Fixed asset registers that maintain themselves

Ask to see a fixed asset register and you will usually get a spreadsheet. It will be arithmetically correct, carefully maintained, and describe a set of assets that does not quite match what the business owns.

The arithmetic was never the problem. Depreciation is a formula. The problem is that the register drifts from reality in three specific ways, and only one of them is a calculation issue.

The three ways registers go wrong

Assets that were expensed instead of capitalised. Somebody coded a RM 6,000 purchase to repairs. It never reached the register, and it never will, because nothing prompts a review of what was coded to expense.

Assets that no longer exist. Scrapped, stolen, replaced, or simply gone. The register carries them at written-down value and depreciates them faithfully for years. Nobody tells finance when equipment is thrown out.

Assets that exist and were never recorded. Bought outside the normal purchasing process, acquired with a business, or transferred between sites.

All three produce a register that reconciles internally, agrees to the ledger, and misstates the position. That is what makes it a durable problem: every internal check passes.

What automation genuinely fixes

Depreciation. Runs every period, on every asset, correctly. Removes the missed month, the wrong rate, the asset that was never added to the schedule.

Disposals. Gain or loss on disposal calculated correctly, with the asset properly removed and the accumulated depreciation cleared. Manual disposal accounting is a common source of error.

Additions from invoices. This is the useful part. When an invoice is coded to a capital account, the system can create the register entry from the invoice — description, supplier, date, cost, document reference attached — rather than someone retyping it into a spreadsheet weeks later.

Reconciliation to the ledger. Continuous rather than annual, so a divergence is found when it happens.

Flagging capitalisation candidates. Purchases above the capitalisation threshold coded to expense accounts, surfaced for review. This directly attacks the first failure mode, and it is the single most valuable check in the list.

What it does not fix

Whether the asset still exists. No system knows the machine was scrapped. That requires someone to look, and the only reliable mechanism is a physical verification.

Whether it should have been capitalised. Capital versus revenue is judgement against your policy, informed by intent that usually is not in the document. Automation can flag candidates by amount; it cannot make the call. It is the most common automated coding error for exactly this reason — see AI coding of general ledger transactions.

Useful life and residual value. Estimates requiring a view about how long you will use something and what it will be worth.

Impairment. Whether an asset is worth less than its carrying value depends on facts about the business and the market.

The capture point is the whole game

Registers drift because assets enter and leave the business through processes that do not talk to finance.

The fix is upstream, not in the register:

• Capitalisation decision at the point of purchase, not at year end. If a purchase order above the threshold requires a capital classification before approval, the asset is identified before the invoice arrives.
• A disposal process that someone actually uses. Usually the weakest link. If throwing out a broken machine requires a form nobody knows exists, the register will carry it forever. The process has to be lighter than the alternative of ignoring it.
• A tag or reference linking the physical item to the register entry, so verification is possible without guesswork.

Automation applied to good capture produces an accurate register with no effort. Applied to poor capture it produces a beautifully calculated, precisely depreciated, incorrect register — faster than before.

The verification nobody schedules

At least annually, someone should physically verify a sample of the register, and separately walk the premises looking for assets that are not on it.

The second direction is the one that gets skipped and finds the most. Checking that register items exist is standard; checking that existing items are on the register is not, and it is where unrecorded assets and unrecorded disposals both surface.

The check worth running now

Take the last twelve months of purchases coded to repairs, maintenance and consumables, filter for anything above your capitalisation threshold, and look at what comes back.

In most businesses that have not run it, something belongs on the register. It is a fifteen-minute exercise and it is the fastest way to find out whether this is a problem for you.

Common questions

Can a fixed asset register maintain itself?

The arithmetic can — depreciation, disposals, gain or loss calculations and reconciliation to the ledger all run without intervention, and register entries can be created directly from capital invoices rather than retyped. What cannot be automated is knowing whether an asset still physically exists, whether a purchase should have been capitalised, and estimates of useful life and impairment.

Why is my fixed asset register wrong?

Almost always for one of three reasons that have nothing to do with calculation: assets that were coded to expense and never reached the register, assets that have been scrapped or replaced without anyone telling finance, and assets acquired outside the normal purchasing process. All three produce a register that reconciles internally and agrees to the ledger while misstating what the business owns.

How do I find assets that were expensed by mistake?

Filter the last twelve months of purchases coded to repairs, maintenance and consumables for anything above your capitalisation threshold, and review what comes back. In most businesses that have not run this check something belongs on the register, and a system can surface these candidates automatically as they occur rather than leaving it as an annual exercise.

What should a physical verification cover?

Both directions. Checking that items on the register still exist is standard practice, but walking the premises looking for assets that are not on the register is what finds unrecorded additions and unrecorded disposals, and it is the direction most commonly skipped.

Related: keeping track of your equipment · AI coding of general ledger transactions · year-end adjustments and the audit file