Project Financial Management: The Cost Case That Has to Survive Review

Project financial management is usually taught as budgeting with a longer horizon. On publicly funded projects in Ireland it is something more specific: the production of a financial case that has to survive three formal approval gates, an external assurance process if the project is large enough, and a post-completion review that asks whether the money bought what the business case promised.

That framing matters because the common failure is not arithmetic. Submissions are rarely sent back because the sums are wrong. They are sent back because a cost element was left out, because a single number was presented where a range was required, or because a decision was visibly taken before the analysis that was supposed to inform it.

The lifecycle the numbers have to move through

The Infrastructure Guidelines, published in December 2023, replaced the Public Spending Code for capital projects. The change was structural rather than cosmetic. The lifecycle condensed from six stages to five, and the Public Spending Code’s four decision gates, DG0 to DG3, became three approval gates.

• AG1 — Strategic Assessment and Preliminary Business Case. Government consent is required for major projects.
• AG2 — the pre-tender deliverables: Detailed Business Case, Project Execution Plan and Procurement Strategy. For major projects, Ministerial approval is required before tendering.
• AG3 — Final Business Case, submitted after tender. Government consent is required again for major projects.

Stages 4 and 5, implementation monitoring and then Post Completion Review and Benefits Realisation, sit in the lifecycle but are not approval gates. Nothing is held up waiting for them, which is exactly why they are the stages most often skipped.

The thresholds decide how much of this applies. A Strategic Assessment is required above a €20 million capital expenditure threshold, raised from €10 million. Below €20 million the AG2 deliverables are not required and only AG1 and AG3 apply. The threshold for formal External Assurance Process and Major Projects Advisory Group review was raised from €100 million to €200 million, applying at the Preliminary Business Case stage.

Two cases in one submission, and VAT is the tell

The economic appraisal and the financial appraisal answer different questions, and the quickest way to see whether a team understands the difference is to look for VAT.

The economic appraisal excludes VAT, because from the point of view of the State it is a transfer rather than a cost. The financial appraisal must include it, because the sponsoring agency actually pays it. The guidance is unambiguous that VAT should always be included in the financial appraisal, and the recurring mistake is teams dropping it from both cases on the strength of the economic rule. The result is a financial case that understates what the organisation will have to fund.

Which appraisal method is required also depends on size: cost-benefit analysis should be undertaken for projects with an estimated cost greater than €30 million. Our guide to building a project business case under the Irish public spending framework covers the appraisal documents in sequence, and the discounting mechanics behind them are set out in our guide to NPV, IRR and payback in capital budgeting.

What a total cost actually contains

The single most common defect in submissions is reporting capital cost as though it were total cost. A complete cost table carries capital cost, VAT, contingency, risk, inflation and operating costs, and the guidance asks explicitly that all cost elements appear in one table rather than being scattered through the narrative.

Operating costs are the quietest omission. Schemes concentrate on capital spend and overlook operating costs beyond the base case, which understates the project in both the economic and the financial appraisal at once.

Contingency and inflation carry a second obligation. Where the rates applied do not come from the published sources a sponsoring agency is expected to use, it has to explain why it chose different values. An unexplained contingency percentage is not treated as a judgement call; it is treated as a gap.

Note the direction of all of this. Every omission on that list, without exception, makes the project look cheaper than it is. That is not usually deliberate, but it is systematic, and it is why reviewers check completeness before they check accuracy.

Ranges, not points

Project appraisal groups recommend cost ranges rather than single-point estimates, particularly at the early stages, with the range narrowing as the design develops. A single figure at AG1 is a claim to precision the project does not yet have.

This is harder to present than it sounds, because a range invites the question of which end of it will be approved. The answer is that the range is the honest output of an early-stage estimate, and narrowing it is work, not drafting. Where flexibility itself has value, that value can be quantified rather than described, which is the subject of our guide to real options in investment appraisal.

Where the benefit-cost ratio goes wrong

Two errors recur, and both inflate the ratio.

The first is the demand year. Some projects calculate the benefit-cost ratio using demand from the end of the appraisal period rather than the opening year. Demand grows over the appraisal period, so using the later figure overstates the ratio. The opening-year demand is the one to apply.

The second is the shadow price of public funds, which should be applied to the net public financial cost of a project rather than split between non-labour and labour costs. It is a mechanical error with a material effect on the result.

Alongside these sits a methodological drift: reliance on multi-criteria analysis and cost-effectiveness analysis where cost-benefit analysis is required. The guidance is blunt that the multi-criteria approach is highly subjective, and a poorly specified or selectively weighted matrix can undermine an otherwise sound scheme.

The sequencing failure

The most damaging finding in the review literature is not a number at all. Reports sometimes show that key decisions about the design of a project or programme had already been made before the optioneering phase. Once that is visible on the page, the appraisal reads as justification rather than analysis, and no amount of rigour in the later sections repairs it.

A related weakness starts earlier still. Objectives set at Preliminary Business Case stage are often high-level and lack specification, which then weakens every KPI and option derived from them. Objectives that are not specific do not produce options that can be compared.

Cost control after the money is approved

Approval moves the problem rather than solving it. From AG3 onward the question is whether spend to date can be reconciled against work actually completed, which is the function of earned value management, and whether the cash profile holds, which is ordinary cash flow management under unusually long timescales. Where the project is delivered under long-term contracts, the revenue and cost recognition pattern is governed by IFRS 15 and its over-time recognition and contract asset rules, which frequently diverge from the approved cash profile and need to be reconciled deliberately rather than discovered at year end. The forecasting discipline underneath all of it is covered in our guide to budgeting and forecasting.

Stage 5, and why it is usually the weakest section

Two requirements at the end of the lifecycle are routinely under-built.

The first is a confirmation step that is easy to miss because it is new: the assumptions made at AG1, including the economic appraisal assumptions, should be reviewed and confirmed at Detailed Business Case stage. A business case that carries forward an eighteen-month-old assumption set without revisiting it has skipped a required step.

The second is the ex-post evaluation plan. These are commonly weak in a specific way: KPIs that do not clearly link back to the stated objectives, and data sources and baseline timing left unclear. A logic path model is the recommended remedy, and the reason baseline timing matters so much is that the baseline usually has to be captured before implementation begins. We set out why in our guide to benefits realisation and why most projects cannot prove they worked.

The underlying disciplines

Project financial management draws on the same core disciplines as the rest of the finance function, applied over a longer horizon and against an approval process. The planning cycle itself is covered in our guide to budgeting and forecasting, the liquidity side in cash flow management, and the general management responsibility in financial management for managers. Where the cost table has to be defensible to a reviewer, the system it is produced from matters too — see our note on accounting software with an audit trail.

Our suggestions

Three things follow for anyone carrying the financial case on a project.

Build the cost table before the narrative, with all six elements in it, and treat any blank cell as an open question rather than a zero. Second, present early-stage costs as ranges and say what work will narrow them, because a single number at AG1 will be read as either precision you do not have or a figure you will later have to defend. Third, write the evaluation plan at business case stage and collect the baseline before delivery starts, because that is the one part of the financial case that cannot be reconstructed afterwards.

The discipline is less about forecasting accurately than about being complete and honest about what is not yet known. Reviewers are considerably more tolerant of a wide range than of a confident number with something missing from underneath it.