The item that generates the most friction on an initial diligence request list is rarely the audited annual financial statements; it is the trailing thirty-six months of monthly management accounts prepared on a consistent basis. Annual statements exist in most companies because statutory obligation produces them, whereas the monthly series either does not exist at all or is assembled retrospectively once the request arrives. Retrospective assembly is technically achievable, yet what the reviewing party reads from it is not the number but the moment at which the number came into being, since a time series produced after a process has commenced functions as a transaction-specific output rather than as an instrument the business actually used to manage itself. A second pattern observed in the same room aligns with the first: when the seller-side EBITDA adjustment schedule appears for the first time within the process, nearly every item on it runs in one direction.
That one-directionality is a consequence of when items are recorded rather than of bad faith. An expense enters the schedule not because it is genuinely non-recurring but because it appears non-recurring when examined after a price expectation has taken shape, while one-off revenues arising in the same period and lifting the margin are seldom screened with equal rigour. The reviewing party knows this and behaves accordingly: each adjustment is expected to stand on its own documentation, any item that cannot be evidenced is stripped from the base, and because every removed item is subsequently multiplied, its effect reaches several times its own magnitude. This is what the quality of earnings discussion means in practice — not how large the profit is, but which portion of it can be priced.
Technically, quality of earnings is the composite of three properties of reported profit: repeatability, cash conversion and operational origin. Repeatability asks whether a comparable result can be produced in future periods under comparable conditions; cash conversion asks how much of the accrued profit is collected within a defined interval; operational origin asks whether the result derives from the core business model or from incidental items. The mechanisms that erode quality along these three axes are well known, and most are individually defensible: recognising revenue by reference to invoicing practice rather than contractual delivery terms, applying an inconsistent threshold when capitalising development and maintenance expenditure, pricing related-party transactions on a logic that departs from market terms, and concentrating shipments near period end in a way that pulls forward demand belonging to the following period.
None of these choices is irrational in the conditions that produced them. In a mid-sized company the accounting infrastructure is typically configured around three external audiences — the tax authority, the lending bank and the shareholders — and each of the three expects a different presentation. The founder, meanwhile, carries the real economics of the business in his own head, because carrying it there has been sufficient until now and no external party has ever required that a single coherent picture be produced. The cost of the shortcut genuinely remains low until the conditions change, and the conditions change at a single moment: for the first time, the company will be priced by a party with no access to the founder's mind. From that moment onward, a multiplicity of views that caused no difficulty for years is reclassified as unverifiability.
The cash conversion axis is where that reclassification becomes visible fastest. The divergence between profit and cash does not surface in the income statement; it accumulates in receivables ageing, in inventory turnover, in the pre-depreciation balance of capitalised expenditure and in the tenor of prepaid items. A company whose operating profit rises across three years while free cash flow moves sideways has not thereby declared a problem, since working capital naturally expands through growth phases; but where the line item in which the divergence has accumulated cannot be identified, and the mechanism by which it is expected to reverse cannot be explained, the reviewing party tends to model the divergence as a permanent loss of quality. The criticality of the measurement dimension becomes apparent here: if cash conversion ratios, collection days and segment-level gross margin are not tracked on a regular cadence, the company does not hold the same information about the nature of its own profit as the party sitting across the table.
Contrary to common expectation, weak earnings quality rarely translates into valuation through the multiple. Multiples are anchored to market references and are comparatively rigid in negotiation; what yields is the base to which the multiple is applied and the architecture of the consideration. Unsupported adjustments come out of the base, the working capital target is set from a more conservative level than the trailing twelve-month average, a completion accounts mechanism is introduced, and a portion of the consideration is retained in escrow. Where quality remains genuinely uncertain, a meaningful slice of the price is tied to an earn-out — which is to say the seller accepts the obligation to prove, after closing, the repeatability of the profit he has asserted. Each of these structures looks reasonable in isolation; taken together they open a pronounced gap between the headline price and the amount actually received.
The second channel is quieter and accumulates across the process. Because every undocumented item transfers the evidentiary burden to the seller, the review period lengthens; the lengthening consumes the exclusivity window, multiplies second-round information requests and requires that the company be taken back to the buyer's investment committee a second time. The scope of representations and warranties widens, warranty periods extend, and the exclusions falling outside insurable risk multiply. On the lending side the effect appears on a different surface: the spread between EBITDA as defined in the credit agreement and EBITDA as used by management compresses the room available within leverage calculations, obliging the company to draw on its borrowing capacity at a level below its own internal reckoning. That is a cost which persists whether or not any transaction ever completes.
The mechanism that neutralises this tendency is not individual vigilance but record architecture, and it separates into four components. The first is a monthly close calendar with a fixed cut-off: each month closes within a defined number of days, no post-close adjustments are made, and where they are made they are recorded separately. The second is a written and approved revenue recognition and capitalisation policy; alongside the existence of the policy, the date from which it took effect and whether it was applied to prior periods must themselves be on record. The third is a contemporaneous normalisation log — any unusual revenue or expense item is entered in the same ledger, in the month in which it arises, irrespective of the direction in which it moves the result. The fourth is a reconciliation between management accounts and statutory statements, prepared and held ready for every period; a difference between the two is not a problem, whereas an unexplained difference is.
Once those four components are in place, the sequence moves to measuring earnings quality, assigning ownership for it and detaching it from any individual. On the measurement side, cash conversion ratios, collection and inventory days, segment-level gross margin and the share of recurring revenue in total revenue become standing items in the monthly report, presented to the board or the shareholders in the same format and on the same cadence. On the ownership side, a single role is defined as accountable for completing the close on time and in accordance with policy, and that role is positioned on a reporting line unconnected to any sales target. On the continuity side the test is straightforward: with neither the founder nor the finance director engaged, how many days does the existing team require, and on the basis of which documents, to reproduce the normalised profit bridge for the last three years. The answer alone reveals whether earnings quality is an institutional capability or a body of personal knowledge.
BEIREK typically establishes three things in this area. The first is a reporting cadence that binds close discipline to a calendar and fixes the cut-off date, the accountable role and the reconciliation output for every month; that cadence is run before a transaction is even contemplated rather than at the start of a process, since a series not produced contemporaneously does not carry the same evidentiary weight when reconstructed. The second is the normalisation ledger: unusual items are recorded in the month they arise, without regard to direction and together with their supporting documentation, and that ledger is presented throughout the process as the single reference against the counterparty's adjustment schedule. The third is committing the method of producing the profit bridge to writing — which account, through which adjustment, arrives at which base, documented so that another person following the same steps reaches the same result.
The purpose of this intervention is not to present profit as larger, but to ensure the company knows which portion of its profit is defensible before the counterparty does. The outcome observed in practice is generally this: in companies maintaining a normalisation ledger, a subset of adjustment items is eliminated by the company itself ahead of negotiation, while the items that remain survive in the base precisely because they are documented; the aggregate effect is a more defensible base and a materially shorter diligence timetable than an unprepared process would produce. The shortened timetable is itself an element of value, since a review completed within the exclusivity period forecloses the buyer's option to develop an alternative.
Quality of earnings is ultimately not an accounting heading but the degree to which a company's statement about its own economics can be verified from outside, and that degree is determined in the months in which the records were created rather than at the moment a transaction arises. The question genuinely under discussion at an investment committee table is not how much the company earned last year, but whether the same earnings can be produced and demonstrated with the founder's mind out of the loop.
Quality of earnings is ultimately not an accounting heading but the degree to which a company's account of its own economics withstands external verification, and that degree is fixed in the months the records were created rather than in the moment a transaction arises.