The control system a board is shown is denominated in accrual, the constraint that ends an enterprise is denominated in cash, and no instrument in the delivery discipline joins them.
The monthly portfolio review ends well. Three of the four major programmes are earning close to a dollar of value for each dollar of cost recognised. The fourth has a cost variance the delivery director can explain and a recovery plan with dates on it. Nobody in the room is uncomfortable.
Four days later the group treasurer asks the chief executive to sign a request to extend a working capital facility, because the drawdown profile for the next two quarters no longer fits the existing limit. The request is not linked to any of the four programmes by any document either party holds. It is caused by all of them.
Both accounts are accurate. They are also expressed in different units, produced by different functions on different cycles, and reconciled by nobody. The delivery report is denominated in cost recognised against work performed; the treasury request, in money that has actually moved. Between them sits a gap no instrument in the delivery discipline is built to close, and it is the gap through which solvent enterprises walk into insolvency while their performance reporting stays green.
This is not a lapse of diligence in one enterprise. It is a structural property of the control system almost every large organisation uses. The apparatus producing variance, efficiency indices and completion forecasts was built to answer a question about productive efficiency: are we getting the work we are paying for. That question is worth answering, but it is not the one that determines whether the enterprise survives the next eleven months — and the discipline that built the apparatus knows it.
The Strategic Context
The two ledgers are not a subtlety. The delivery discipline's own foundational teaching draws the distinction crisply: a time-phased budget records cost at the point work is committed and performed, while a cash flow records money at the point it changes hands. The same material names what lives in the gap — supplier terms, invoicing lags, retention held against defects, the cost of carrying working capital — and states plainly that an enterprise can be profitable on paper and still fail because it cannot fund the difference.
Then it makes a choice, and the choice is the whole problem. Having drawn the distinction, it declares that for performance measurement the accrual side is the side that matters, and builds every instrument there. The cost baseline, planned value, earned value, actual cost, and every variance, index and completion forecast derived from them are reported against an accrual reference.
What is built on the cash side is a definition and a warning. No baseline, no variance, no index, no forecast any governance body must receive, no threshold, no owner. The discipline names the constraint that ends companies, classifies it as somebody else's, and returns to the ledger it can measure.
An enterprise adopting this control system inherits that choice without being told a choice was made, and what reaches its boardroom is a complete, internally coherent, professionally produced picture whose completeness is the danger. It does not look partial; it looks like the answer.
What Leaders Commonly Misread
The most common misreading is that a spend profile is a funding requirement. The confusion is embedded in the language the discipline uses about itself: the cumulative cost curve is routinely described as the enterprise's cash curve, existing so the finance chief knows when the money goes out. That collapses the very distinction the same body of teaching sets out elsewhere. A curve showing when cost will be recognised is not a curve showing when the bank balance falls, and on procurement-heavy work the two differ by months in both directions.
The second is that a favourable cost variance means a favourable cash position. It is frequently the opposite: efficient early work accelerates the outflow, so a programme running ahead of baseline draws money sooner than the funding plan assumed while every index on the report improves.
The third does the most damage. The material teaching the cost baseline also instructs practitioners to shape it to payment reality on procurement-heavy packages, placing the supplier deposit in the first period rather than spreading the package evenly. That makes the baseline resemble cash and degrades it as an accrual measure, because earned value is then computed against a curve whose shape reflects payment terms rather than work content. The system is corrupted towards a job it still cannot perform, and the corruption surfaces as variance nobody can attribute.
Whether the accrual baseline is even internally consistent with the other baselines the enterprise holds is a separate question, and not this article's: four baselines produced by four processes with nothing built to detect that they disagree belongs to [Related article: Four Baselines, One Project, No Referee].
Reframing the Issue
Treated as a reporting deficiency this problem is insoluble; treated as a denomination mismatch it is tractable. Adding cash commentary to the delivery pack fixes nothing, and nor does asking finance for a cash forecast built from a different work breakdown, on a different calendar, by people who do not attend the delivery review. What is missing is a conversion instrument.
Consider a specialty chemicals manufacturer building capacity for a new product line. The illustration is hypothetical; the mechanism is general. Long-lead reactor vessels and instrumentation are ordered against staged deposits, so cash leaves at order, at shipment and at site delivery, while cost is recognised as equipment arrives and installation is performed. Raw material for the qualification campaigns is bought ahead and sits in inventory — cash out carrying no project cost line at all. The revenue justifying the undertaking does not begin at mechanical completion but when the customer's own qualification protocol clears, perhaps two or three quarters later. The accrual picture peaks and settles; the cash picture is still deeply negative long after the delivery report has closed the project.
Independent film and television production compresses the same mechanism. Cash leaves during principal photography, cost is recognised across the production, and receipts arrive on delivery acceptance, then across distribution windows and residual streams, with financing bridged against pre-sales and rebates that settle only after acquittal. A slate can read as complete on the delivery ledger while still carrying a facility nothing in that ledger describes.
The Denomination Gap and Where It Is Widest
Every number in the delivery report is a cost, not a payment
Actual cost, in every definition the discipline uses, is cost incurred. Incurred is not paid. An invoice received and approved but unpaid is fully present in actual cost, in the cost variance, in the efficiency index and in the completion forecast, and entirely absent from the bank balance. The delivery report cannot distinguish a supplier paid in advance from one paid ninety days after acceptance, though the funding requirements they produce differ by a full quarter of peak exposure.
The gap widens as the level rises
At project level the accrual system is dense and the cash consequence is bounded. At programme level, aggregation preserves the accrual denomination and adds nothing in cash. At portfolio level, where funding, drawdown and facility headroom are held, the constraint bites hardest and the instrumentation is thinnest. The enterprise has built its most detailed measurement where the risk of ruin is lowest, leaving the level that can end the company reliant on a forecast assembled outside the delivery discipline entirely.
There is a further asymmetry. Cost variance is bounded by the budget at completion; cash exposure is bounded by the facility, and the facility is shared. One programme's early drawdown consumes another's headroom without appearing in either report.
The two pictures diverge most when the enterprise is growing
Growth consumes cash. An enterprise winning more work, starting more units and building more inventory is spending ahead of receipts by construction. Its delivery portfolio therefore looks strongest — more units in flight, more value earned, indices holding — at the point its cash position deteriorates fastest. The control system's optimism and the treasury's alarm are not contradictory readings of one data set. They are correct readings of two, and the enterprise experiences the divergence as a dispute about who is being pessimistic.
Where the headroom comes from — the capital structure that sets it, the cost of holding more of it, the trade-off between funded flexibility and returns — is the province of ERANORTH's Commercial Engine collection, not of this article. The claim here is narrower and prior to it: the delivery control system cannot see the constraint it is spending against.
Decision Framework
The two-ledger reconciliation. Run it at the cadence of the delivery report, for every delivery unit above a stated materiality threshold, and table it in the same meeting. Four parts and one rule.
Part one — the accrual line. Cost incurred to date, value earned to date, forecast cost at completion — reproduced unchanged from the existing report, so the comparison is honest.
Part two — the cash line. Cash out to date, cash in to date, net position to date, and the forecast peak net cash requirement with the date it occurs. The date is the point of the exercise: a peak without a date cannot be tested against anything.
Part three — the bridge. The named items explaining the difference between the two lines, each with a value and an owner: deposits and prepayments; work performed but not invoiced; invoiced but not collected; retention held; long-lead material bought ahead of use; duty, tax and freight timing; rebates awaiting acquittal; financing drawn against the unit. If the bridge does not reconcile to the difference, report it as failed rather than adjust it.
Part four — two tests. The headroom test: does the forecast peak, on its date, fit inside available facility with a margin the board stated in advance, and what is that margin now. The divergence test: is the interval between accrual completion and cash completion widening, and is any decision now before the enterprise one that improves the accrual line while worsening the cash line.
The rule. Any change approved on accrual grounds that moves cash beyond a stated threshold requires separate approval from the cash line's owner. Acceleration, early procurement, prepayment for a discount and scope brought forward all pass the accrual test easily, and are exactly the decisions that consume headroom.
| Accrual ledger | Cash ledger | |
|---|---|---|
| Records | Cost when work is performed | Money when it moves |
| Bounded by | Budget at completion | Facility headroom, shared |
| Instrumented in the delivery discipline | Completely | Not at all |
From Strategy to Execution
Immediate. Name the owner of the cash line for each material delivery unit and require the bridge once before the next review, however rough — one that does not reconcile is more informative than none. Then identify every decision approved in the last two quarters on accrual grounds that moved cash materially, and establish whether anyone assessed that effect.
Medium term. Put the forecast peak requirement and its date into the standard delivery pack beside the completion forecast, and set the headroom margin the board expects maintained. Route changes above the threshold past the cash owner before commitment. Make payment terms a visible attribute of every material supply arrangement in the portfolio view: terms are the largest single lever on peak exposure, and are currently negotiated where nobody governing the portfolio can see them.
Long term. Treat the divergence between accrual and cash completion as a portfolio design parameter rather than an accounting outcome, and decide deliberately how much of the portfolio may consist of units whose receipts depend on acceptance rather than progress. That is a question about the shape of the business, and belongs to the people who choose the portfolio.
Re-baselining will be proposed as a remedy somewhere in this work, and it is not one. What it does to the record of commitments already broken is treated in [Related article: Re-Baselining Erases the Record of Every Promise You Broke], a separate problem from this one.
Signals to Monitor
The interval between the accrual peak and the cash peak, tracked as a trend. Growth in work performed but not invoiced, the purest form of the gap. The ageing of retention balances, which are cash the enterprise has earned and cannot use. Supplier deposits rising as a share of committed spend, converting negotiated discount into funding exposure. Financing drawn increasing while cost variance stays favourable — the most diagnostic pairing available, because it can only mean the ledgers have separated. And a rising share of portfolio revenue contingent on acceptance rather than progress, which lengthens the tail on every unit at once.
Whether anything in the enterprise is authorised to test that its plans are complete before it commits against them is a related but distinct exposure, and is examined in [Related article: Nothing Is Authorised to Test Whether the Plan Is Complete].
Questions for the Leadership Team
- On what date does our portfolio's peak net cash requirement fall, what is the figure, and which units drive it?
- What is the average interval between cost recognition and cash settlement across our material supply arrangements, and how has it moved?
- Which decisions approved in the last two quarters improved a cost variance and worsened our cash position, and who assessed the second effect?
- How much of our earned value to date sits in amounts invoiced but uncollected, and what is the ageing profile?
- If our three largest units all ran two months ahead of baseline at once, what would that do to headroom, and has anyone modelled it?
- Who is accountable, by name, for the bridge between the delivery report and the treasury forecast, and when did they last reconcile them?
Closing Perspective
The choice before a board is not whether to keep the accrual control system. It is a good system, comparable across units, and the language contracts are written in. The choice is whether to go on believing it describes the enterprise's condition.
It describes one of two conditions. The other ends companies, and it is not measured, not forecast, not thresholded and not owned inside the discipline that produces the report. That absence is not an oversight to delegate downwards. It is a governance position the enterprise holds by default, and the responsibility for holding it deliberately, or ending it, sits with whoever reads the report and mistakes it for the whole picture.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.