Inventory and Replenishment Management

Introduction and Context

Inventory is the stock of materials, components, work in process, finished goods, spare parts and supplies held to support future operations or customer demand. Inventory management is not simply a matter of keeping quantities low. It is the coordinated process of deciding what to stock, how much to hold, when to replenish, where to position stock and how to control its risk and cost.

In logistics and distribution, inventory performs several functions. It decouples stages of a supply chain, protects operations from uncertain demand and lead time, permits economic purchasing or production quantities, supports product variety and improves customer service. However, inventory also ties up working capital, occupies space, incurs handling and insurance costs, becomes obsolete and may conceal process problems.

The central managerial problem is therefore a balance: sufficient inventory must be available when required, but unnecessary inventory must be avoided. This chapter develops the principal analytical tools for that balance: ABC analysis, economic order quantity (EOQ), reorder-point systems, safety stock, periodic and continuous review, service-level measures and multi-echelon inventory management.

Learning Objectives

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

• classify inventory using ABC analysis and interpret its managerial implications;
• calculate and critique EOQ, reorder points and safety stock;
• distinguish continuous-review and periodic-review systems;
• relate inventory decisions to cycle service level, fill rate, working capital and obsolescence;
• explain multi-echelon inventory positioning and the bullwhip effect;
• design practical replenishment policies using data, technology and governance controls.

1. Inventory as a Supply-Chain Resource

Inventory may be classified by its position and purpose:

The main inventory costs are ordering or setup cost, holding cost, purchase cost and shortage cost. Holding cost commonly includes capital cost, storage, insurance, damage, shrinkage and obsolescence. A useful annual approximation is:

where H is annual holding cost per unit (currency/unit/year), i is the annual holding-rate fraction, and C is unit acquisition cost (currency/unit). The relation is an approximation: the holding rate must be defined consistently and should not double-count costs already included elsewhere.

Inventory decisions should also recognize different demand types. Independent demand arises from external customers and is forecast statistically. Dependent demand is derived from a production plan or bill of materials and is normally managed through material requirements planning (MRP), rather than by applying a simple independent-demand reorder formula.

2. ABC Analysis and Inventory Segmentation

ABC analysis ranks items according to annual consumption value, usually calculated as:

where ACV_j is annual consumption value for item j (currency/year), D_j is annual unit demand (units/year), and C_j is unit cost (currency/unit). Items are then ordered from highest to lowest ACV and cumulative percentages are calculated.

A conventional segmentation is:

These percentages are guidelines, not laws. ABC analysis is a value-prioritisation technique, not a complete risk classification. A low-value fire extinguisher, pharmaceutical component or safety-critical spare may deserve tighter control than its monetary value suggests. Organisations therefore often combine ABC with criticality, demand variability, lead time, substitutability, shelf life and supply risk. A resulting matrix may identify, for example, “AX” items as high-value and stable, “AZ” items as high-value and erratic, and “CX” items as low-value and stable.

ABC analysis should be refreshed when prices, demand, product portfolios or supply conditions change. Its managerial purpose is differentiated control: not every item warrants the same counting frequency, approval level, forecasting effort or service target.

3. Economic Order Quantity

The basic EOQ model determines the order quantity that minimises relevant annual ordering and holding costs. Under the standard assumptions of constant known demand, constant lead time, instantaneous replenishment, no shortages, fixed ordering cost and linear holding cost, the total relevant annual cost is:

where Q is order quantity (units/order), D is annual demand (units/year), S is ordering or setup cost (currency/order), and H is annual holding cost per unit (currency/unit/year). The first term is annual ordering cost and the second is average cycle-stock holding cost.

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

The square-root result is valid only under the stated assumptions. If replenishment is gradual during production, the production-order quantity model is more appropriate; if quantity discounts, shortages, capacity constraints or uncertain demand are material, the basic EOQ must be modified or used only as a benchmark.

Worked Example: Basic EOQ

A distributor sells 24,000 units per year. The ordering cost is $45 per order and annual holding cost is $3 per unit. Thus:

Q* = √(2 × 24,000 × 45 ÷ 3) = √720,000 ≈ 848.5

The recommended order quantity is approximately 849 units. The approximate number of orders per year is D/Q* = 24,000/848.5 ≈ 28.3. EOQ is not a mandatory order size; it is a cost-based reference point that must be checked against supplier pack sizes, truckload economies, cash constraints and shelf-life limits.

The model’s key insight is the trade-off: larger orders reduce ordering frequency but increase average cycle stock; smaller orders do the reverse. At the optimum, annual ordering cost and annual holding cost are equal under the basic model.

4. Reorder Point and Lead-Time Demand

EOQ answers “how much?” A reorder point (ROP) answers “when?” Under continuous review, an order is triggered when inventory position reaches the ROP. Inventory position is:

The simplest reorder point is expected demand during lead time:

where d is average demand per time unit and L is replenishment lead time in the same time unit. Unit consistency is essential: daily demand must be multiplied by lead time in days.

Because demand and lead time may vary, safety stock is added:

where μ_L is expected lead-time demand (units) and SS is safety stock (units). If daily demand is independent and normally distributed with standard deviation σ_d, and lead time is fixed at L, then:

Here z is the standard-normal factor corresponding to the selected cycle service level. This formula assumes that demand observations are appropriately represented by a normal distribution and that lead time is fixed.

When both demand and lead time vary independently, a commonly used approximation is:

where the first σ_L on the left is the standard deviation of lead-time demand, the σ_L in the final term is standard deviation of lead time, d is average demand per period, and σ_d is standard deviation of demand per period. To avoid ambiguity, denote lead-time standard deviation as σ_{LT]:

Then SS = zσ_LD. This expression assumes independent demand and lead time and uses consistent time units. If they are correlated, a covariance term is required:

where ρ is the correlation coefficient. In practice, empirical simulation or historical lead-time-demand observations may be safer than imposing a normal model.

Service-Level Interpretation

Cycle service level is the probability of not experiencing a stockout during one replenishment cycle. For a normal lossless model:

where Φ is the standard normal cumulative distribution function. A 95% cycle service level corresponds to approximately z = 1.645, while 99% corresponds to approximately z = 2.326. These values do not mean that 95% of demand will necessarily be filled. That is a different measure, the fill rate, defined as the proportion of demand units supplied immediately from stock.

Worked Example: Reorder Point

Daily demand has mean 100 units and standard deviation 20 units. Lead time is constant at four days. For a 95% cycle service level:

• expected lead-time demand = 100 × 4 = 400 units;
• standard deviation of lead-time demand = 20√4 = 40 units;
• safety stock = 1.645 × 40 = 65.8 ≈ 66 units;
• reorder point = 400 + 66 = 466 units.

The system therefore triggers replenishment when inventory position falls to approximately 466 units. If on-hand inventory is 300, on-order inventory is 250 and backorders are 50, inventory position is 500 units, so no order is triggered yet.

5. Continuous and Periodic Review Systems

A continuous-review system, often called an (Q, R) system, continuously monitors inventory position. When IP ≤ R, an order of approximately Q is placed. It provides rapid response and can use lower safety stock when demand information is timely, but requires reliable transaction recording, system integration and monitoring.

A periodic-review system, often called an (P, S) or order-up-to system, reviews inventory every P time units and orders enough to raise inventory position to target level S. Because no order can be placed between reviews, protection is required over the review period plus lead time:

where μ_P+L and σ_P+L are mean and standard deviation of demand over the review-plus-lead-time interval. The order quantity at review is:

subject to practical minimums, pack sizes and outstanding orders. Periodic review is convenient for route delivery, vendor-managed inventory and coordinated purchasing, but usually exposes the item to more uncertainty and may require greater safety stock.

Min–max systems are a practical variant. The minimum is broadly equivalent to a reorder point, while the maximum is the replenishment target. In all systems, inaccurate stock records can cause both stockouts and excess stock; physical counting and transaction discipline are therefore part of the replenishment design.

6. Working Capital, Obsolescence and Inventory Risk

Inventory is a current asset, but it is not automatically liquid or productive. A useful diagnostic is inventory turnover:

Days inventory outstanding can be approximated by:

These measures must use compatible accounting bases and should be interpreted by product family. High turnover may indicate efficiency, but may also indicate under-stocking and lost sales. Low turnover may be justified for strategic or critical spares, or may signal obsolete inventory.

Risk controls include:

• expiry-date and batch tracking using FEFO (first expired, first out);
• cycle counting focused on value, accuracy risk and criticality;
• approved disposal, markdown and return processes for obsolete stock;
• segregation of quarantined, damaged and saleable inventory;
• demand-sensing reviews for slow-moving items;
• supplier agreements covering returns, postponement or consignment;
• periodic reassessment of service targets and safety-stock parameters.

Inventory valuation and write-down practices must comply with applicable accounting standards and organisational policy. Operational teams must not inflate service performance by hiding shortages in inaccurate records or by shifting inventory between locations without visibility.

7. Multi-Echelon Inventory Management

A supply chain may contain suppliers, factories, regional distribution centres, local warehouses and retail outlets. Managing each location independently can create excessive total stock because every echelon protects itself against the same uncertainty. Multi-echelon inventory management (MEIM) coordinates inventory across these levels.

Two concepts are important. Echelon inventory is stock at a location plus all stock downstream from that location. Installation inventory is stock physically at that specific location. An upstream distribution centre may hold less local stock when downstream demand and replenishment information are visible. Conversely, strategic stock may be positioned upstream when products are common components, while differentiated finished goods are postponed downstream.

A multi-echelon design considers total landed cost, transport time, facility capacity, risk pooling, service commitments and inventory lead time. Risk pooling occurs because aggregated demand is often less variable relative to its mean than separate local demands, although correlation between markets reduces the benefit. Centralisation is not always superior: transport distance, response time, customs exposure and disruption risk may justify decentralised stock.

The bullwhip effect describes amplification of order variability as one moves upstream. Causes include forecast updating, order batching, price promotions, rationing and long lead times. Countermeasures include sharing point-of-sale data, reducing order and transport batch sizes, stabilising prices, shortening lead times and using common replenishment rules. A warehouse should not optimise its own stock metric at the expense of total supply-chain cost and service.

Figure: Physical product flow and information flow in a multi-echelon network.

8. Applied Case: Illustrative Distribution-Centre Policy Redesign

Background. A fictional consumer-electronics distributor supplies 180 retail outlets from one regional distribution centre. Its high-value wireless router is purchased from an overseas supplier. Annual demand is 18,000 units, average daily demand is 60 units, demand standard deviation is 15 units per day, and supplier lead time is normally four days. The firm previously reviewed stock weekly and experienced both emergency shipments and excess stock.

Problem. The item was classified only by unit cost. Managers used on-hand stock rather than inventory position, so open purchase orders were ignored. The weekly review interval also allowed demand to rise substantially before action was taken.

Analysis. The item’s annual consumption value was high, so it was classified as A. Its service consequence was also high because retail customers readily switched to competitors. Management selected a 95% cycle service level. For illustration, assume lead-time standard deviation is one day and demand and lead time are independent. The standard deviation of lead-time demand is:

σ_LD = √(4×15² + 60²×1²) = √3,?

Correcting the arithmetic: 4×225 = 900 and 60² = 3,600, so σ_LD = √4,500 ≈ 67.1 units. Mean lead-time demand is 60×4 = 240 units. Safety stock is 1.645×67.1 ≈ 110 units, giving ROP ≈ 350 units.

The organisation also estimated EOQ using annual demand of 18,000 units, ordering cost of $80 and annual holding cost of $4 per unit:

Q* = √(2×18,000×80÷4) = √720,000 ≈ 849 units.

Approach and solution. The distributor retained periodic review for low-value routine products but adopted continuous review for this A-critical item. The system calculated inventory position from receipts, issues, transfers and open orders. The target service level was reviewed with sales and finance rather than selected automatically. Supplier performance data were used to update lead-time variability monthly. The company also introduced cycle counting, supplier advance-shipment notices and a dashboard showing fill rate, aged stock, forecast error and emergency freight.

Outcome. In the illustrative scenario, record accuracy improved, emergency shipments declined and average stock was reduced after obsolete models were identified. The result did not come from applying EOQ mechanically; it came from segmenting the item, defining the service objective, using the correct inventory position and governing the data.

Learning points. Analytical formulas support judgement but do not replace it. Service level, lead-time reliability, product life cycle, supplier constraints and financial cost must be considered together. A policy should be tested through historical back-testing or simulation before full deployment.

9. Technology, AI and Governance

Warehouse-management systems and enterprise-resource-planning systems can automate inventory transactions, purchase suggestions, lot control and exception alerts. Barcode scanning, RFID, electronic proof of delivery and supplier portals improve visibility. Advanced planning systems can optimise safety stock across locations, while machine-learning models may improve forecasts when sufficient clean data exist.

AI should be treated as decision support rather than an unquestioned authority. Forecast models can reproduce promotional distortions, missing-data patterns or historical bias. Governance should include data ownership, model validation, explainable override rules, security controls, audit trails and monitoring of forecast error and service outcomes. Human approval remains important for critical spares, regulated goods, recalls and unusual demand events.

Contemporary practice increasingly combines postponement, demand sensing, real-time inventory visibility, supplier collaboration, digital twins and sustainability measures. Carbon emissions, packaging, waste and reverse logistics should be included in replenishment decisions where material. More stock is not automatically resilient: resilience also depends on qualified alternative suppliers, flexible capacity, information quality and recovery plans.

10. Implementation Best Practices and Limitations

A practical implementation sequence is:

1. cleanse item masters, units of measure, lead times and supplier data;
2. segment items by value, criticality, variability, shelf life and supply risk;
3. select review policies and service targets by segment;
4. estimate demand and lead-time distributions from relevant history;
5. calculate initial parameters and validate them with operational experts;
6. pilot the policy, back-test stockouts and excess inventory, then refine;
7. monitor service, inventory, cost, accuracy and obsolescence together;
8. establish ownership, approval limits and periodic parameter review.

The principal limitations of textbook models are predictable demand assumptions, normality assumptions, independence assumptions, fixed costs and perfect replenishment. Promotions, new products, intermittent demand, supplier minimums, capacity constraints and correlated disruptions require specialised methods. For intermittent spare-parts demand, Croston-type forecasting or probabilistic models may be more appropriate than a normal-demand safety-stock formula. For highly seasonal demand, parameters should be time-phased rather than based on a single annual average.

The most common implementation failures are poor master data, confusing on-hand with inventory position, setting a high service target without valuing its cost, ignoring stock at other echelons, and measuring each department against conflicting objectives. Effective governance aligns procurement, warehousing, sales, finance and customer service around total cost and agreed service outcomes.

Chapter Summary

Inventory enables supply-chain continuity, customer service and economic operations, but it consumes working capital and creates obsolescence, damage and shortage risk. ABC analysis prioritises managerial attention using annual consumption value, but should be combined with criticality and supply-risk factors. EOQ balances ordering and holding cost under restrictive assumptions and is best used as a benchmark. Reorder-point systems determine when to replenish by combining expected lead-time demand with safety stock. Safety stock depends on the desired service level and the variability of demand and lead time; cycle service level must be distinguished from fill rate.

Continuous review responds promptly to inventory-position changes, whereas periodic review is simpler but generally requires protection over a longer uncertainty interval. Working-capital indicators, ageing and obsolescence controls complement service metrics. Multi-echelon management coordinates stock across the network, captures risk pooling and reduces local optimisation and bullwhip effects. Technology improves visibility and analysis, but data quality, governance and managerial judgement remain essential. Sound replenishment management therefore integrates quantitative models, operational constraints, financial consequences and strategic risk.

Key Terms

ABC analysis: Classification of items according to annual consumption value.

Annual consumption value: Annual unit demand multiplied by unit cost.

Cycle stock: Inventory used between normal replenishments.

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

Echelon inventory: Stock at a location plus downstream stock assigned to that echelon.

Economic order quantity: Order size that minimises relevant ordering and holding costs under specified assumptions.

Fill rate: Proportion of demand units supplied immediately from available stock.

FEFO: First-expired, first-out inventory issue rule.

Holding cost: Annual cost of carrying one unit in inventory.

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

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

Multi-echelon inventory management: Coordinated inventory control across several supply-chain levels.

Obsolescence: Loss of usability or value because inventory is outdated or no longer required.

Periodic review: Inventory review performed at predetermined time intervals.

Reorder point: Inventory-position level that triggers replenishment.

Safety stock: Buffer inventory held against uncertainty.

Service level: Agreed probability or proportion of demand served, depending on the defined measure.

Continuous review: Inventory system that monitors inventory position and triggers replenishment dynamically.

Review Questions

Short Answer Questions

1. What are the four principal inventory cost categories?
2. Define annual consumption value and explain its use in ABC analysis.
3. Distinguish inventory position from on-hand inventory.
4. What is the difference between cycle service level and fill rate?
5. State one advantage and one limitation of periodic review.

Long Answer Questions

1. Explain the assumptions of the basic EOQ model and discuss situations in which those assumptions may fail.
2. Describe how safety stock is calculated when both demand and lead time are variable.
3. Compare continuous-review and periodic-review systems in terms of information requirements, responsiveness and inventory risk.
4. Explain how ABC analysis should be combined with criticality, variability and supply-risk segmentation.
5. Discuss the contribution of multi-echelon inventory management to total supply-chain performance.

Analytical / Application Questions

1. A product has annual demand of 12,000 units, ordering cost of $50 per order and annual holding cost of $2 per unit. Calculate EOQ and explain two operational checks required before adopting the result.
2. Daily demand averages 80 units with a standard deviation of 12 units. Lead time is fixed at five days. Using a 95% cycle service level and z = 1.645, calculate safety stock and reorder point.
3. A retailer has on-hand stock of 420 units, 300 units on order and 60 units backordered. Its reorder point is 700 units. Should a continuous-review system place a replenishment order? Explain the calculation and identify one data-quality risk.

Multiple Choice Questions with Answers

Question 1

Which measure is used most directly to rank items in conventional ABC analysis?

A. Annual consumption value

B. Supplier lead-time variance only

C. Warehouse cubic capacity only

D. Number of customer complaints

Correct Answer: A. Annual consumption value

Question 2

Under the basic EOQ assumptions, which two costs are balanced at the optimum?

A. Purchase cost and shortage cost

B. Ordering cost and holding cost

C. Transport cost and insurance cost

D. Labour cost and tax cost

Correct Answer: B. Ordering cost and holding cost

Question 3

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

A. 200 units

B. 400 units

C. 632 units

D. 1,000 units

Correct Answer: C. 632 units

Question 4

If average daily demand is 50 units and fixed lead time is six days, what is lead-time demand?

A. 56 units

B. 300 units

C. 500 units

D. 3,000 units

Correct Answer: B. 300 units

Question 5

In a continuous-review system, a replenishment order is normally triggered when:

A. On-hand inventory reaches annual demand

B. Inventory position falls to or below the reorder point

C. The warehouse becomes completely empty

D. The supplier changes its price

Correct Answer: B. Inventory position falls to or below the reorder point

Question 6

Daily demand has a standard deviation of 10 units and fixed lead time is four days. Using z = 1.645, what is safety stock under the fixed-lead-time normal-demand model?

A. 16.45 units

B. 20 units

C. 32.90 units

D. 65.80 units

Correct Answer: C. 32.90 units

Question 7

What does a periodic-review order-up-to policy protect against?

A. Only demand during the current day

B. Demand during the review period plus lead time

C. Only supplier purchase price changes

D. Only warehouse damage

Correct Answer: B. Demand during the review period plus lead time

Question 8

Which statement correctly distinguishes cycle service level from fill rate?

A. They are always mathematically identical

B. Cycle service level concerns a stockout during a cycle; fill rate concerns demand units supplied

C. Fill rate measures only supplier reliability

D. Cycle service level measures inventory value in dollars

Correct Answer: B. Cycle service level concerns a stockout during a cycle; fill rate concerns demand units supplied

Question 9

Which situation is most likely to justify combining ABC analysis with a criticality classification?

A. A low-value but safety-critical spare part

B. An item with no demand and no operational use

C. A product whose unit cost is already known

D. A warehouse with barcode scanners

Correct Answer: A. A low-value but safety-critical spare part

Question 10

What is a principal purpose of multi-echelon inventory management?

A. To optimise every location independently

B. To coordinate stock and service across supply-chain levels

C. To eliminate all safety stock immediately

D. To replace demand information with purchase price data

Correct Answer: B. To coordinate stock and service across supply-chain levels