Your Fixed Asset Register Is a Spreadsheet Nobody Has Verified in Years

Anything your capitalisation policy says is an asset rather than an expense, recorded at its full capitalised cost including delivery, installation, testing and any non-recoverable duty. That means the machine, the plant it sits on, vehicles, buildings, fit-out, IT hardware above the threshold, and intangibles such as purchased software licences where the policy allows. Consumables, spares held for repair and anything below the threshold stay out and get expensed. The register is not a list of everything the company owns. It is the list of everything carried on the balance sheet.

Straight line and reducing balance cover most requirements, and any serious module handles both. Usage-based depreciation, where the charge follows machine hours or units produced, matters if you run heavy plant and want cost to track output. Beyond the method itself, check the detail: part-period conventions in the month of acquisition, whether it depreciates from the in-service date or the invoice date, residual value handling, and what happens when you change the remaining useful life part way through. Those settings cause more reconciliation trouble than the method choice does.

It has to, if your tax rules allow different methods or lives from your statutory accounts, which is common. The same asset then carries two valuations: one posting to the general ledger, one feeding the tax computation, with different methods, lives and accumulated balances. This is a genuine dividing line between capable and basic fixed asset modules. Some handle it natively, some bolt it on with a second asset record that then drifts, and some cannot do it at all. Test it with your own numbers during evaluation rather than accepting a yes in a demo.

As a dated transfer transaction against the asset record, not as an edit to a location field. The transfer carries an effective date, the source and destination location or cost centre, the custodian handing over and the one receiving, and an approval where the destination sits in a different part of the business. From that date the depreciation charge moves to the new cost centre, so the department carrying the asset carries its cost. The old location stays in the asset history, which is what you need when a physical verification finds the machine somewhere unexpected.

Tag each asset with a barcode or RFID label carrying the register number, then scan against a verification list, either everything once a year or a rolling cycle that covers high-value locations more often. The system produces a variance report: found where expected, found somewhere else, not found, and found but not in the register. Each variance is then cleared with a transfer, a write-off or a late capitalisation, approved and posted. Assets on the books that no longer physically exist are both an audit finding and an overstatement of the balance sheet, so the write-offs are the point of the exercise, not an embarrassment.

You raise a disposal against the asset with a date and a sale value, which may be zero for a scrapping. The system depreciates up to the disposal date, removes the cost and the accumulated depreciation from the balance sheet, compares the proceeds against the remaining net book value, and posts the gain or loss to the profit and loss account. Partial disposal of a componentised asset works the same way against the component. The important part is that it is one posting, so the register and the ledger cannot disagree afterwards.