Inventory Management and Optimization

Introduction

Inventory is the stock of materials, components, work-in-process, finished goods, spare parts and consumables held to support operations and customer service. In procurement and materials management, inventory is both an asset and a potential liability. It protects the organisation from demand and supply uncertainty, but it also consumes working capital, storage capacity and managerial attention. Excess inventory can become obsolete, damaged or technologically outdated; insufficient inventory can cause stockouts, production stoppages and lost sales.

Inventory management therefore seeks an economically appropriate balance between availability, cost, uncertainty and obsolescence. The central managerial question is not “How much inventory can be stored?” but “What inventory position and control policy provide the required service at acceptable total cost and risk?”

This chapter integrates classification systems, lot-sizing, reorder-point design, safety stock, service levels, periodic and continuous review, single-period decisions and multi-echelon inventory principles.

Learning Objectives

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

1. Explain the economic and operational roles of inventory.
2. Classify items using ABC, XYZ and FSN methods.
3. Calculate economic order quantity, reorder points and safety stock under stated assumptions.
4. Distinguish cycle service level from fill rate.
5. Compare continuous-review and periodic-review policies.
6. Apply newsvendor logic to short-life and uncertain-demand products.
7. Explain multi-echelon inventory concepts and practical implementation issues.
8. Evaluate the limitations, governance requirements and digital-enablement opportunities of inventory policies.

1. The Role and Economics of Inventory

Inventory performs several functions:

• Decoupling: separates successive operations so that a local disruption does not immediately stop the entire system.
• Anticipation: supports seasonal demand, planned shutdowns, promotions or expected price increases.
• Protection against uncertainty: absorbs variation in demand, lead time, quality and transportation.
• Economies of scale: permits economical purchasing, production and transportation quantities.
• Product availability: supports customer service and production continuity.

The principal inventory categories are raw materials, work-in-process, finished goods, maintenance-repair-operating supplies and goods in transit. Inventory decisions concern quantity, timing, location, ownership and condition—not merely physical counting.

Relevant costs include purchase cost, ordering or setup cost, holding cost, shortage cost, transportation cost, inspection cost, handling cost and obsolescence cost. A useful annual approximation is:

where TC is annual relevant total cost, D is annual demand in units, C is unit purchase cost, Q is order quantity, S is ordering cost per order and H is annual holding cost per unit. This basic expression assumes constant demand, instantaneous replenishment, no shortages and a constant unit price. If purchase price changes with quantity, DC and the price-break structure must be included explicitly.

Inventory turns and days of supply are common diagnostic measures:

The numerator and denominator must be expressed on a consistent basis. High turns are not automatically desirable: an item with high turns but frequent stockouts may indicate understocking.

2. Classification for Differential Control

A single policy for every stock-keeping unit (SKU) is inefficient. Classification directs analytical effort and control intensity toward items whose financial, operational or demand characteristics matter most.

2.1 ABC analysis

ABC analysis ranks items by annual usage value:

Items are sorted in descending order and cumulative percentages are calculated. Typical, but not universal, categories are:

The percentages are conventions rather than laws. A low-value item may be operationally critical, hazardous or legally regulated and should not be treated casually.

2.2 XYZ analysis

XYZ classification concerns demand predictability. A common measure is the coefficient of variation:

where σ_D is the standard deviation of demand over a chosen period and μ_D is mean demand. X items have relatively stable demand, Y items show moderate variation or trend/seasonality, and Z items are highly irregular or intermittent. Boundaries must be calibrated to the organisation and data frequency.

XYZ is not the same as ABC. An AX item is financially important and predictable; an AZ item is financially important but uncertain and deserves forecasting, supplier responsiveness and management attention. A CZ item may be inexpensive and erratic, making order-on-demand, substitution or deliberate nonstocking appropriate.

2.3 FSN analysis

FSN classifies movement:

• Fast-moving: frequent issues or consumption.
• Slow-moving: occasional consumption and prolonged residence.
• Non-moving: no consumption during a defined review period.

FSN is particularly useful for identifying dormant stock, obsolete materials and warehouse-layout opportunities. The time threshold must reflect item type; a spare part with no issue for one year is not necessarily obsolete if it protects an expensive asset.

2.4 Combined classification

A practical inventory segmentation combines financial importance, uncertainty and movement. For example, an A-X-F item may receive continuous review and statistical safety stock. A C-Z-N item may be investigated for deletion, return, substitution or disposal. Classification should be refreshed after major changes in demand, product design, supplier performance or service strategy.

3. Economic Order Quantity

The basic EOQ model determines the order quantity that balances ordering and holding costs. With constant demand and instantaneous replenishment, the average cycle stock is Q/2. The relevant cost is:

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

At Q^*, annual ordering cost equals annual holding cost. The formula requires demand and cost units to be compatible: for example, annual demand in units per year, ordering cost in currency per order and holding cost in currency per unit per year.

Worked example: basic EOQ

Annual demand is 12,000 units, ordering cost is $50 per order and annual holding cost is $2 per unit.

Orders per year are approximately 12,000 ÷ 775 = 15.5. If operating 300 days per year, the average time between orders is approximately 300 ÷ 15.5 = 19.4 working days.

EOQ is a benchmark, not an automatic instruction. Quantity discounts, minimum order quantities, truckload economics, shelf life, supplier capacity, cash constraints and storage limits may justify a different quantity. EOQ also excludes safety stock because safety stock addresses uncertainty, whereas EOQ addresses the cycle-stock trade-off.

Economic production quantity

When replenishment occurs gradually during production, the maximum inventory is less than Q. If production rate p exceeds demand rate d, the economic production quantity is:

This applies to a finite production rate with no shortages and stable rates. It should not be used for purchased items delivered instantaneously.

4. Reorder Points, Lead Time and Safety Stock

A reorder point triggers replenishment. Under constant demand and lead time:

where d is average demand per time unit, L is replenishment lead time in the same time unit and SS is safety stock. The term dL is expected lead-time demand.

4.1 Safety stock for demand variability

If demand during lead time is approximately normally distributed and lead time is constant:

where σ_L is the standard deviation of demand during lead time and z is the standard-normal quantile associated with the target cycle service level. If independent-period demand has standard deviation σ_d and lead time is L periods:

This assumes independent demand observations and stable variance. For correlated, seasonal or intermittent demand, the variance must be estimated using a suitable forecasting model rather than blindly applying this formula.

4.2 Safety stock with demand and lead-time variability

If period demand and lead time are independent, with mean demand μ_d, demand standard deviation σ_d, mean lead time μ_L and lead-time standard deviation σ_L, the variance of lead-time demand can be approximated by:

A more exact expression includes the product term:

The latter is appropriate when both variables are random and independent under the stated distributional assumptions. Then SS = zσ_DL. Correlation between demand and lead time, or non-normal distributions, requires a different treatment, often simulation or empirical quantiles.

Worked example: reorder point

Daily demand has mean 100 units and standard deviation 20 units. Lead time is constant at 4 days. The target cycle service level is 95%, for which z ≈ 1.645.

A continuous-review system places an order when inventory position reaches approximately 466 units. Inventory position is normally defined as on-hand inventory plus on-order inventory minus backorders; using on-hand stock alone can create duplicate orders.

5. Service Levels and Review Policies

A cycle service level is the probability of experiencing no stockout during a replenishment cycle. A 95% cycle service level means that approximately 5% of cycles may experience a stockout under model assumptions. It does not mean that 95% of all demand units are filled.

The fill rate is the proportion of demand units supplied immediately from available stock:

These measures can differ substantially. A system may have occasional stockouts but lose many units during each event, producing a poor fill rate despite a moderate cycle service level. Service targets should therefore be selected according to customer commitments, shortage consequences and economics, not habit.

5.1 Continuous review: (Q, R)

Under a continuous-review policy, inventory position is monitored continuously. When it reaches reorder point R, a fixed quantity Q is ordered. The policy is often represented as (Q,R). It is suitable for high-value, critical or fast-moving items when transaction data and rapid response are available.

Advantages include precise triggering and lower reaction delay. Limitations include system and monitoring costs, transaction errors and the need for reliable lead-time data.

5.2 Periodic review: (P, S)

Under periodic review, inventory is checked every P time units and replenished up to an order-up-to level S. The protection period includes review interval plus lead time. With constant lead time and normally distributed demand:

The order quantity at review is approximately:

Periodic review can consolidate orders and reduce administrative effort, but it requires more protection because demand can occur before the next review. It is common for low-value items, route-based supplier deliveries and organisations with scheduled purchasing cycles.

6. Single-Period Decisions and Newsvendor Logic

The EOQ and reorder-point models are appropriate for replenishment settings where unsold stock can generally be carried forward. For newspapers, fresh food, seasonal fashion or event merchandise, a decision may be made only once before demand occurs. The newsvendor model balances underage and overage costs.

Let c_u be the cost of ordering one unit too few and c_o the cost of ordering one unit too many. The critical fractile is:

where F(Q^*) is the cumulative probability that demand is no greater than the optimal order quantity. If salvage value is v, selling price is p and purchase cost is c, then, in a basic model:

These equations assume that a shortage loses the sale and excess units can be salvaged at v. If shortage causes goodwill loss, markdown effects or substitution, those consequences must be included in c_u.

The model does not seek maximum availability. It deliberately accepts some probability of excess and shortage because their costs are asymmetric. Forecast quality, markdown policy, salvage channels and product life strongly affect the decision.

7. Multi-Echelon Inventory Principles

A supply chain may hold inventory at suppliers, plants, central distribution centres, regional warehouses and customer-facing locations. These levels form a multi-echelon system. Optimising each location independently can increase total inventory because every echelon protects itself against the same uncertainty.

Important principles include:

1. Position inventory by postponement: delay final configuration or deployment until demand information is clearer.
2. Pool uncertainty: centralised inventory can serve several markets and may require less aggregate safety stock when demands are not perfectly correlated.
3. Differentiate service roles: a central warehouse may hold component variety while local facilities hold fast-moving finished goods.
4. Use echelon inventory: evaluate stock relative to downstream commitments, not only physical stock at one location.
5. Avoid double counting: safety stock at upstream and downstream locations should reflect the actual replenishment relationship.

Risk pooling is strongest when demands are not perfectly positively correlated. However, centralisation may increase transport time, delivery cost or disruption exposure. The best network balances inventory pooling benefits against responsiveness and resilience.

8. Integrated Application: Illustrative Case

Background

The illustrative company, MedEquip Services, distributes replacement pumps for hospital water-treatment systems. It stocks 2,400 SKUs. Some pumps are expensive and critical; seals and fasteners are inexpensive but numerous. Stockouts can delay hospital maintenance, while obsolete models have little resale value.

Problem

The company used the same monthly review rule for every item. An internal review found excessive slow-moving inventory, emergency purchases of critical pumps and poor visibility of supplier lead-time variation.

Analysis

The team calculated annual usage value for ABC classification, demand variation for XYZ classification and issue frequency for FSN classification. A high-value, predictable pump was classified AXF. A high-value, irregular legacy component was AZS. Several low-value, non-moving seals were CZN. The team also discovered that planners were using on-hand stock rather than inventory position, ignoring open purchase orders.

For the AXF pump, annual demand was 12,000 units, ordering cost $50 and annual holding cost $2, producing an EOQ of approximately 775 units. Daily demand averaged 100 units with a standard deviation of 20, and lead time was four days. At a 95% cycle service target, safety stock was approximately 66 units and the reorder point 466 units.

For the AZS component, EOQ was not used mechanically. The organisation established a criticality review, supplier commitment, engineering-substitution assessment and a deliberately higher service target. Demand was forecast using intermittent-demand methods and reviewed by an engineer. For CZN items, the company checked maintenance bills of material, discontinued obsolete codes and introduced controlled order-on-demand purchasing.

Approach and solution

MedEquip implemented differentiated policies:

• AX items: continuous review, validated lead times and supplier performance monitoring.
• B items: periodic review with standard parameters.
• C items: simplified two-bin or min–max control, subject to criticality screening.
• Z items: exception-based review, engineering input and explicit obsolescence decisions.
• Non-moving items: monthly disposition review, including return, transfer, repair, reuse or disposal.

The planning system calculated inventory position automatically and separated cycle stock from safety stock. Service metrics included fill rate, cycle service level, backorders, inventory turns, aged inventory and emergency freight.

Outcome

Within the illustrative six-month evaluation, average inventory fell while availability of critical pumps improved. The principal improvement did not come from a single formula; it came from segmentation, reliable master data, explicit service targets and governance of obsolete stock.

Learning points

The case demonstrates that mathematical policies are useful only when their assumptions, data and managerial purpose are understood. Criticality can override ABC rank, and an item with low financial value may require a high service level. Inventory optimisation is therefore a cross-functional activity involving procurement, operations, engineering, finance, sales and suppliers.

9. Technology, Governance and Contemporary Practice

Enterprise resource planning, warehouse-management systems, barcode scanning, RFID and supplier portals improve transaction visibility. Advanced analytics can estimate lead-time distributions, detect parameter drift, identify abnormal consumption and simulate service-cost trade-offs. Machine-learning forecasts may help where sufficient, stable and relevant data exist, but they do not eliminate the need for causal understanding, human review or inventory-policy discipline.

Governance should define item ownership, parameter approval, review frequency, exception thresholds and audit trails. Important controls include cycle counting, segregation of duties, approved supplier records, prevention of duplicate SKUs and reconciliation between physical, system and financial inventory.

Ethical and regulatory considerations are important in pharmaceuticals, food, medical equipment, chemicals and hazardous materials. Traceability, expiry control, recalls, temperature records, lot control and disposal regulations may dominate economic calculations. Deliberately reducing stock below a safe level to improve financial metrics is unacceptable where human safety or legal compliance is at risk.

Contemporary practice increasingly combines demand sensing, postponement, supplier collaboration, repair and refurbishment, circular-economy principles and scenario-based resilience planning. Sustainability analysis should include energy, packaging, waste, reverse logistics and the environmental cost of obsolescence—not only warehouse carrying cost.

10. Limitations and Best Practices

Inventory models simplify reality. Demand may be intermittent, seasonal or promotion-driven; lead times may be correlated with demand; supplier minimums may invalidate EOQ; costs may be nonlinear; and normal-distribution assumptions may be poor for slow movers. Forecast error, inaccurate bills of material, unrecorded withdrawals and parameter inertia can undermine even a theoretically correct policy.

Best practices are therefore to:

1. Segment by value, variability, movement and criticality.
2. Use inventory position, not merely on-hand quantity.
3. Match service measures to customer and operational consequences.
4. Estimate demand and lead-time variability from clean, relevant data.
5. Review obsolete and non-moving stock explicitly.
6. Treat EOQ as a benchmark subject to operational constraints.
7. Coordinate policies across echelons and avoid double safety stock.
8. Monitor forecast bias, supplier reliability and parameter performance.
9. Integrate procurement, engineering, finance and operations decisions.
10. Test policies through sensitivity analysis and scenario simulation.

Chapter Summary

Inventory management balances availability, cost, uncertainty and obsolescence. ABC classifies items by annual usage value; XYZ describes demand predictability; and FSN identifies movement. Their combination enables differentiated control rather than a one-policy-fits-all system.

EOQ balances ordering and holding costs under restrictive assumptions. Reorder points combine expected lead-time demand with safety stock. Safety-stock calculations depend on variability, service-level definitions and the statistical behaviour of demand and lead time. Cycle service level and fill rate measure different aspects of performance. Continuous review provides precise triggers, while periodic review offers administrative and consolidation benefits but requires protection over a longer period.

Newsvendor logic applies to single-period or short-life products by balancing underage and overage costs. Multi-echelon management coordinates inventory across locations, exploits risk pooling and prevents duplicated protection. Effective implementation requires accurate data, governance, cross-functional judgement, technology and explicit treatment of criticality, regulation and sustainability.

Key Terms

Inventory: Stock held to support operations or customer demand.

Stock-keeping unit (SKU): A distinct item identified for planning and control.

Cycle stock: Inventory created by ordering or producing in batches.

Safety stock: Additional inventory held against uncertainty.

ABC analysis: Classification by annual usage value.

XYZ analysis: Classification by demand variability or predictability.

FSN analysis: Classification by movement frequency.

EOQ: Order quantity that balances ordering and holding costs under stated assumptions.

Reorder point: Inventory position at which replenishment is triggered.

Lead-time demand: Demand expected during replenishment lead time.

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

Cycle service level: Probability of no stockout during a replenishment cycle.

Fill rate: Proportion of demand units supplied immediately.

Continuous review: Inventory monitoring performed continuously or transaction by transaction.

Periodic review: Inventory checked at fixed time intervals.

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

Multi-echelon inventory: Inventory distributed across successive supply-chain levels.

Obsolescence: Loss of economic or functional value due to age, change or discontinuation.

Risk pooling: Reduction of aggregate uncertainty through demand aggregation.

Review Questions

5 Short Answer Questions

1. What are the principal operational functions of inventory?
2. How does XYZ classification differ from ABC classification?
3. Define inventory position.
4. Distinguish cycle service level from fill rate.
5. State two limitations of the basic EOQ model.

5 Long Answer Questions

1. Explain how ABC, XYZ and FSN classifications can be combined to design differentiated inventory policies.
2. Derive the basic EOQ relationship and discuss its assumptions and managerial limitations.
3. Explain the calculation and interpretation of safety stock when demand varies but lead time is constant.
4. Compare continuous-review and periodic-review systems in terms of responsiveness, cost and data requirements.
5. Discuss how multi-echelon principles and risk pooling influence supply-chain inventory decisions.

3 Analytical / Application Questions

1. Daily demand is 80 units with a standard deviation of 15 units. Lead time is five days and the target cycle service level is 98% (z = 2.054). Calculate safety stock and reorder point, assuming independent daily demand and constant lead time.
2. A seasonal product sells for $40, costs $18 and has a salvage value of $8. Calculate the critical fractile and explain what additional information is needed to determine the order quantity.
3. An organisation has a low-value spare part with no issues for 18 months, but failure of the related machine would stop a production line. Recommend an appropriate classification and control approach, explaining why financial value alone is insufficient.

Multiple Choice Questions with Answers

Question 1

Which classification primarily ranks inventory items by annual usage value?

A. ABC

B. XYZ

C. FSN

D. Newsvendor

Correct Answer: A. ABC

Question 2

Annual demand is 10,000 units, ordering cost is $40 per order and annual holding cost is $2 per unit. What is the approximate EOQ?

A. 200 units

B. 447 units

C. 632 units

D. 1,000 units

Correct Answer: C. 632 units

Question 3

Which quantity represents expected demand during replenishment lead time when demand is constant?

A. Safety stock

B. dL

C. Q/2

D. zσ

Correct Answer: B. dL

Question 4

A 95% cycle service level means that the system will:

A. Fill exactly 95% of all demand units

B. Avoid stockout in approximately 95% of replenishment cycles

C. Hold 95% more inventory than before

D. Have a 95-day lead time

Correct Answer: B. Avoid stockout in approximately 95% of replenishment cycles

Question 5

Which policy monitors inventory position continuously and orders a fixed quantity at a trigger level?

A. (P,S) periodic review

B. Newsvendor policy

C. (Q,R) continuous review

D. FSN policy

Correct Answer: C. (Q,R) continuous review

Question 6

If daily demand standard deviation is 10 units and constant lead time is 9 days, what is the standard deviation of lead-time demand under independent demand?

A. 10 units

B. 19 units

C. 30 units

D. 90 units

Correct Answer: C. 30 units

Question 7

In a basic newsvendor model, the critical fractile is calculated as:

A. c_o ÷ c_u

B. c_u ÷ (c_u + c_o)

C. c_u − c_o

D. c_u × c_o

Correct Answer: B. c_u ÷ (c_u + c_o)

Question 8

Which item is most likely to require special attention despite low annual usage value?

A. A noncritical office supply

B. A critical spare part whose failure stops production

C. A regularly consumed packaging label

D. A low-cost item with reliable local supply

Correct Answer: B. A critical spare part whose failure stops production

Question 9

What is a major benefit of risk pooling in a multi-echelon network?

A. It always eliminates transportation cost

B. It can reduce aggregate safety stock when demands are not perfectly correlated

C. It guarantees zero stockouts

D. It removes the need for forecasting

Correct Answer: B. It can reduce aggregate safety stock when demands are not perfectly correlated

Question 10

Which action best addresses non-moving inventory?

A. Automatically increase its safety stock

B. Ignore it because it has no current demand

C. Review return, transfer, reuse, substitution or disposal options

D. Classify it as fast-moving

Correct Answer: C. Review return, transfer, reuse, substitution or disposal options