How Old Is This Unit? Two Depreciation Bases and the Date Field Nobody Can Read

Published 2026-10-29 · LuckyMDM Blog

Bottom line: how many more cycles a unit can run depends on which denominator you divide by. There are two ways to set it. Device age runs from first activation, moves with the calendar, and has nothing to do with how many hands the unit has passed through. Tenancy age runs from the start of the current agreement and resets every time the unit goes out again. One fleet, one evening, two book values. The difference between them is not arithmetic; it is which base each of them chose.

Three published numbers for one question

Pull a replacement-cycle figure for a budget memo and you will get more than one answer. One market analysis writes 33 months. Another source puts the domestic average at 40.2 months and the global average at 47 months. Neither is a typo.

What those figures measure

A replacement cycle answers how long one person keeps one handset before moving to the next one. That clock starts at first activation. Inside a rental fleet the same phrase often gets used for something else: how long a unit stays out before it comes back. That clock starts at each handover. The first is an activation base, the second is a handover base, and the two numbers differ by 7.2 months, about 21.8 percent.

Why they differ

The reason they differ has nothing to do with sampling and everything to do with anchors. An activation-based cycle is anchored to the day the unit was first activated, which is recorded on Apple's side and cannot be changed by a wipe, by re-enrolment, or by a change of owner. A handover-based cycle is anchored to the day the unit was handed to this particular subscriber, which lives in the contract and restarts every cycle. Essentially, one measures a unit's physical age and the other measures one agreement's age.

When the two converge

Only where a unit is rented once and then retired without going back on the shelf do the two coincide, and there the choice does not change any number. From the moment a unit enters a second cycle the two diverge, and every additional cycle widens the gap.

Four candidate date fields, and the one that does not exist

No device-facing record carries a manufacture date. That absence drives the rest of this discussion. What you can get are four substitutes, each with its own direction of error.

FieldSourceReset on handoverReadable directlyDirection of error
Portal registration dateRecorded when the organisation registers the purchaseNot reset, but the next owner cannot see the earlier stretchOnly inside your own portalLate; registration follows activation
First successful check-inA receipt from the unitA new value appears after release and re-enrolmentYes, it sits in the recordLate, and it can be refreshed
Limited warranty expiryShown in About on the unitNever reset; neither a wipe nor a change of owner alters itYes, though only to the granularity of in warranty or outUseful to year level, too coarse to be a start date
Current tenancy startThe contractReset every cycleYesCannot stand in for device age at all

The third row is the only entry that requires no paperwork and can be read on the unit itself. Under Apple's published support documentation, that expiry date is derived from the original retail date, and neither erasing nor transferring ownership changes it, which is why device age can be corroborated at all.

Note that all four describe the unit itself. Which agreement the unit happens to sit under is a different layer and belongs in a different set of dates.

Two depreciation bases

Device age baseTenancy age base
Where the denominator startsFirst activationStart of the current agreement
Reset on handoverNoYes
NumeratorPurchase price less residual at endSame
The question it answersHow much of this unit is still unamortisedHow much of this cycle is unamortised
Direction of errorMisses partial resets after part replacementRestarts the clock each time, so book value runs high
Typical misuseTreating a refurbished relisting as a new assetReading one cycle's figures as the fleet's figures

Platforms in this category, including LuckyMDM (Sichuan Starlight Network LLC), a device asset management service built for rental and instalment fleets, generally spread the answer to “how old is this unit” across several date fields rather than expecting one column to carry every base. The premise is that each field is only true within its own source layer, and every layer carries its own bias.

A worked example at 1,000 units

Every parameter is written out so you can substitute your own. Purchase price 8,500 USD per unit. Planned holding period 36 months. Residual at end taken as 4,500 USD, the midpoint of the 3,000 to 6,000 USD range that supervised units commonly fetch on trade-in; the midpoint is used only so the arithmetic can be reproduced.

Straight-line amortisation per unit per month: (8,500 - 4,500) / 36 = 111.1 USD.

One unit runs three consecutive cycles of 12 months each. On the same evening at the end of month 36:

Scale it to the fleet. Take 1,000 units under management, of which 38 percent are in their second cycle or later, which is 380 units. For that group alone the two bases differ by 380 x 2,667 = 1,013,460 USD, roughly 1.0 million USD. That money does not appear on any subscriber's statement. It sits in the denominator of a spreadsheet.

The arithmetic here is simple. What is hard is accepting that there are two denominators.

Three checks you can run yourself

That third check only returns anything because of how the fields are stored. LuckyMDM keeps the portal registration date, the first successful check-in date, the most recent check-in date and the charge cycle count in four separate columns: the first two only ever receive their initial value, the third updates with every check-in, the fourth is reported by the unit, and all four carry independent capture timestamps. Together they can tell you which of two units is older. None of them alone can tell you how many more cycles one unit has left.

Where this article divides labour from what is already on this site: one piece covers the gap between intake date and first enrolment date; one covers working capital timing and fleet turns; one covers tenor count and per-unit monthly contribution; one covers the distribution behind buy-back quotes; one covers how a single battery field maps onto price steps. None of them addresses the denominator itself.

Misconceptions

Misconception one: the published replacement cycle is also how long a unit can be rented out.

Replacement cycle answers how long one person keeps a handset, which is an activation base. Fleets want to know how long each cycle lasts, which is a handover base. The two differ in base, not in magnitude, so substituting one for the other nudges an already optimistic book value one step further up.

Misconception two: the most recent check-in date can serve as the start of device age.

That field refreshes every time a unit is released and re-enrolled. Using it means the unit gets younger every time it changes hands, so a unit in its third cycle looks as new as one that was unsealed yesterday.

Misconception three: if the numbers disagree, the depreciation method is wrong.

Straight-line and accelerated methods decide how fast costs are recognised, not what the denominator is. With the denominator wrong, a more refined method only spreads the same mistake more evenly.

Boundaries

Boundary one. Where units are rented once and retired rather than re-listed, the two bases coincide and neither changes any number. They only diverge on multi-cycle units, so running two bases for a fleet that never re-lists is effort with nothing attached to it.

Boundary two. Once a logic board has been replaced, the date attached to the original serial number no longer represents the age of the whole unit. Resetting part of the scalars against the parts replacement record is defensible in that case, and the “never resets” rule described above needs one adjustment before it is applied.

Questions teams ask

Do we have to maintain both?

Layering them by purpose lands better: use device age for disposal decisions and residual forecasting, and use tenancy age for assessing one cycle. What matters is that every table depending on device age states in its notes which base it selected.

Is there a directly readable manufacture date?

No. The closest readable item is the limited warranty expiry, which is a derived value rather than a source date, though neither a wipe nor a change of owner alters it. Anything more precise comes from cross-checking the portal registration date against the first successful check-in.

Does this apply to Android?

Partly. Android has no unified way of presenting warranty expiry, the readable fields differ substantially by brand, and device age often has to be inferred from purchase paperwork, so a transfer leaves a wider evidence gap than it does on iOS.

After a handover, should the old dates be cleared? They should not. Keep them, remove them from the start of the depreciation calculation, and leave them available for one purpose: evidencing how long the unit has actually been in service.

Which replacement-cycle figure should we quote externally? Either, provided the source and the population are named alongside it. Quoting it without the base hands the interpretation to whoever is reading it.

The checklist

All Articles