Accrual conversion gave three years. It did not give anyone a register.

Public sector accrual conversion looks, from the outside, like an accounting reform. Cash in, cash out becomes assets, liabilities and depreciation. Most of it is genuinely an accounting exercise, and accountants handle it.

One part is not. Before an entity can report the value of its infrastructure, somebody has to establish what infrastructure exists. That is not accounting. It is fieldwork, and it is usually the largest and least understood part of the programme.

What the standard actually offers

IPSAS 33 governs first-time adoption of accrual basis IPSAS. It grants a transitional relief of three years during which an entity need not recognise or measure certain classes of asset, including property, plant and equipment. The stated purpose is to allow time to develop reliable models for recognition and measurement, and it exists because reliable historical cost information frequently does not.

Two things follow from that, and they are easy to conflate.

The relief is not permission to leave assets out. It is time to find them. And three years is measured from adoption, not from the point at which anyone starts looking.

An entity that spent the relief building a register emerges with a number it can defend. An entity that spent the relief waiting emerges with the same absence it started with, and a reporting obligation that has now matured.

A valuation and a register are different purchases

This is where the procurement usually goes wrong, and it goes wrong for an understandable reason.

The obligation is expressed as a financial statement line. So the work gets scoped as a valuation exercise, procured against a deliverable, and priced accordingly. The output is a dataset that supports a number at a cut-off date.

An asset register is not that. ISO 55000 defines asset management as the coordinated activity of an organisation to realise value from assets, and ISO 55001 sets out the requirements for a management system to do it. A register is an output of such a system. It is not a substitute for one, and it cannot be bought as a finished object, because the thing it describes does not hold still.

Four ways the scope goes wrong

None of the following requires anyone to have behaved badly. Each follows from procuring a function as though it were a deliverable.

Scope written around available information rather than criticality. A specification that says "review the assets recorded in the entity's systems" produces a survey of what was already known. For an organisation established decades ago, with successive reorganisations and inconsistent record-keeping, the assets that are missing from the records are precisely the ones the exercise needed to find.

Budgets set without reference to portfolio size. Where the fee is set by procurement convention rather than by the extent of the estate, coverage is thin by arithmetic. A bidder pricing to win a competitive tender on an unscaled budget cannot fund the fieldwork that the specification implies, and everybody involved knows it at bid stage.

Stakeholder requirements untethered from criticality. Without an agreed view of which assets matter, requests expand to fill the scope. Somebody wants coordinates for every sign and lighting column. Somebody else wants condition on street furniture. These are not unreasonable questions in isolation. They are unanswerable inside a single engagement, and asking them signals that no one has yet decided what the register is for.

Methodology set by whoever is closest to the sponsor. Where the specification does not fix a method, competing advisers apply different ones, and the output stops being comparable across entities. That matters most at the consolidating level, where figures from many entities are meant to add up.

The register starts decaying on acceptance

Assets change. Roads are widened, networks are extended, master plans are revised, whole districts are re-planned. A register produced as a one-off deliverable is accurate on the day it is accepted and progressively less accurate every day after.

The organisations that feel this first are the ones whose estates are growing fastest, which in this region is most of them.

There is a second decay, and it is quieter. A programme of this kind produces more than a dataset. It produces an understanding of where the information was, which functions held it, and what was wrong with it. That understanding usually lives in a handful of people. When the sponsoring official moves to another portfolio, it leaves with them, and the next programme rediscovers the same deficiencies at full cost.

Say what the number is worth

There is an established answer to the discomfort of publishing a figure built on records of uneven quality, and it is not to avoid publishing.

Mature infrastructure practice grades the confidence of asset information alongside the information itself. The International Infrastructure Management Manual sets out confidence grading for this purpose, and Audit New Zealand applies the same logic in its guidance to auditors, looking at whether data is sufficiently accurate, sufficiently complete, current, and adequate in spatial coverage.

An unqualified asset figure is not necessarily wrong. It is incomplete, because it does not disclose how much of it rests on verified field data and how much on desktop inference. A figure published with a confidence grade is more useful and more honest than one published without, and it gives the next programme a baseline to improve on rather than a number to defend.

What the international frameworks say to buy instead

The idea that an organisation should build a capability rather than commission a dataset is not a contrarian position. It is the mainstream international one, and it is specified in some detail.

The GFMAM Asset Management Landscape, now in its third edition, sets out a consolidated international view of the discipline across seven groups and forty subjects. It exists so that capability can be assessed and compared rather than assumed. The Institute of Asset Management's Asset Management Anatomy maps those subjects onto capability models with maturity levels, where the level described as competent corresponds broadly to an effective management system conforming to ISO 55001.

The IAM's Competences Framework makes the resourcing implication explicit. Asset management spans engineering, finance, operations, information systems, contract and supplier management, and organisational development. No single discipline covers it, which is why placing it inside the finance function, or inside engineering alone, tends to produce a partial result.

What that looks like in practice

Building the function rather than buying the dataset changes the shape of the procurement rather than only its size.

• Establish where asset management sits in the organisation, with named accountability, before specifying any survey. The location question is not administrative. It determines who can compel other functions to hand over what they hold.
• Define criticality first. What the register is for decides what goes in it. Without that decision, scope is set by whoever asks loudest.
• Scale the budget to the estate, not to procurement precedent. An unscaled fee produces thin coverage and a second programme later.
• Include an operating period. A capability is demonstrated by running it. An adviser who builds a register and leaves has not transferred anything; one who operates it for a defined period while training the permanent team has.
• Specify the maturity level being bought, using a recognised model, so that the outcome is testable and comparable across entities.
• Grade and publish data confidence, so that the register improves against a known baseline rather than being rebuilt from zero.

Done that way, the register outlives the accounting obligation that prompted it. The same information base supports maintenance planning, renewal forecasting, performance-based contracting and concession structures, none of which are available to an organisation that can produce a valuation but not a current view of what it owns.

The deadline is the opportunity

In Saudi Arabia the conversion to accrual accounting under IPSAS was mandated for government institutions by Royal Decree No. 13059 in December 2016, which obliges entities to recognise, tag and value their assets. Comparable transitions are under way across the region.

The transitional relief in IPSAS 33 was designed for exactly this situation, where the records do not support the standard and time is needed to build something that does. It is a well-judged provision.

It is also finite, and it ends whether or not the register exists. The entities that will find the deadline straightforward are those that treated it as a mandate to build an asset management function. The ones that treated it as an accounting deferral will meet it again, at the next set of financial statements, with less time and the same absent information.