Asked at an investment committee meeting or across a board table why a production line, a warehouse bay, or a pool of software licences was acquired, the answer given is typically an account of why the asset was necessary — capacity was insufficient, lead times had lengthened, a customer had demanded it. Asked at the same table how much profit that asset has generated since it was purchased, the answer usually drifts upward to a company-level ratio and never descends to the individual asset. The distance between those two answers explains why, in the substantial majority of companies, return on assets remains not a management instrument but an outcome indicator computed once the year has closed. Assets are justified in operational language and measured in financial language, and the two vocabularies are never translated into one another.
The second and less noticed face of this gap is how quietly idle assets can sit on a balance sheet. A machine whose utilisation has fallen, a floor left empty after a relocation, a receivable pool whose collectability has weakened, a stock line whose turnover has slowed — each remains an asset in the ledger and generally reaches no one's agenda, for the simple reason that none of them has an owner and none of them has a review date. When a single company-level return on assets is reported, the yield produced by well-performing assets masks the dilution created by idle ones, and to the extent the resulting figure looks tolerable, no question follows. The ratio has been aggregated to precisely the degree required to hide its own components.
The mechanism operating underneath is not a management failure but, under particular conditions, a wholly functional shortcut. In a company in its growth phase, an asset acquisition decision is generally taken to relieve a bottleneck rather than to earn a computed return; if capacity is short, a machine is bought, and if the team no longer fits, space is leased. Deciding quickly is rational to the extent that it avoids the cost of delay. The difficulty lies not in the shortcut itself but in its persistence once the conditions have changed: when a company stops relieving bottlenecks and begins allocating capital, the same decision reflex now produces a stack of assets whose returns were never measured. The same mechanism also delays exit, since a body that authorised a purchase bears an institutional cost in certifying that purchase as unproductive by its own hand; the sunk cost fallacy — the tendency of an unrecoverable outlay to govern a future decision — appears here as the perpetual deferral of a disposal decision.
Looking at this picture, the reviewing party is generally expected to take an interest in the level of the ratio, whereas its actual interest lies in the ratio's decomposability. The question posed is not what return on assets amounts to in percentage terms, but which asset group carries that return, which group dilutes it, and whether the company has ever drawn that distinction on its own initiative. The question a company most commonly never asks itself runs roughly as follows: of the fixed assets acquired over the past three years, did each in fact produce the contribution promised in its acquisition case, and is that comparison recorded anywhere? The number of companies able to answer with a document is limited irrespective of revenue scale, because the justification lives in an investment proposal while the outcome appears only in a consolidated income statement, and the two are never set side by side on the same page.
The documentation dimension is narrower here than is generally assumed. Fixed asset registers, depreciation schedules, and inventory count records exist in nearly every company; these evidence the existence of an asset, not its productivity. The record treated as verifiable in a review is the one showing above which threshold an asset was submitted to which authority, under what return assumption it was approved, and when that assumption was tested retrospectively. Absent such a record, the assertion that an asset acquisition discipline exists remains an oral claim, and oral claims do not enter a valuation. The shortfall arises less from missing documentation than from the wrong layer having been documented.
The implementation and measurement dimensions then weaken one another in a reinforcing loop. Where return on assets is calculated only at year-end, during the audit or the accountant's closing work, the figure loses its character as a management input; returning today to a decision taken twelve months ago serves to explain that decision, not to correct it. Once the measurement rhythm becomes quarterly and descends to the asset-group level, the ratio begins to change behaviour: a stock line whose turnover has slowed, or a line whose utilisation has dropped, becomes visible while the cost of disposal remains bearable. Frequency of measurement proves more determinative here than precision of measurement, since what opens an intervention window is timing rather than accuracy.
Ownership is the layer most frequently left vacant under this heading. The owner of a purchase decision is generally identifiable — the founder, the general manager, or the relevant unit head; the responsibility for tracking the return on an acquired asset and, where warranted, proposing its disposal is typically assigned to no one at all. Finance keeps the asset in the register, operations uses it, and neither carries a mandate to ask whether that asset covers its cost of capital. This unowned space produces both an implementation gap and a dependence on the founder: in practice the only person following asset productivity is the founder, working from memory and instinct, resting on no record whatsoever. Such a configuration yields the conclusion that performance, however genuinely strong, is not transferable.
The institutional cost reaches valuation through three distinct channels. The first is the direct multiple effect: an asset base whose returns cannot be decomposed is read conservatively by a buyer, and in the normalised earnings calculation the dilution caused by idle assets is treated as permanent. The second is the working capital channel, where inefficiency on the inventory and receivables side enters price through the pre-closing working capital adjustment, the target level being anchored to historical averages — an adjustment that has, in most cases, never appeared in the seller's own arithmetic. The third is structural: an unmeasured asset base supplies grounds for broadening the representation and warranty package, raising the escrow proportion, or tying a portion of consideration to post-closing performance. What these three channels share is that none of them turns on the company's actual efficiency, only on its demonstrability.
The mechanism that neutralises this tendency is decision architecture rather than individual vigilance. Four workable components separate out. First, every asset acquisition above a defined amount is tied to a single-page record carrying, prior to acquisition, the expected return and payback assumption. Second, that record is created at the moment of proposal rather than at the moment of approval, since a justification written after approval is constructed backwards to legitimise a decision already taken. Third, a fixed review rhythm classifies asset groups quarterly by their contribution to return, incorporating utilisation and turnover data. Fourth, a line of responsibility charged with bringing disposal proposals forward sits apart from the authority that made the purchase. This fourth component is institutionally the hardest to establish and the highest-yielding, because where the person deciding and the person interrogating the decision are the same, interrogation is systematically late.
In capital-intensive projects and multi-asset structures, BEIREK typically opens this intervention by establishing an asset-level decision record: the acquisition case for each material asset line, the numerical assumption underlying that case, and the date on which the assumption will be tested are held in a single record, operated independently of the approving authority. Onto that is fitted a quarterly review rhythm separating asset groups by contribution to return; the object is not to improve the ratio but to break it into components, making visible which line carries and which line dilutes. The third layer defines a distinct role obliged to bring forward disposal and reassessment proposals, a role not consolidated in the same hands as purchase authority.
What this architecture produces on the continuity dimension is precisely what a reviewing party is looking for. Where a company's asset efficiency is high and that efficiency is explained by the founder's sectoral instinct, supplier relationships, or personal negotiating capability, what has been shown is a performance and not a capacity; it carries no evidence of recurrence following the founder's departure. Where the same efficiency is produced through defined thresholds, recorded assumptions, and a fixed review calendar, it becomes for a buyer a system capable of being assumed, and what is priced becomes future reproducibility rather than historical outcome. That difference is among the principal explanations for two companies arriving at an identical ratio and trading at materially different multiples.
Return on assets therefore merits being read as a governance indicator rather than a performance indicator. The ratio itself states where a company has placed its capital; how the ratio is produced and by whom it is monitored states where that company will place its capital next, and it is the second of these that an investor is buying. Where every line on a balance sheet carries an owner, a justification, and a review date, that balance sheet has ceased to be a record and become an instrument of decision. One question remains: which portion of the present asset base, were it presented for purchase today, would be bought again on the same reasoning?