When depreciation policy is requested in an investment review, the file that lands in the data room is typically not a policy but a schedule: asset code, capitalization date, rate, accumulated depreciation, net book value — a direct export from the accounting system. Every company has this list, because financial statements cannot be produced without it. What the reviewing party is looking for is not the list itself but the judgment standing behind each of its lines: why that asset is amortized over seven years rather than ten, why one maintenance invoice was capitalized while a comparable one was expensed, who set that rate, on what reasoning, and under whose approval. The schedule is an output; the policy is the decision set that generates the output, and at the diligence table the difference between the two becomes visible almost immediately.
In companies without a written policy, three patterns tend to appear together. The first is the same asset class carrying different useful lives across different facilities, usually not because of any technical distinction but because the facilities were acquired in different years under different accounting practices. The second is the size of the population of assets carried at zero net book value while continuing to run on the production line. The third is that maintenance and overhaul expenditures of comparable character are expensed in some periods and capitalized in others. None of these patterns, taken alone, indicates irregularity; observed together, they indicate that the policy lives inside the company not as a document but as a convention held in the memory of particular individuals.
The origin of this condition is not negligence but a shortcut that genuinely lowers cost under certain conditions. Tax legislation publishes ready-made useful life tables by asset class; where those tables are followed, the calculation becomes mechanical, audit and inspection exposure is minimized, and no technical assessment is required. For a closely held company with limited external financing, the choice is rational, since the depreciation figure binds no decision beyond the tax base. The difficulty lies not in the shortcut but in its persistence after the conditions change: from the moment the company begins to borrow, to report under financial reporting standards, or to admit an outside shareholder, depreciation ceases to be a component of the tax computation and becomes the lens through which operating performance and maintenance capital requirements are read.
The judgments embedded in a depreciation policy extend well beyond a single rate. The capitalization threshold determines below what amount expenditure is charged directly to the income statement, and in practice decides whether low-value, high-frequency consumables and spare parts appear on the balance sheet or in the profit and loss account. Component separation determines whether the differently aged parts of a plant or line — building shell, mechanical equipment, control systems, wear groups with markedly shorter lives — are depreciated separately or as a single asset. Residual value; whether depreciation commences on the invoice date or the date of commissioning; the election between straight-line and reducing-balance methods; the asset groups for which units-of-production amortization is appropriate; whether major overhauls are capitalized as a separate component; whether right-of-use assets are amortized over lease term or economic life; and the definition of the conditions that trigger an impairment review — each of these is an assumption made about the future. Their sum produces a figure that, in most companies, no one reopens.
The first channel into valuation opens from a place that appears, at first sight, unrelated. Because transaction price is usually built on an EBITDA multiple and depreciation sits by definition below that line, the policy may seem to have no bearing on price. What the policy actually governs, however, is not the amount of depreciation but which expenditures never become subject to depreciation at all, being charged directly to expense. Where the capitalization threshold is set high, or maintenance spending is systematically expensed, operating profit is suppressed; under the opposite configuration the same expenditure is carried to the balance sheet and lifts EBITDA. In a transaction priced on a multiple, that classification preference enters the outcome not as the difference of a single period but as a price differential enlarged by the multiple — and the buy-side quality-of-earnings exercise is constructed precisely to test that boundary.
The second channel is the maintenance capital expenditure bridge. In building normalized free cash flow, a buyer must estimate the scale of recurring replacement investment, and the readiest available proxy is the depreciation charge; that proxy works only to the extent that useful lives reflect the actual replacement cycle. Where lives are defined materially longer than the real interval of replacement, the bridge breaks: historical periods show low depreciation and high profitability, while the first years after closing produce an unbudgeted renewal burden. The population of fully depreciated assets still in production is the most direct signal of this exposure, representing as it does a wave of renewal that is undated but effectively certain. A finding of this kind typically produces one of three outcomes — a direct reduction in price, a technical condition assessment imposed as a pre-closing requirement, or an escrow item held for a defined period.
The measurement dimension of the review is, in most companies, a door never opened. Useful life is an assumption, and like any assumption it can be tested retrospectively; the data required for the test already sits inside the fixed asset register, since the capitalization and disposal dates of every retired asset are recorded there. A structure in which the gap between assumed life and realized service life is compared periodically at asset class level constitutes evidence that a reviewing party rarely encounters and that, when encountered, markedly shifts its view on the accuracy of management estimates. A second verification layer attaches to this: reconciliation among physical count, register, general ledger, and insured values. Inconsistency across those four surfaces reads less as a depreciation issue in isolation than as a signal about the state of asset management as a whole.
The structural gap that emerges under the ownership and continuity dimensions is the following: the nominal owner of the policy is finance, while the substance of the judgments it contains originates with engineering and maintenance. The party that knows an item of equipment's real service life, its wear behavior, its overhaul interval, and the supply horizon for its spare parts is not accounting; accounting, nevertheless, holds sole authority to fix the life. Absent a defined interface between the two, the decision devolves in practice to whoever books the invoice, and on the day that person leaves, the only source of the policy leaves with them. Founder dependency accumulates here with particular quietness: a configuration in which capitalization decisions are taken by a single individual against the profit expectation of the period is by definition not reproducible, and is therefore priced not as institutional capability but as a temporary arrangement resting on a person.
The mechanism that neutralizes this tendency is architecture rather than awareness, and it comprises four separable components. The first is a useful life and residual value table constructed at asset class level on a written rationale drawn from the company's own maintenance history and replacement cycle rather than from the tax schedule; the difference between the tax base and the reporting basis is not eliminated but held visible in a separate reconciliation. The second is that the record of the capitalize-or-expense decision is created at the point the expenditure is requested rather than when the invoice is posted, with a brief technical opinion on life-extending effect attached to the request. The third is an annual back-testing rhythm: comparing the realized service life of retired assets against the assumed life, and carrying the deviation, with reasoning, into the following period's table. The fourth is separation of authority — engineering proposes, finance applies, the body responsible for audit approves — and this separation is the component that actually detaches the policy from individuals.
BEIREK's intervention in this area is not a recalculation of the existing depreciation schedule but the surfacing of the decision chain behind it. Realized service lives by asset class are derived from historical disposal data in the fixed asset register and compared against assumed lives, with component separation established for the groups where the deviation is largest; where shell, mechanical equipment, and short-lived wear groups are separated within the highest book value asset classes, the depreciation charge typically changes not in aggregate amount but in timing profile, and the maintenance capex bridge becomes coherent for the first time. Running in parallel are a decision record capturing capitalization judgments at the point of request, a four-surface reconciliation cycle spanning count, register, ledger, and insurance, and a review calendar for the annual back-test.
The consequence of this structure in a sale or capital-raising process is direct: at the diligence table the policy is presented not as an assertion but as a record demonstrably applied under the same rules across at least three periods, and the ability to demonstrate consistency moves the discussion from the accuracy of the figure to the reasonableness of the method. What the counterparty is doing at that point is not recomputing the number but establishing whether the number issued from a judgment or from a default; because a company's depreciation policy is, in the end, its written view on how long its own productive capacity will last, and the way that view was formed is an indicator, not easily falsified, of how well the company knows itself.