Monthly operations reviews tend to repeat a particular scene: the warehouse report appears on screen, weeks of cover have fallen on several critical items, the procurement lead presents a replenishment proposal, and the proposal is approved after a short discussion. No one in the room disputes the figures, because the figures are correct — the warehouse does hold precisely that quantity. What the same page does not contain is a line showing what has been ordered but not yet received; that information sits in the purchasing module, under a different breakdown and a different owner, and enters the agenda only once something has already gone wrong. The decision is taken on the most visible data available, and in that narrow sense it is entirely rational.
The same pattern reappears in capital projects at materially larger amounts. When material status is discussed in a site progress meeting, the conversation centres on what has physically landed and stands ready for installation; long-lead equipment still on the fabricator's floor, a shipment covered by an opened letter of credit waiting at the load port, a package held in customs clearance — none of these contribute to the reported progress percentage, and none appear in the body of the report. What the site observes is an empty lay-down area, and the demand the site generates follows that observation in the form of additional ordering. The distance between committed value and visible physical reality then accumulates in the project's cash curve without ever becoming a heading that anyone owns.
This accumulation has a name — pipeline inventory, stock whose title has passed to the buyer but which has not yet reached a usable position — and its magnitude is not a matter of preference. In-transit volume is, to a close approximation, the product of consumption rate and lead time; in an operation whose weekly consumption is unchanged, a lead time extending from eight weeks to fourteen therefore enlarges the pipeline position by roughly seventy percent without a single new decision being taken. No one has resolved to carry more stock, the ordering policy has not changed, no budget approval has been sought; an external parameter has moved and the balance sheet has quietly adjusted to it. Growth in in-transit inventory consequently arrives as the output of arithmetic rather than the output of management intent.
Pipeline inventory is not in itself a defect; it is the price of distance, lot sizing and supply uncertainty. Shipping by sea rather than by air reflects cost discipline; enlarging an order to reach a price break is the exercise of purchasing leverage; carrying additional cover on a critical component is reasonable insurance against single-source exposure. The difficulty lies not in the shortcut but in the shortcut persisting after the conditions that justified it have shifted: to the extent that the reorder trigger remains tied to on-hand quantity alone while lead times lengthen, the system will predictably issue replenishment for goods already moving. That duplication is the most common institutional source of the amplification of small demand variations as they travel upstream — the bullwhip effect.
Visibility is typically lost where accounting and operations look at different moments. Most ERP configurations open the inventory record at goods receipt, whereas title, governed by the applicable Incoterm, often passes considerably earlier — on an FOB shipment, when the goods are loaded at the named port of shipment. Through the intervening four to eight weeks the material is legally and for accounting purposes the company's, while in the operational system it belongs to no one; insurance obligation, currency exposure and financing cost all continue to run through that window. A comparable gap recurs in work in process held at a toll processor, in consignment stock standing at a customer, in a lot prepaid and segregated in a supplier's warehouse, and in goods moving through a returns process — each of these is economically inventory, and none of them appears in the warehouse report.
The first institutional cost is read in working capital. Where days inventory outstanding is derived from warehouse data, the cash conversion cycle systematically understates the number of days actually carried, and the business structurally underestimates its own liquidity requirement. On the financing side the picture inverts: most borrowing base arrangements admit in-transit inventory as eligible collateral only where a bill of lading, an insurance certificate and customs documentation support it, and exclude it altogether where that chain is absent. The result is an asset funded with cash that produces no collateral value; and as it grows, revolver utilisation rises at an unchanged revenue level, interest expense follows, and demurrage, detention and bonded storage charges disperse across the income statement under headings that obscure their common origin.
The second cost appears at the valuation table. Although price in a company sale is usually discussed through an EBITDA multiple, the second mechanism determining the amount actually paid at closing is the net working capital peg, typically set against a trailing twelve-month average. Where the pipeline position was inflated during the reference period, the peg is fixed high and the seller is left funding it; where the inflation occurs after the reference period, the difference is deducted from consideration at closing. Beyond that, reconciliation between the physical count date and the dates on which title passed is a standard procedure in quality of earnings work, and a discrepancy surfacing there tends to be read not as an isolated adjustment but as a signal about inventory controls generally — a signal usually priced through broader representations and warranties, a higher escrow percentage, or an added condition precedent.
The third cost is charged directly to the project. Where committed but not yet incurred value is not maintained in a separate record, the estimate at completion is structurally low, and when an unforeseen item arises the contingency assumed to remain available proves to have been consumed some time earlier. Material arriving on site ahead of need produces a further cost layer of its own: lay-down capacity tightens, second handling becomes necessary, damage rates rise, and once the storage duration permitted under the construction all-risks policy is exceeded, a coverage discussion begins. In a project setting, pipeline inventory is therefore not merely a financing item but a schedule and liability item as well.
The mechanism that neutralises this tendency is definitional discipline rather than individual attention, and it separates into four components. The first is a single inventory position: quantities on hand, in transit, segregated at the supplier, held at a toll processor and standing on consignment are consolidated into one table under one owner. The second is tying the reorder trigger to that consolidated position rather than to warehouse quantity; a technically modest change, it removes the greater part of duplicated replenishment. The third is converting lead time from a master-data constant into a measured variable whose variance is tracked — the most frequently observed structural defect being a lead-time parameter entered at implementation and never revised thereafter. The fourth is a monthly reconciliation of title transfer dates against goods receipt dates, with in-transit balances monitored through an ageing schedule.
The intervention BEIREK operates in capital projects and in multi-site industrial groups builds these components around a commitment register. For each procurement package three dates are maintained separately — order date, date on which title passes, date of acceptance on site — and cost reporting never collapses the committed, incurred and installed amounts corresponding to those three dates into a single figure. The monthly rhythm is constructed on the reconciliation of the three columns, so that a widening gap between committed and incurred value becomes an agenda item on its own terms, without waiting for a delay report to surface it.
The second layer is the separation of accountability, since the same figure carries a different meaning for each party reading it. For procurement, pipeline inventory is a supplier delivery performance indicator; for finance, it is cash immobilised outside the collateral base; for project controls, it is a precondition for the accuracy of the estimate at completion; for the sponsor and the lender, it is the question of whether the physical progress underlying a drawdown request is real. Producing those four readings from one record moves the discussion away from who is right and toward which parameter has shifted. Revision of the lead-time parameter is accordingly handled not as a master-data correction but as a management decision whose rationale is recorded and whose working capital effect is calculated.
Pipeline inventory reflects less a company's inventory policy than its decision about where inventory is counted; and an item that is not counted is never managed, only financed. In assessing the supply discipline of a business, the question that carries information is not how many weeks of stock are carried, but at which moment the counting of those weeks is deemed to begin.