A property group running on a standard accounting setup is usually running blind in a very specific way. The ledger is accurate. The bank reconciles. The annual accounts are signed. And nobody in the business can answer, on any given Tuesday, what the current project is now forecast to cost and what that does to the equity return. That gap is not a bookkeeping problem. It is a design problem, and it has to be solved before the first drawdown rather than after the first cost overrun.
The unit of account is the project, not the company
Generic accounting systems are built to report a trading entity: revenue this month, expenses this month, a profit figure at the bottom. Property does not work that way. A development consumes cash for two or three years and produces revenue in a compressed burst at the end. A commercial building produces steady income but its value moves independently of that income. A hotel does both at once. In each case the meaningful unit of account is the project or the asset over its life, not the company over a calendar month.
Which means a monthly profit and loss for a development entity is close to meaningless. It shows capitalised costs sitting on the balance sheet and almost nothing in revenue, and it tells a director nothing about whether the project is on track. The information a board actually needs is a cost report against feasibility, a cash forecast against the facility, and a forecast final return — none of which a standard chart of accounts produces.
The practical fix is dimensional. Every transaction carries a project code and a cost category that maps to the feasibility line items, so the ledger can be pivoted by project and by trade package rather than only by account. Most cloud accounting systems support this through tracking categories, but they support it badly beyond a handful of projects and they do not carry commitments at all. At some point — usually the third or fourth concurrent project — the group needs a dedicated development management system alongside the ledger, with a defined reconciliation between the two. The decision to make deliberately is which system is the source of truth for cost, because if both claim to be, neither is.
Feasibility to actuals
Every project starts with a feasibility. It sets the land cost, the construction budget by package, the professional fees, the finance costs, the contingency, the revenue assumptions and the resulting margin and return on cost. It is the document the equity was raised on and the facility was approved against. It is, in most groups, the last time anyone looks at it as a whole.
What typically happens is that the feasibility gets superseded by a live cost report, and the live cost report becomes the only reference point. Each month it is compared to last month. Movements are small, they are explained, and everyone is satisfied. By month four the cumulative drift from the original feasibility is material and nobody has measured it, because the comparison being made is against the previous version of the forecast rather than against the case the project was approved on.
The discipline is to freeze the approved feasibility as a locked baseline and report against it for the life of the project, alongside the current forecast. Where a re-forecast genuinely supersedes the baseline — a scope change the board has approved, a redesign — that becomes a formally re-based feasibility with a version number and an approval date, and the original stays visible underneath it. A director should be able to see, in one view, the approved case, the current forecast, and every re-basing in between with the reason recorded. Groups that do this catch problems in month three. Groups that do not catch them at practical completion.
Development cost control
The core instrument is the cost-to-complete report, and its value comes from separating three things that are routinely conflated.
| Column | What it means | Why it matters separately |
|---|---|---|
| Budget | The approved feasibility line, plus approved variations to it | The benchmark. Movements here need a decision, not an entry |
| Committed | Contracted or ordered, whether or not work has been done | Cash that is already spent in substance. It is the number that constrains you, and it never appears in the ledger |
| Incurred | Work certified or goods received, whether or not invoiced | The accounting position. Drives WIP and accruals |
| Paid | Cash out the door | Reconciles to the bank and to the facility drawn |
| Cost to complete | Forecast remaining spend to finish the line | The judgement. Budget less incurred is an arithmetic answer, not a forecast |
| Forecast final cost | Incurred plus cost to complete | The only number that answers "what will this project actually cost" |
The distinction that causes the most damage is committed versus incurred. A group looking only at incurred cost sees comfortable headroom against budget while the trade packages have already been let at prices that consume it. Commitment is the leading number; incurred cost is the trailing one. If the cost report does not carry a commitment column sourced from executed contracts and purchase orders, it is not a control document.
Variations need the same rigour. Each one carries a status — notified, assessed, approved, rejected — and the report should show approved variations inside the budget column and unapproved but likely ones disclosed separately, so the board can see the exposure that has not yet been decided. A variations log that only records approved items understates risk by design.
Track the contingency drawdown curve against the construction completion curve from the first month. Contingency is priced to cover the unknown across the whole build, and the unknown is weighted towards the back end. If contingency is being consumed faster than the build is progressing, the project is in trouble and the cost report has not caught up yet. This single comparison is the most reliable early warning in development finance, and it costs nothing to produce.
Capitalisation and revenue recognition
Development costs, including borrowing costs and holding costs, are generally capitalised into the carrying value of the project while it is being brought to its intended condition. The rules on borrowing cost capitalisation are specific: costs are capitalised on qualifying assets while expenditure is being incurred, borrowing costs are being incurred and the activities necessary to prepare the asset are in progress. Two consequences are regularly missed. Capitalisation must be suspended during extended periods in which active development is paused — a consenting delay, a funding stall, a dispute that stops work — and it must cease when the asset is substantially complete, not when the last invoice arrives or the last unit settles. Interest on unsold completed stock is a period expense, and for a project with a slow settlement run-off that is a meaningful hit to the reported result at exactly the point the project looks finished.
Holding costs on land held for future development deserve the same scrutiny. Rates, insurance and interest on a site that is not yet in active development are not being incurred to bring an asset to its condition, and treating them as capital because it flatters the interim result is a position that does not survive review.
Revenue recognition on development sales is the question most likely to be answered wrong, because it is usually answered by intention rather than by contract. Under IFRS 15 the test is whether the performance obligation is satisfied over time or at a point in time, and the over-time test turns on things like whether the asset has an alternative use to the developer and whether there is an enforceable right to payment for performance completed to date. Those are legal characteristics of the sale and purchase agreement and the relevant law, not commercial preferences. Two developments on the same street, with different contract terms, can properly reach different answers. The consequence is large: over-time recognition spreads revenue and margin through the build, point-in-time recognition puts all of it at settlement. Get the assessment done in writing, by contract template, before the first presale is signed — and redo it whenever the template changes. Retrofitting the analysis after a year of presales is expensive and it is the kind of thing an auditor will challenge.
Funding and drawdown mechanics
A typical development capital stack runs senior debt, sometimes mezzanine, and equity, with each layer priced for its position and each carrying conditions. The senior facility will usually require a level of qualifying presales, will fund construction progressively against quantity surveyor certification, and will require the equity to go in first or alongside on a defined ratio. Mezzanine sits behind senior, costs materially more, and brings its own conditions and intercreditor arrangements that constrain what the senior lender and the borrower can agree between themselves.
The mechanical point that causes the most avoidable stress is timing. A drawdown is not available on request. It follows a claim from the contractor, certification by the QS, submission to the lender and processing within the facility's notice period — a cycle that takes weeks, not days. Meanwhile the contractor's payment terms run on their own clock, and the two are not synchronised. The result is a working capital gap that recurs every single month of the build and is entirely predictable.
Which is why the drawdown schedule must sit inside the cash forecast rather than beside it. A forecast that shows the facility as available capacity, and construction cost as an outflow, will always look solvent. A forecast that models each claim, its certification date, the notice period, the funding date and the payment date against contractor terms will show the gaps — and the gaps are the thing you need to know about, because they are covered with equity, retentions timing or a negotiated facility change, all of which need lead time. This is a weekly-resolution problem, not a monthly one.
Covenant testing in a property context
Property facilities are covenanted differently from trading facilities, and the ratios have property-specific definitions that carry more weight than the ratio names suggest.
| Test | Broadly measures | Where the definition bites |
|---|---|---|
| Loan to value (LVR) | Debt against asset value | Which valuation — as-is, as-if-complete, on a vacant possession or investment basis; which valuer; how often; who commissions it |
| Loan to cost (LCR) | Debt against total project cost | Whether "cost" includes capitalised interest, land at cost or at value, and the developer's own margin |
| Interest cover (ICR) | Income against interest expense | Whether income is passing rent, contracted rent or effective rent; whether interest is actual, hedged or a notional stressed rate |
| Debt yield | Net income against debt, valuation-independent | The definition of net income and what is deducted for vacancy, capex and non-recoverables |
| Presale coverage | Qualifying presales against debt | What makes a presale "qualifying" — deposit level, unconditional status, purchaser type, related-party exclusions, single-purchaser caps |
The covenant definitions in the executed facility document are the only ones that matter. Not the term sheet, not the credit paper, not the ratio as it is generally understood in the market. Build the covenant model directly from the defined terms, line by line, with a clause reference against each input, and test it monthly against forecast as well as actual. A covenant breached on a forecast basis is a conversation with the lender. A covenant breached without warning is a default, and in a property facility default rights typically include repricing, suspension of further drawdowns and enforcement over the security — mid-build, with a contractor on site.
Commercial leasing and investment property
The first thing to be clear about in a commercial portfolio is that the rent on the lease is not the rent you are earning. Face rent is the headline figure. Effective rent is what remains after incentives — rent-free periods, fit-out contributions, cash inducements — are spread across the term. Under lessor accounting the incentive is recognised over the lease term, which means reported rental income is smoothed while the cash is lumpy and front-loaded against the landlord. A portfolio reporting on face rent overstates its income, and a valuation conversation conducted on face rent misleads everybody in the room. Report both, and hold the incentive amortisation schedule per tenancy.
The rent roll itself should be treated as a forecasting instrument rather than an administrative schedule. It holds every date that drives future income: rent review dates and mechanisms, expiry dates, rights of renewal and their exercise windows, incentive run-off, outgoings recovery structure and the split between recoverable and non-recoverable costs. From that, the group should be producing a forward income profile, an expiry profile by year, and a weighted average lease term. WALT is the summary metric a lender or a purchaser will ask for first, and it needs a stated basis — weighted by income or by area, to expiry or to first break — because the two versions differ and are frequently quoted interchangeably.
Measurement of investment property is a policy choice between fair value and cost, and it is not a neutral one. Fair value brings the asset's real economics onto the balance sheet and is what most investors and lenders expect to see; it also injects valuation movements straight into reported profit, so earnings swing with the property cycle rather than with operating performance. That volatility has to be managed in board and investor reporting by separating operating result from revaluation movement every period, without exception, so that a good year is not mistaken for a good year of trading. Set the valuation cycle deliberately — independent valuations at a defined frequency, directors' valuations in between on a defined basis, with the valuer's instructions and assumptions on file — because the cycle is exactly what an auditor and a lender will test.
Hotels: an operating business wearing a property asset
The most common analytical mistake in hotel ownership is to treat the hotel as a building that generates income. It is a trading business with high fixed costs, daily pricing decisions and a labour-intensive operation, which happens to occupy an asset on your balance sheet. The finance function has to serve both layers, and the operating layer moves far faster than the property layer.
Operating performance is driven by occupancy, average daily rate and the product of the two, revenue per available room. These are not supplementary statistics — they are the actual drivers, and they need to be reported daily or weekly with forward booking pace alongside them, because pace is what allows a rate decision to be made while it still matters. Departmental reporting on the uniform industry basis, with each department's revenue and direct cost shown before undistributed expenses and ownership costs, is what makes performance comparable to the market and to prior periods.
The structural question is how the hotel is operated, and the two main answers produce very different economics and very different accounting. Under a lease, the owner receives rent and the operator takes the trading risk and reward. The owner's numbers look like property: lessor accounting, rental income, covenants tested on rent, and the operator's credit quality as the principal risk. Under a management agreement, the owner retains the trading result and the operator manages the asset for base and incentive fees. The entire hotel profit and loss consolidates into the owner — every operating cost, every employee, every seasonal swing — and the finance function has to be resourced accordingly. A franchise or brand licence sits over either arrangement and brings brand fees and mandated standards, including capital spend obligations that fall on the owner regardless of how the year has gone.
Under a management agreement the owner should read the fee mechanics carefully: how the incentive fee is calculated, what it is subordinated to, what the owner's priority return is, what the termination rights and performance tests are, and what the operator can commit the owner to without consent. These are the terms that determine whether the owner's return survives a soft year.
Two further items. Furniture, fittings and equipment reserves are usually contractually required — a percentage of revenue set aside for asset replacement — and they need to be modelled as a genuine cash claim and reconciled, not treated as an accounting provision that quietly funds working capital. And hotels are seasonal, often severely so, with a cost base that does not flex nearly as fast as revenue. A thirteen-week rolling cash forecast is not a good practice item for a hotel owner; it is the instrument that gets the business through the low season with the covenant intact.
Tax, structure and sequence
Three areas of New Zealand tax catch property groups repeatedly, and all three are technical enough that the right move is to describe the mechanism and then get current advice.
The first is the land taxing provisions. New Zealand taxes gains on land in a range of circumstances well beyond the bright-line test that gets the public attention: land acquired with a purpose or intention of disposal, disposals by people in the business of dealing in, developing or building on land, and land subject to development or division work. The bright-line rule is one provision among several, its period has changed more than once, and it is not the provision that most often catches a developer. The critical point is that these tests apply at acquisition based on purpose and association, so the analysis needs to be done and documented when the land is bought, not when it is sold.
The second is GST on land. Supplies of land between GST-registered persons, where the purchaser acquires it with the intention of using it to make taxable supplies and it is not intended as a principal place of residence, are compulsorily zero-rated. Compulsorily — it is not an election. The mechanism relies on warranties given in the sale and purchase agreement about registration status and intended use, and if those warranties are wrong the GST consequences land somewhere unwelcome, usually with interest. Change of use, apportionment where a development has mixed residential and commercial components, and the timing of input tax claims across a long build are all live issues on almost every project.
The third area is the associated persons rules, which can attribute the activities and status of an associate to an entity that has none of its own. A passive investment entity associated with a builder or a land dealer may be taxed on a disposal it assumed was on capital account. Associations run through shareholding, trusts, relatives and control relationships, and they are broader than people expect. Because these tests apply at acquisition, the structure has to be right before the contract is signed. There is no clean way to restructure out of a purpose or an association after the fact.
Structurally, most groups run a special purpose vehicle per project, which is the right default. It ring-fences risk, gives lenders clean security over a defined asset, and lets an investor take a position in one project without exposure to the others. The cost is administrative: multiple entities, multiple sets of statutory accounts, a consolidation, and intercompany funding flowing between them. That intercompany funding is where groups get sloppy. Loans between project entities need to be documented as loans, with terms, interest and repayment mechanics, and reconciled monthly on both sides. Undocumented intercompany balances create tax exposure, distort project returns by putting one project's finance cost in another project's accounts, and become a diligence problem the moment the group raises capital or sells an asset.
The sequencing below is what we would want in place at each gate.
| Gate | In place before it | Because |
|---|---|---|
| Before land settlement | Entity and ownership structure confirmed, land tax and associated persons position advised in writing, GST treatment and SPA warranties reviewed | These tests apply at acquisition and cannot be fixed later |
| Before first drawdown | Covenant model built from the executed facility definitions, drawdown schedule inside the cash forecast, cost report with budget, committed, incurred and paid columns live, feasibility baseline locked | The controls have to exist before the money moves, not after the first claim |
| Before first presale | IFRS 15 over-time versus point-in-time assessment documented against the contract template, deposit handling and stakeholder arrangements confirmed | The answer follows the contract terms, and the template is signed hundreds of times |
| Before first valuation | Measurement policy decided, valuer instructed with documented assumptions, valuation cycle and directors' valuation basis set | Fair value drives reported earnings and covenant headroom; the basis must be defensible |
| Before practical completion | Capitalisation cessation date determined, settlement run-off and unsold stock holding costs forecast, defects and retentions provisioned | Interest stops being capital at completion, and the run-off is longer than anyone plans for |
| Ongoing | Monthly cost report against locked feasibility, monthly covenant test on actual and forecast, weekly cash to thirteen weeks, quarterly rent roll and WALT refresh | Property problems are slow to appear and fast to become terminal |
None of this is exotic. It is ordinary financial discipline applied to a unit of account that most accounting systems were not built to report. The groups that handle a downturn well are not the ones with better instincts. They are the ones who could see the contingency curve bending in month three and had time to act on it.
XLCFOWe run finance for property groups — project cost reporting against feasibility, drawdown and covenant models, multi-entity consolidation, rent roll forecasting and hotel departmental reporting — alongside your QS, valuer and tax adviser.
