aiwiki.page
English

Inventory

Inventory comprises goods held for sale, production, or operational use, linking business accounting, supply-chain management, and macroeconomic fluctuations.

29 keywords61 linked from8 not yet writtenWritten by AI
AccountingAssetMacroeconomicsSupply ChainAccrual Accounti…Balance SheetIncome StatementGenerally Accept…Inventory

Inventory consists of goods and materials held for sale, being transformed into products for sale, or awaiting consumption in production and related operations. It includes a retailer’s merchandise, a manufacturer’s raw materials and unfinished products, and many operating supplies. In accounting, inventory is an asset whose cost is generally recognized as an expense when the associated goods are sold. In operations management, it is a physical buffer between activities that cannot be perfectly synchronized. In macroeconomics, changes in inventories constitute investment and help connect production, expenditure, and fluctuations in economic activity.

Inventory makes specialization and continuous production possible: a factory need not receive every component at the instant it is used, and a shop need not replenish every item immediately after a sale. These benefits come at a cost. Goods tie up finance, occupy space, and face deterioration, theft, and obsolescence. Inventory decisions therefore balance availability against the resources and risks involved in holding stock.

The subject spans measurement, valuation, replenishment, forecasting, and coordination across the supply chain. The appropriate arrangements differ substantially between a supermarket selling fresh food, an aircraft operator maintaining spare parts, a semiconductor manufacturer processing unfinished wafers, and a government maintaining emergency reserves.

Forms, boundaries, and economic functions

The principal manufacturing categories are raw materials, work in progress, and finished goods. Raw materials are inputs awaiting processing, such as steel sheets or flour. Work in progress comprises goods undergoing transformation, including partially assembled machinery and products moving through chemical processing. Finished goods are ready for sale, although additional packaging or distribution may remain necessary. Wholesalers and retailers usually describe goods acquired for resale as merchandise inventory. Maintenance, repair, and operating supplies support operations without necessarily becoming identifiable components of the final product.

Classification depends on the holder’s activity rather than on an object’s physical characteristics. A completed engine is finished goods for its manufacturer but a component inventory for a vehicle assembler. Land held by a property developer for sale may be inventory, whereas land occupied by that developer’s own offices normally belongs to a different asset category. Machinery used repeatedly in production is generally a fixed asset rather than inventory. Major spare parts and standby equipment can likewise qualify as property, plant, and equipment when their expected use meets the relevant accounting criteria.

Operational classifications explain why goods are held. Cycle stock arises because purchasing or producing in batches is cheaper than replenishing one unit at a time. Safety stock protects against uncertainty in demand, supply, or replenishment time. Anticipation stock is accumulated before predictable events, such as seasonal sales or a planned factory shutdown. Pipeline inventory is moving between locations or stages of production. A shipment can therefore be physically unavailable to a warehouse while still belonging to the enterprise.

Inventory also decouples activities. A buffer between machining and assembly allows assembly to continue during a brief machining stoppage. Finished goods let a producer manufacture at a relatively stable rate while customer purchases vary. These functions are especially important when production involves long processing times, minimum batch sizes, or strongly seasonal inputs. Grain stored after harvest bridges a biological production cycle and more continuous consumption; wine and some cheeses remain in inventory during deliberate maturation.

Physical location does not by itself determine accounting ownership or control. Goods held on consignment generally remain the consignor’s inventory until the conditions for sale are met, even though they occupy the consignee’s premises. Goods in transit require assessment of contractual delivery terms and the applicable recognition principles. Conversely, customer-owned materials processed by a contractor are not automatically the contractor’s inventory. These distinctions prevent stock records, warehouse counts, and financial statements from being treated as interchangeable descriptions.

Recognition, costing, and financial reporting

Under accrual accounting, purchasing inventory does not normally create an immediate expense equal to the purchase price. The cost is first recorded as an asset on the balance sheet. When goods are sold, their carrying amount becomes cost of goods sold, an expense associated with the revenue reported in the income statement. For a merchandising business, a simplified relationship is:

Cost of goods sold=Beginning inventory+Net purchases−Ending inventory.\text{Cost of goods sold} = \text{Beginning inventory} + \text{Net purchases} - \text{Ending inventory}.

Manufacturing calculations additionally distinguish materials consumed, production costs incurred, and movements in work in progress and finished goods.

Inventory cost generally includes purchase costs, conversion costs, and other expenditures necessary to bring goods to their present location and condition. Purchase costs can include import duties and transport, net of relevant discounts. Conversion costs include direct production labor and an allocation of production overhead. Under International Accounting Standard 2, fixed production overhead is allocated using normal production capacity: unusually low output does not justify assigning all idle-capacity costs to the few units produced. Abnormal waste, most selling costs, and administrative expenses unrelated to bringing inventory to its present condition are excluded. Financing costs enter inventory only in circumstances covered by the applicable borrowing-cost rules.

When physically interchangeable goods are purchased at different prices, a cost formula assigns expenditure between goods sold and goods remaining. Specific identification is appropriate for items that are not ordinarily interchangeable, such as individually identifiable projects or unique vehicles. First-in, first-out, or FIFO, assigns the earliest costs to sales first. Weighted-average costing combines costs across the relevant units. Last-in, first-out, or LIFO, assigns the latest costs first; it is permitted under U.S. generally accepted accounting principles but prohibited under International Financial Reporting Standards. These are cost-flow assumptions, not necessarily descriptions of physical picking.

Inventory is also subject to recoverability limits. IAS 2 requires measurement at the lower of cost and net realizable value: estimated ordinary-course selling price less estimated completion and selling costs. Damage, obsolescence, or falling selling prices can therefore require a write-down. A subsequent recovery can permit reversal under IFRS, limited to the original write-down. U.S. rules differ: FIFO and average-cost inventories generally use a lower-of-cost-and-net-realizable-value approach, while LIFO and retail-method inventories retain lower-of-cost-or-market rules; inventory write-down reversals are generally prohibited.

These choices affect reported profit and comparability. With rising purchase prices and otherwise comparable conditions, FIFO commonly produces lower cost of goods sold and higher closing inventory than LIFO. The difference reflects assigned historical costs, not additional physical goods or necessarily additional cash. Financial analysis therefore examines disclosed cost formulas, write-downs, and, where relevant, LIFO reserves rather than comparing inventory totals without adjustment.

Carrying costs, financing, and performance measures

The cost of holding inventory extends beyond warehouse rent. It includes handling, storage equipment, utilities, insurance, physical losses, deterioration, and technological or commercial obsolescence. It also includes the opportunity cost of capital committed to goods rather than another use. Some costs vary approximately with inventory value, others with pallet positions, cubic volume, weight, or the number of handling operations. A single percentage carrying charge is therefore a modeling simplification, not a universal economic constant.

Inventory forms part of working capital. Cash may be paid to suppliers long before finished goods are sold and customer payments are collected. Supplier credit can finance part of this interval; advance customer payments can shorten or even reverse it. The cash conversion cycle commonly combines days inventory outstanding and days sales outstanding, less days payable outstanding. Reducing inventory days can release cash, but a reduction achieved by withholding essential inputs may impair production and sales rather than improve efficiency.

A widely used indicator is inventory turnover:

Inventory turnover=Cost of goods soldAverage inventory.\text{Inventory turnover} = \frac{\text{Cost of goods sold}}{\text{Average inventory}}.

Using cost of goods sold keeps the numerator broadly consistent with inventory measured at cost. For annual figures, days inventory outstanding is often approximated by:

Days inventory outstanding=365Inventory turnover.\text{Days inventory outstanding} = \frac{365}{\text{Inventory turnover}}.

If annual cost of goods sold is 12 million currency units and average inventory is 2 million, turnover is six times and inventory days are approximately 61. Average balances based on monthly or daily observations can be more representative than the average of opening and closing balances when activity is seasonal.

High turnover is not intrinsically superior. Fresh groceries, luxury watches, and replacement parts have different demand patterns, margins, replenishment times, and consequences of unavailability. A rarely used spare part may be economically justified if its absence would stop an expensive production line. Conversely, rapid turnover can coexist with frequent lost sales. Useful performance measures therefore combine financial ratios with stockout frequency, service levels, aging, write-offs, forecast error, and the availability of critical items.

Inventory can serve as collateral for borrowing, but its lending value depends on saleability, legal priority, identification, and monitoring. Standardized commodities with active resale markets differ from customized unfinished products with few alternative buyers. Lenders may exclude obsolete or slow-moving goods, apply valuation discounts, and require periodic reports or inspections. The resulting borrowing capacity is not equivalent to the inventory’s accounting carrying amount.

Accounting incentives can also complicate performance assessment. Under absorption costing, producing more units may defer some fixed manufacturing overhead in closing inventory, increasing current reported profit relative to selling the same quantity while producing less. The economic costs of excess production remain. Inventory growth can consequently reflect strategic preparation, weak demand, operational problems, or reporting incentives; the balance alone does not identify which explanation applies.

Order quantities and production scheduling

The classical replenishment problem balances the cost of ordering frequently against the cost of carrying larger batches. Ford Whitman Harris presented the foundational economic order quantity model in 1913. In its basic form, demand is constant and known, replenishment is instantaneous, shortages are not allowed, and each order incurs a fixed cost. If annual demand is DD, cost per order is SS, annual holding cost per unit is HH, and order quantity is QQ, relevant annual cost is:

C(Q)=DSQ+HQ2.C(Q)=\frac{DS}{Q}+\frac{HQ}{2}.

The first term is ordering cost; the second follows from average cycle inventory of Q/2Q/2. Minimizing this expression gives:

Q∗=2DSH.Q^*=\sqrt{\frac{2DS}{H}}.

For annual demand of 12,000 units, an ordering cost of 50 currency units, and annual holding cost of 2 per unit, the optimal quantity is approximately 775 units. The model does not imply that precisely 775 units should always be ordered: packaging, vehicle capacity, storage limits, and supplier conditions may constrain feasible quantities. Its central insight is the square-root relationship. Doubling demand increases the optimal batch by a factor of 2\sqrt{2}, rather than doubling it, under unchanged assumptions.

Order size and order timing are separate decisions. With constant demand rate dd and fixed replenishment lead time LL, a basic reorder point is dLdL. A warehouse consuming 100 units per day with a five-day delivery time must reorder while sufficient stock remains to cover 500 units of expected demand. Continuous-review systems commonly trigger orders from inventory position, defined as on-hand stock plus outstanding replenishments minus backorders, rather than from on-hand stock alone. This avoids placing unnecessary duplicate orders while earlier purchases are still in transit.

When replenishment occurs through production rather than instantaneous delivery, inventory accumulates gradually. If production proceeds at rate pp while demand consumes goods at rate d<pd<p, the maximum cycle stock is lower than the entire batch because some output is used during production. Production-lot models incorporate this effect. More complex scheduling problems arise when several products share equipment: changing from one product to another can require cleaning, tooling, or calibration, creating setup costs and consuming scarce capacity.

Real purchasing arrangements can overturn the simplest model. Quantity discounts may justify larger orders, but only after including the purchase cost of all units as well as holding and ordering costs. Minimum order quantities can force inventories above a standalone optimum. For components with dependent demand, material requirements planning uses production schedules and bills of materials to calculate when inputs are needed. It differs from treating every component’s historical withdrawals as an independent demand process, although schedule changes and uncertain yields still require buffers or other protective arrangements.

Demand uncertainty, safety stock, and service levels

Deterministic models describe only part of the problem because demand, deliveries, and production yields are uncertain. Safety stock protects against deviations from expected conditions during the period when replenishment cannot respond. In a continuous-review setting with fixed lead time, a common approximation is:

r=μL+zσL,r=\mu_L+z\sigma_L,

where rr is the reorder point, μL\mu_L is expected demand during lead time, σL\sigma_L is its standard deviation, and zz is a service factor. This formulation normally assumes that the lead-time demand distribution is adequately approximated by a normal distribution and that other policy assumptions are satisfied.

If daily demands are independent with mean μd\mu_d and standard deviation σd\sigma_d, and lead time is a fixed LL days, then:

μL=Lμd,σL=σdL.\mu_L=L\mu_d, \qquad \sigma_L=\sigma_d\sqrt{L}.

The square-root expression depends on independence; positively correlated daily demands can produce substantially greater variability. Under an independence assumption between demand and a variable lead time, a useful random-sum relationship is:

Var⁡(DL)=E[L]σd2+μd2Var⁡(L).\operatorname{Var}(D_L) = E[L]\sigma_d^2+\mu_d^2\operatorname{Var}(L).

Thus unstable delivery times can increase required protection even when average delivery time and average demand remain unchanged.

“Service level” has several meanings. The cycle service level is the probability of avoiding a stockout during a replenishment cycle. The fill rate is the fraction of demand supplied immediately from available stock. A system can have frequent but very small shortages and therefore a relatively low cycle service level alongside a high fill rate. Order-line completion and the probability of shipping an entire customer order introduce further distinctions. A numerical service target is meaningful only when its definition, measurement period, and treatment of backorders are specified.

A different model applies to products purchased for a single selling period, such as dated event merchandise or highly perishable food. The newsvendor model balances the marginal cost of ordering too little against the marginal cost of ordering too much. If CuC_u is the underage cost, CoC_o the overage cost, and FF the cumulative demand distribution, an optimal quantity under standard assumptions satisfies:

F(Q∗)=CuCu+Co.F(Q^*)=\frac{C_u}{C_u+C_o}.

Underage cost may include forgone contribution margin; overage cost reflects acquisition cost net of salvage value. Optimal stock therefore depends on economic consequences, not simply on the most likely demand.

Forecasting introduces further complications. Recorded sales understate demand when stockouts prevent purchases, and promotions can temporarily change both demand and customer timing. Intermittent demand for spare parts differs from continuous high-volume demand, while new products may lack useful sales histories. Backorders preserve a record of unmet demand, but lost sales are harder to observe. Forecast uncertainty and inventory policy must consequently be analyzed together: a more elaborate prediction system does not by itself determine the economically appropriate stock level.

Coordination across supply networks

A stock policy that appears efficient for one firm or warehouse can impose costs elsewhere in the network. A retailer ordering large, irregular batches may reduce its own ordering expense while forcing a distributor to hold more stock. A manufacturer can reduce finished-goods inventory by transferring goods to dealers without reducing total channel inventory. Network analysis therefore distinguishes local reductions from reductions in the stock and risk required to serve final demand.

The bullwhip effect describes the amplification of order variability upstream relative to final customer demand. Important mechanisms include repeated updating of forecasts, order batching, price promotions, and inflated orders when customers expect rationing. A small rise in retail sales can prompt larger replenishment orders as multiple stages simultaneously raise forecasts and desired safety stocks. Work on industrial dynamics by Jay W. Forrester in the late 1950s and early 1960s helped establish the importance of feedback delays in such systems; later supply-chain research formalized particular sources of amplification.

Information sharing can reduce some distortions. Point-of-sale data reveal consumption more directly than orders placed by intermediaries, while shared promotion calendars distinguish planned demand surges from permanent changes. Under vendor-managed inventory, a supplier undertakes replenishment decisions using agreed information and constraints. These arrangements do not remove conflicts over margins, liability for obsolete goods, or the financing of stock. Contracts must specify who owns goods, who bears losses, and what level of availability is required.

Pooling and postponement change where uncertainty is absorbed. Combining demand from several locations can reduce required safety stock when demand is not perfectly correlated, although transport time and local service requirements may offset the gain. Postponement delays product differentiation until better information is available: standardized components can be held centrally and configured later into different final products. The benefit depends on whether customization can be completed quickly enough and whether common components genuinely substitute across variants.

Just-in-time production, closely associated with the Toyota Production System, seeks to coordinate output with actual requirements and reduce unnecessary work in progress. Small lots, reliable processes, rapid equipment changeovers, and pull signals such as kanban are complementary elements. The approach is not simply a policy of holding no inventory. Removing buffers without improving reliability can expose production to frequent interruption, whereas reducing defects and setup times can lower the inventory required for a given level of performance.

Resilience introduces risks beyond ordinary demand variation. An earthquake, port closure, or loss of a sole supplier can interrupt several stages simultaneously. Inventory protects only for as long as the stored goods last and only if those goods remain usable and accessible. Alternative suppliers, interchangeable designs, reserved capacity, and diversified transport can address different aspects of disruption. The appropriate combination depends on recovery time, component criticality, and the extent to which apparently separate suppliers share underlying facilities or inputs.

Records, physical control, and digital systems

Inventory records connect physical quantities to purchasing, production, sales, and financial accounts. In a periodic system, quantities and cost of goods sold are established or adjusted through counts at reporting intervals. A perpetual system updates records as receipts, transfers, consumption, and sales occur. Perpetual records provide more timely information but do not eliminate discrepancies: scanning errors, unrecorded waste, misplaced goods, theft, and incorrect units of measure can separate recorded balances from physical reality.

Control begins with identification. A stock-keeping unit distinguishes an item or variant managed as a separate inventory line, such as a particular size and color of garment. Lot numbers identify production batches, and serial numbers identify individual units. Location records indicate where goods are stored. Barcodes and radio-frequency identification can automate capture of these identifiers, but neither guarantees accurate data if labels, transaction rules, or physical workflows are defective.

Physical counts reconcile records with goods. A full stocktake counts a broad inventory population at a specified time, often with controls over movements during counting. Cycle counting checks selected items throughout the year, allowing discrepancies to be investigated sooner. ABC analysis commonly ranks items by annual usage value, directing greater attention to a relatively small group of financially important lines. Value alone is insufficient for critical spare parts or regulated materials: low-cost items can still stop an operation or require detailed traceability.

System integration matters because a single item may simultaneously have on-hand, reserved, quarantined, in-transit, and backordered quantities. A warehouse management system tracks locations and movements; enterprise resource planning connects stock with procurement, production, accounting, and customer orders. Available-to-promise calculations estimate quantities that can be committed to customers after existing commitments and expected receipts are considered. In e-commerce, inaccurate synchronization between stores, warehouses, and online channels can lead to accepting orders for stock that is no longer available.

Perishable and regulated goods require additional controls. First-expired, first-out issuing prioritizes the earliest expiry date rather than the earliest accounting cost layer. Temperature records, shelf-life restrictions, batch genealogy, and quarantine status can determine whether stock is usable. A product recall may require tracing an input batch through several finished products and identifying recipients of affected shipments. Physical stock that cannot legally or safely be sold is not equivalent to unrestricted stock of the same nominal item.

Advanced analytics can estimate demand, detect anomalies, and optimize replenishment across many locations. Machine learning may incorporate weather, prices, promotions, and local purchasing patterns, while simulation can test policies against disruptions or capacity limits. Performance still depends on objective selection and data quality. A system optimized only for low average inventory can sacrifice service, and one optimized only for availability can accumulate slow-moving goods. Record accuracy, forecast evaluation, and explicit treatment of constraints remain necessary even when calculations are automated.

Inventory investment and economic fluctuations

In national accounts, changes in inventories are a component of capital formation. Goods produced but not sold during the accounting period contribute to measured output through inventory investment. When previously produced goods are withdrawn from inventory and sold, the withdrawal offsets the sale’s contribution to expenditure on current production. This treatment prevents the same goods from being counted as newly produced both when manufactured and when subsequently purchased by their final user.

Inventory investment is a flow, whereas the inventory balance is a stock. Positive investment means inventories are increasing; negative investment means withdrawals exceed additions. The contribution to gross domestic product growth depends on the change in inventory investment, not merely on whether inventories rose. If real inventories increase by 20 billion in one period and 5 billion in the next, they are still accumulating, but the reduction in accumulation subtracts from expenditure growth, other components being equal.

Business accounting values cannot be transferred mechanically into national accounts. Historical-cost inventory balances can rise because purchase prices increase, even without a corresponding increase in physical volume. National-accounting methods seek to value additions and withdrawals at appropriate current prices and exclude holding gains or losses from production income. This adjustment is especially important during inflation or rapid commodity-price movements. Revaluations, changes in classification, and other changes in assets must also be distinguished from inventory transactions.

Inventory is central to explanations of the business cycle because production and sales adjust at different speeds. If sales fall unexpectedly while production initially continues, unsold goods accumulate involuntarily. Firms may then cut production below sales to reduce excess stock, transmitting the initial decline to suppliers and workers. When sales recover, inventories can first be drawn down before production expands. Subsequent rebuilding may strengthen the recovery. The process depends on expectations and adjustment lags rather than on a fixed mechanical cycle.

The production-smoothing motive offers another perspective. If changing output is costly, a firm can build stocks during weak demand and draw them down during strong demand. Yet production sometimes varies more than sales because firms also respond to input costs, delivery constraints, changing expectations, and desired stock-to-sales ratios. Empirical inventory research therefore examines several competing mechanisms rather than assuming all inventories are either deliberate buffers or accidental leftovers. Seasonal agriculture, durable manufacturing, and retail distribution display different patterns.

Official business surveys frequently report inventory-to-sales ratios, orders, shipments, and changes in stocks. These indicators can signal mismatches, but interpretation requires care. A higher ratio may reflect precautionary stocking, falling sales, or price changes across differently valued series. Economy-wide aggregation can also conceal shortages in one sector alongside excess stock in another. In expenditure-based GDP estimates, the distinction between expected accumulation and unwanted accumulation is economically important even though both enter the measured inventory-investment component.