A fixed asset register can be wrong in two opposite directions. It can list equipment that's gone, which an existence test is built to catch, or miss equipment that's working on the floor, which a completeness test is built to catch. They're two different tests because checking what's on the books can't show what's missing from them.
Walk a plant floor with the fixed asset register in hand and two kinds of surprise turn up. Some lines point to equipment nobody can find: a generator sold two years ago, a forklift scrapped after an accident. Some equipment on the floor has no line at all: a compressor bought on a maintenance budget, a machine transferred in from another site and never added.
Both are register errors, and they pull the numbers in opposite directions. The first overstates what the company owns. The second understates it. Auditors test for them separately, and a process built to catch one can miss the other entirely. When a federal audit of routine equipment purchases checked items against the records, 43 percent of the items tested, worth $1.8 million, could not be verified as existing. This piece is for COOs, CEOs and operations leaders who run the equipment, and for the finance and audit teams who have to stand behind the register.
The existence test checks whether the assets recorded on the books exist. The completeness test checks whether everything that should be on the books is recorded. One catches overstatement, the other catches understatement.
Auditing standards (for US public companies, the PCAOB's AS 1105) list these as separate assertions, which is audit language for the claims a company's financial statements make. Two plain-language terms go with them.
The standards' own wording makes the point directly. Auditing standards separate the existence assertion, which tests whether recorded assets actually exist, from the completeness assertion, which tests whether everything that should be recorded is recorded. They are two different tests because sampling what is on the books cannot tell you what is missing from it.
The direction of the check decides what it can find. Start from the register and walk to the floor, and you'll find the ghosts. Start from the floor and walk to the register, and you'll find the zombies. A count that only runs one way only finds one kind of error.
A single transfer can create both errors at once. A machine moved from one plant to another can stay on the first plant's register as a ghost and turn up at the second plant as a zombie. Each site's register is wrong, in opposite directions, and the company-wide total can still look about right.
Registers fail both tests for the same underlying reason: equipment changes on the floor, and the register only changes when someone tells it. Five causes drive most of the gap:
|
|
Existence test |
Completeness test |
|
The question |
Does everything on the books exist? |
Is everything that exists on the books? |
|
Where the check starts |
The register, then the floor |
The floor, then the register |
|
The error it finds |
Ghost assets, which overstate the books |
Zombie assets, which understate the books |
|
A typical case |
A generator sold years ago, still recorded |
A compressor bought on a maintenance budget, never recorded |
|
What it can't see |
Equipment missing from the register |
Recorded equipment that's gone |
Neither column is the right one. A register that's checked in only one direction can look clean and still be wrong in the other.
Ghosts and zombies distort different numbers, and both stay invisible until someone checks in the right direction:
The 43 percent figure above comes from one specific audit, GAO-08-333, a 2008 review of government purchase card buys. GAO itself calls its selection nonrepresentative, so it shows what checking finds, not a rate for any other organization.
None of this puts a dollar figure on a given company's exposure. The size depends on its own equipment and records, and the only way to know is to check in both directions. The more useful question for an executive isn't how big the gap is somewhere else. It's which direction your own register has been checked in, and when.
"Evidence beats debate. Every time."
Tim Harris, CEO, SoloTruth
The exposure is highest where equipment moves often and the people who move it are far from the people who record it:
A verification process that holds up checks the register in both directions and records the evidence it finds. When evaluating approaches, look for six capabilities:
A register that passes both tests is simple to describe, and it depends on routine more than on any one system:
The ways ghosts get onto a register are covered in our post on six ways ghost assets get onto the register, and the completeness side in our post on zombie assets.
Four assumptions let a register look clean while it fails one of the two tests:
Reality: a sample drawn from the register can only confirm what's listed. It says nothing about equipment that was never recorded.
Reality: an invoice proves a purchase happened and a work order proves a repair happened. Neither shows the machine is still on site today.
Reality: ghosts at one site and zombies at another can offset each other in the total while every site's register is wrong. The total is the last place the errors show.
Reality: there's no general requirement to physically check fixed assets the way inventory is observed, so an annual audit shouldn't be assumed to look at every machine.
It checks whether the assets recorded in the register exist. It starts from the books and looks for the equipment, so it finds ghost assets that overstate what the company owns.
It checks whether everything the company owns is recorded. It starts from the floor and looks for the record, so it finds zombie assets that understate what the company owns.
A ghost asset is recorded but no longer there or usable. A zombie asset is there and in use but was never recorded. One is an existence failure, the other a completeness failure.
A sample drawn from the register only contains items the register already lists. Equipment that was never recorded can't be sampled, so finding it takes a check that starts from the floor.
No. Audit standards say a management representation is not sufficient evidence on its own, and obtaining one does not reduce the other procedures the auditor has to perform.
Not generally. Audit standards require inventory to be physically observed but set no equivalent general requirement for property, plant and equipment, outside narrow cases such as federal grant equipment.
Asset relationship management is a control layer above the ERP that captures physical evidence about assets and routes it through approval before the financial record changes.
A fixed asset register can be wrong because it lists what's gone or because it misses what's there, and a check that runs in one direction only finds one of them. Passing both tests takes evidence from the floor, captured where the equipment changes and reviewed before the register moves.
This is the gap SoloTruth's Asset Relationship Management platform (ARM) was built to close. Your own crews gather photographic, condition and other asset evidence in SoloTruth's mobile inspection app during work they already do, scanning the asset's QR code or barcode on the phones they already carry, with no extra hardware, and ARM verifies it with location and a timestamp. ARM routes every finding, including equipment with no record, to your accounting team for approval before the ERP is updated. SoloTruth has used the SoloTruth/VirtuSpect platform to manage and complete inspections, verifying condition and location, at over 1,000 businesses across the US.
Book a 30-minute strategy call at calendly.com/tim-harris-solotruth/30min to see how continuous verification changes what your fixed asset register is capable of.
Last Updated: October 2026