. For each price, calculate purchase cost plus ordering and holding costs, verify that the candidate quantity qualifies for the price, and compare total annual cost. The lowest EOQ at a price may be infeasible if it does not meet the minimum quantity for that price.

EOQ is less reliable when demand is highly intermittent, lead time is unstable, prices change frequently, shortages are strategically unacceptable, or fixed costs are not approximately constant. In such settings, simulation, optimization, or service-based policies may be preferable.

4. Uncertainty, safety stock, and service levels

EOQ determines cycle stock but does not protect against uncertainty. Safety stock is additional inventory held to reduce the probability or expected magnitude of shortage. It is not a substitute for improving forecasts, reducing lead time, or correcting supplier unreliability.

Two service measures must be distinguished:

• Cycle service level (CSL): probability of no stockout during a replenishment cycle.
• Fill rate: proportion of demand units supplied immediately from available stock.

A 95% CSL does not necessarily mean a 95% fill rate. A single stockout may affect many units, and the relationship depends on demand size and shortage timing.

Reorder point with variable demand and constant lead time

For continuous review, the reorder point is:

where μ_L is expected demand during lead time and SS is safety stock. If average demand per period is μ_d and lead time is L periods:

If demand during lead time is approximately normal and the desired cycle service level corresponds to standard-normal quantile z, then:

where σ_L is the standard deviation of lead-time demand. For independent, identically distributed demand with standard deviation σ_d per period and fixed lead time L:

The units are important: μ_d and σ_d must refer to the same time period as L.

Variable demand and variable lead time

If demand and lead-time uncertainty are independent, average demand per period is μ_d, demand standard deviation is σ_d, mean lead time is μ_Lt, and lead-time standard deviation is σ_Lt, then a commonly used approximation is:

The first term represents demand variation over the average lead time; the second represents variation caused by lead-time uncertainty. Safety stock is then approximately zσ_{DL}. This approximation assumes independent periods and independence between demand and lead time. If they are correlated, a covariance term is required and empirical lead-time-demand observations are often safer than the approximation.

Worked example: reorder point

Daily demand has mean 100 units and standard deviation 20 units. Lead time is constant at 5 days. The target CSL is 95%, for which z ≈ 1.645. Expected lead-time demand is:

Lead-time demand standard deviation is:

Safety stock is:

Thus:

The operational reorder point is approximately 574 units. When inventory position reaches this level, an order is released. Inventory position equals on-hand inventory plus on-order inventory minus backorders.

A high target service level has a nonlinear cost effect because z rises sharply as the target approaches 100%. Managers should therefore select service levels by item segment and customer consequence rather than adopting 99% universally.

5. Continuous-review and periodic-review systems

Continuous-review (Q, R) policy

A continuous-review policy orders a fixed quantity Q when inventory position reaches reorder point R. It is appropriate when transactions and stock levels can be monitored continuously and when critical items justify frequent control. A typical policy is:

The policy separates the lot-size decision from the timing decision. Q may be EOQ or a practical quantity, while R reflects expected lead-time demand and desired protection.

Periodic-review (P, S) policy

Under periodic review, inventory is inspected every P time units and replenished to an order-up-to level S. The order quantity varies:

Protection is required over the review period plus lead time. With constant demand and normal uncertainty:

If demand is independent with standard deviation σ_d per period:

Periodic review is convenient when suppliers visit on schedules, orders are consolidated, or many low-value items are managed together. Its disadvantage is that stock can fall substantially between reviews, so it normally requires more safety stock than equivalent continuous review.

Two-bin systems and kanban

A two-bin system places a replenishment signal in the first bin. Consumption from the second bin continues while the first is replenished. Kanban extends this logic through visual or electronic pull signals. These systems work best with stable processes, disciplined replenishment, standardized containers, and controlled variability. They are not a remedy for chronically inaccurate records or unstable suppliers.

6. Single-period decisions and the newsvendor model

The newsvendor model applies when inventory has little or no value after a selling period: fresh food, newspapers, fashion goods, event capacity, or seasonal products. The decision is how much to stock before uncertain demand is observed.

Let selling price be p, purchase cost be c, salvage value be v, and shortage or lost-margin consequence be represented by the opportunity cost of underage. Assuming v ≤ c ≤ p, overage cost is:

and underage cost is:

The critical fractile is:

Thus, Q* is the demand quantile whose cumulative probability equals the critical fractile. The formula assumes one replenishment decision, known cost parameters, and a specified demand distribution.

Worked example: newsvendor

A retailer buys a seasonal item for $40, sells it for $70, and can salvage unsold units for $10. Therefore:

The critical fractile is 30 ÷ (30 + 30) = 0.50. If demand is normally distributed with mean 1,000 and standard deviation 200, the 50th percentile equals the mean. The recommended order is therefore approximately 1,000 units. A higher salvage value would reduce overage cost and increase the target quantity; a higher selling margin would increase the target quantity.

The newsvendor model exposes a major distinction: replenishment-system inventory is often evaluated over repeated cycles, whereas single-period inventory is evaluated through the consequences of overage and underage for one decision horizon.

7. Multi-echelon inventory planning

A multi-echelon supply chain contains several linked stocking points, such as a plant, regional distribution centre, and store. Local optimization can produce system-wide inefficiency. For example, every location may hold high safety stock because each faces the same uncertainty independently, even though upstream consolidation could provide protection more economically.

An echelon inventory position includes inventory at a stage plus inventory downstream and in transit, adjusted for committed demand. Echelon-stock policies allocate protection across stages rather than treating each node as an isolated warehouse.

Key principles include:

1. Risk pooling: aggregation reduces relative demand variability when demands are not perfectly correlated.
2. Postponement: delaying final configuration or customization reduces finished-goods variety and obsolescence.
3. Strategic positioning: upstream stock may be cheaper and more flexible, while downstream stock improves response time.
4. Lead-time decomposition: total customer lead time is distributed across transportation, processing, queueing, and decision delays.
5. Allocation discipline: scarce stock should be allocated according to contractual commitments, contribution, criticality, and fairness—not merely political pressure.

Multi-echelon optimization may use stochastic programming, simulation, base-stock policies, or installation-stock models. The practical challenge is data synchronization: demand, inventory, lead time, service requirements, and capacity must be visible across the network. A mathematically optimal solution based on incorrect lead times or phantom inventory is not operationally optimal.

8. A managerial framework for inventory decisions

An effective inventory-control process follows a sequence rather than relying on a single formula.

Define the service promise

Specify the customer, delivery point, response time, acceptable backorder, and relevant service measure. A critical spare part may require a high probability of immediate availability, while a low-value commodity may tolerate a delay.

Establish data integrity

Validate item identifiers, units of measure, bills of material, supplier lead times, minimum order quantities, pack sizes, inventory balances, demand history, and transaction discipline. Separate genuine demand from cancellations, returns, substitutions, promotions, and one-time projects.

Segment items

Combine financial value with criticality, variability, supply risk, lifecycle status, and customer impact. Assign policies, review frequencies, approval limits, and escalation rules to each segment.

Select the policy

Use EOQ or practical lot sizing for cycle stock; reorder-point policies for continuous monitoring; periodic review for scheduled replenishment; newsvendor logic for short-life products; and multi-echelon analysis for network inventory.

Measure performance

Useful measures include:

• inventory turnover = annual cost of goods sold ÷ average inventory value;
• days of supply = average inventory ÷ average daily demand;
• fill rate and cycle service level;
• stockout frequency and backorder age;
• forecast bias and forecast error;
• obsolete and excess inventory;
• inventory record accuracy;
• supplier on-time-in-full performance;
• cash-to-cash cycle time.

No single metric is sufficient. High turnover may reflect excellent flow or unacceptable stockouts. High availability may reflect customer-focused service or excessive inventory.

9. Illustrative case: Northstar Medical Devices

This is an illustrative case, not a report of an actual company.

Background

Northstar manufactures diagnostic equipment. Its service organization supports hospitals with replacement sensors. The sensors are inexpensive relative to the equipment downtime they prevent, but demand is intermittent and supplier lead time varies between 10 and 25 days.

Problem

The purchasing department used a fixed monthly order quantity based on average demand. Hospitals experienced occasional stockouts, while the central warehouse accumulated obsolete sensor variants after engineering changes. Inventory records also included units reserved for field technicians but not physically available for general orders.

Analysis

The item was classified as C by annual dollar usage but V by operational criticality. This demonstrated why ABC analysis alone was inadequate. The team measured actual lead-time demand, separated service demand from planned installations, corrected inventory-status codes, and identified product variants approaching obsolescence.

A continuous-review policy was selected for the critical sensor. The reorder point was based on empirical demand during actual lead-time observations rather than assuming constant lead time. A higher CSL was justified by the downtime cost, but the target was reviewed against the cost of emergency shipment and alternative sensor certification. For less critical accessories, periodic review and pooled stocking were retained.

Approach and solution

Northstar introduced: (1) segmented policies; (2) electronic inventory-position visibility; (3) supplier lead-time agreements; (4) engineering-change alerts linked to purchasing; (5) allocation rules for constrained stock; and (6) a monthly excess-and-obsolescence review. The firm also postponed final packaging for some variants until customer configuration was known.

Outcome

The policy reduced emergency shipments and improved availability without simply increasing total stock. The most important improvement came from correcting item status and aligning control intensity with criticality. The case illustrates that inventory models create value only when supported by governance, accurate data, cross-functional coordination, and differentiated service commitments.

Learning points

1. Criticality can dominate financial classification.
2. Lead-time demand should be measured consistently with the chosen policy.
3. Engineering changes are inventory events, not solely design events.
4. Visibility of on-order, reserved, and usable stock is essential.
5. Reducing uncertainty may be more valuable than adding safety stock.

10. Technology, AI, and governance

Enterprise resource planning, warehouse-management systems, barcode scanning, RFID, IoT sensors, and supplier portals can improve transaction accuracy and inventory visibility. Advanced planning systems can calculate forecasts, safety stocks, allocation plans, and multi-echelon policies.

AI and machine learning are useful when demand is influenced by many variables, such as promotions, weather, web traffic, and customer attributes. However, predictive accuracy does not automatically produce good inventory decisions. Models require stable definitions, representative training data, monitoring for drift, explainable overrides, and explicit treatment of stockouts, which censor observed demand. If a product was unavailable, recorded sales understate true demand.

Governance should define who owns item master data, who may change service targets, how overrides are documented, and how model performance is audited. Ethical concerns include discriminatory allocation of scarce medical products, misuse of customer data, and opaque prioritization. Regulatory requirements may govern traceability, expiry control, cold-chain integrity, recall management, and record retention.

11. Challenges and limitations

Inventory models face several recurring limitations:

• Demand may be nonstationary because of trends, product launches, promotions, and substitution.
• Demand observations may be censored by previous stockouts.
• Lead times may be correlated with demand surges or supplier disruptions.
• Costs such as shortage penalties, obsolescence, and expediting are difficult to estimate.
• Normal-distribution assumptions may be poor for intermittent or lumpy demand.
• Forecast improvements can be offset by poor execution and inaccurate records.
• Local incentives may encourage purchasing discounts while increasing total supply-chain cost.
• Sustainability is not captured by traditional cost alone; emissions, waste, packaging, and energy should be included where material.

Managers should conduct sensitivity analysis. For example, calculate inventory and service consequences under alternative demand variability, lead time, holding cost, and service targets. Robust policies often outperform theoretically optimal but fragile policies.

Best-practice checklist

• Classify by both value and operational consequence.
• Use the correct time units in every calculation.
• Distinguish cycle stock from safety stock.
• Define service level precisely.
• Use empirical lead-time demand when assumptions are doubtful.
• Review parameters after structural changes, not merely on a calendar.
• Include on-order, reserved, damaged, and obsolete stock in inventory visibility.
• Integrate procurement, sales, operations, engineering, finance, and logistics.
• Evaluate total cost and customer impact across the network.
• Monitor exceptions and root causes rather than relying only on aggregate averages.

Chapter Summary

Inventory planning balances product availability, working capital, operating continuity, and uncertainty. Classification systems such as ABC, VED, FSN, and XYZ support differentiated policies, but financial value alone is insufficient. EOQ explains the trade-off between ordering and holding costs under stable assumptions. Reorder-point systems combine expected lead-time demand with safety stock, while periodic-review systems protect demand over the review period plus lead time. Service level must be defined carefully because cycle service level and fill rate are different measures.

The newsvendor model addresses single-period decisions through underage and overage costs. Multi-echelon planning extends the analysis across plants, distribution centres, and customer-facing locations, using risk pooling, postponement, and echelon inventory concepts. In practice, inventory outcomes depend as much on data integrity, segmentation, governance, supplier reliability, and cross-functional coordination as on formulas. Technology and AI can improve visibility and forecasting, but they require sound controls, ethical allocation rules, and continuous validation.

Key Terms

Inventory: Stock held for future use, sale, or operational support.

Cycle stock: Inventory resulting from replenishment in batches.

Safety stock: Additional inventory held against uncertainty.

Anticipation stock: Inventory accumulated for expected future events or demand.

Pipeline inventory: Stock currently in transit.

ABC analysis: Classification based on annual usage value.

VED analysis: Classification by vital, essential, and desirable criticality.

EOQ: Order quantity that minimizes ordering and cycle-stock holding cost under stated assumptions.

Ordering cost: Cost incurred each time an order or setup is made.

Holding cost: Cost of carrying inventory over time.

Reorder point: Inventory position at which replenishment is triggered.

Inventory position: On-hand plus on-order minus backorders or commitments.

Cycle service level: Probability of avoiding a stockout during a replenishment cycle.

Fill rate: Fraction of demand supplied immediately from available stock.

Periodic review: Inventory system reviewed at fixed time intervals.

Newsvendor model: Single-period model balancing overage and underage costs.

Risk pooling: Reduction in relative uncertainty through aggregation.

Echelon inventory: Inventory viewed across a stage and its downstream network.

Obsolescence: Loss of inventory value because of age, technology, regulation, or demand change.

Inventory turnover: Annual cost of goods sold divided by average inventory value.

Review Questions

Short Answer Questions

1. Why is inventory both an asset and a cost in supply-chain management?
2. Distinguish cycle stock from safety stock.
3. What does ABC analysis measure, and why should it be supplemented by criticality analysis?
4. Define cycle service level and fill rate.
5. State two assumptions of the basic EOQ model.

Long Answer Questions

1. Derive the basic EOQ formula and explain the managerial meaning of the cost balance at the optimum.
2. Explain how demand and lead-time variability affect safety-stock requirements.
3. Compare continuous-review and periodic-review inventory systems, including their advantages and limitations.
4. Explain how the newsvendor model differs from a replenishment model used for continuing demand.
5. Discuss the role of multi-echelon inventory planning in a geographically dispersed supply chain.

Analytical / Application Questions

1. Daily demand has a mean of 80 units and a standard deviation of 15 units. Lead time is constant at 4 days, and the target cycle service level is 95% with z = 1.645. Calculate expected lead-time demand, safety stock, and reorder point.
2. A product has annual demand of 18,000 units, ordering cost of $60 per order, and annual holding cost of $2.40 per unit. Calculate EOQ and explain two practical reasons why the implemented order quantity might differ.
3. A seasonal product sells for $90, costs $50, and has a salvage value of $20. Demand is normally distributed with mean 2,000 and standard deviation 400. Calculate the critical fractile and identify the corresponding approximate order quantity using the standard-normal distribution.

Multiple Choice Questions with Answers

Question 1

Which inventory type primarily protects against uncertainty in demand or lead time?

A. Cycle stock

B. Safety stock

C. Pipeline inventory

D. Speculative stock

Correct Answer: B. Safety stock

Question 2

A basic EOQ model assumes that replenishment is:

A. Instantaneous

B. Randomly delayed

C. Limited to one order per year

D. Always subject to shortages

Correct Answer: A. Instantaneous

Question 3

If annual demand is 10,000 units, ordering cost is $50, and annual holding cost is $2 per unit, the EOQ is approximately:

A. 224 units

B. 500 units

C. 707 units

D. 1,000 units

Correct Answer: C. 707 units

Question 4

With constant lead time, reorder point is generally calculated as:

A. EOQ plus annual demand

B. Expected lead-time demand plus safety stock

C. Annual demand divided by holding cost

D. Safety stock minus lead-time demand

Correct Answer: B. Expected lead-time demand plus safety stock

Question 5

A 95% cycle service level means:

A. 95% of annual demand is forecast correctly

B. 95% of orders are placed on time

C. There is approximately a 95% probability of no stockout during a replenishment cycle

D. Inventory turnover must be 95 times per year

Correct Answer: C. There is approximately a 95% probability of no stockout during a replenishment cycle

Question 6

In a periodic-review system, the order quantity is normally:

A. Always equal to EOQ

B. Equal to the order-up-to level minus inventory position

C. Equal to annual demand divided by lead time

D. Fixed at the safety-stock quantity

Correct Answer: B. Equal to the order-up-to level minus inventory position

Question 7

In the newsvendor model, increasing salvage value while other variables remain constant generally:

A. Raises overage cost and lowers the target quantity

B. Lowers overage cost and raises the target quantity

C. Has no effect on the critical fractile

D. Eliminates underage cost

Correct Answer: B. Lowers overage cost and raises the target quantity

Question 8

Which classification focuses on whether an item is vital, essential, or desirable?

A. ABC

B. FSN

C. VED

D. HML

Correct Answer: C. VED

Question 9

Risk pooling is valuable because aggregation can:

A. Eliminate all demand uncertainty

B. Reduce relative variability when demands are not perfectly correlated

C. Guarantee zero holding cost

D. Make lead times identical

Correct Answer: B. Reduce relative variability when demands are not perfectly correlated

Question 10

A product sells for $70, costs $40, and has salvage value of $10. Its critical fractile is:

A. 0.25

B. 0.50

C. 0.67

D. 0.75

Correct Answer: B. 0.50