Financial Due Diligence: What It Covers, and What It Can't See

What a Full Scope Includes, End to End

Financial due diligence is not a single analysis. It is a bundle of them, run against the same books in the same weeks, each answering a different question about the same target.

A full scope usually carries seven parts:

• Quality of earnings. Normalized revenue and EBITDA, every adjustment itemized and defended.
• Revenue analysis. Recognition policy, pricing, concentration, and how durable the recurring base really is.
• Working capital. The level the business needs to operate normally, and the peg that follows from it.
• Net debt. Everything that behaves like debt at completion, not just the bank facility.
• Cash flow. How reliably earnings become cash, and what spending those earnings quietly depend on.
• Tax. The exposures a buyer inherits: filing positions, contractor classification, sales tax across states or borders.
• Forecast support. Whether the budget the price rests on has any history of being met.

The deliverable is two documents, not one. A report that argues, and a databook that shows: every schedule, every reconciliation, every adjustment with its workings attached. Sophisticated buyers spend more time in the second.

Two things are worth knowing before the engagement letter is signed. First, scope is negotiated, not standard. A scope built for a lender skips work an equity buyer needs, and a fast-close scope skips more still. Read the scope letter as carefully as the report itself, because the firm stands behind what it agreed to examine and nothing else.

Second, the work ends later than people expect. The report is not the finish line. After it comes the price bridge, the peg negotiation, the completion mechanism, and a set of purchase agreement definitions that decide what the findings are actually worth. The team that built the analysis should stay in the room for all of it. Every adjustment they found becomes a number someone argues about, and the person who found it argues it best.

The Quality of Earnings Process, Step by Step

A quality of earnings review runs in a sequence, and the sequence matters. Each step earns the next.

It starts with tie-out. The trial balance is reconciled to the financial statements and to the tax returns, so that every later number traces to one source. Gaps here are findings in themselves.

Then proof of cash. Reported revenue and expenses are reconciled to actual bank activity, month by month. It is the simplest test in the process and the hardest to argue with. Books can be adjusted. Deposits are deposits.

Then the management sessions. Not interviews about strategy. Working sessions about entries: why this accrual, why this reserve released in March, why this credit note. The quality of the answers is evidence, and an evasive answer gets noted as carefully as a clear one.

Then the adjustments, which are the heart of the work:

• Non-recurring items. Litigation, a flood, a one-off project. Genuine one-offs are rarer than sellers believe.
• Owner effects. Compensation above or below market, personal costs run through the business, related-party rent.
• Out-of-period items. Revenue and expense recorded in the wrong year, moved back where they belong.
• Pro forma changes. A lost customer removed from history, a signed price increase reflected forward.

Alongside the adjustments runs trend work: revenue and margin by month across the whole period, so that seasonality, run-rate, and the trajectory into close are visible rather than assumed.

Each adjustment moves EBITDA, and each is an argument rather than a fact. A sell-side review finds the adjustments that flatter. A buy-side review finds the ones that discount. The honest number is usually reached by fighting over the list line by line.

The output is a bridge: reported EBITDA at one end, adjusted EBITDA at the other, every step between them named. That bridge is the single most consequential exhibit in the deal, because the price is a multiple of where it lands.

Where Revenue Quality Gets Tested

The same dollar of revenue lives in four places: contracted, billed, recognized, collected. Revenue quality work is largely the discipline of checking that those four agree, and asking why wherever they do not.

Recognition policy comes first. When does the company call revenue earned: at signature, at delivery, over the term? The policy has to match what the business actually does, and a change of policy in the run-up to a sale deserves particular attention. A policy change is the cheapest way to manufacture growth.

Cut-off testing follows. Sales recorded in the last days of a period get traced to shipping documents, project records, or usage data, to confirm the revenue belongs where it sits. Quarter-end spikes that repeat every quarter are a pattern, not a coincidence.

Then the gross-to-net walk. Headline billings shrink through discounts, rebates, credit notes, and refunds before they become real revenue. A widening gap between gross and net is often the first visible symptom of pricing pressure the narrative has not caught up with. Presentation gets its own check here: a company reporting pass-through amounts as its own revenue can look twice its real size while the margin quietly tells the truth.

Beneath the total, cohorts. A flat revenue line can hide a business replacing lost customers with new ones at rising cost. Revenue by customer by year shows retention, expansion, and the true age of the base. Concentration gets measured the same way: by customer, by product, and by the person who owns the relationship.

For any business paid in advance, deferred revenue is the closest thing to truth serum. Cash collected for work not yet delivered sits as a liability and flows into revenue over time. If reported revenue is growing while deferred revenue shrinks, the company is recognizing its backlog faster than it is refilling it. The income statement says growth. The balance sheet says borrowed time.

Working Capital and the Peg

Every operating business traps cash in its own cycle. Receivables not yet collected, plus inventory not yet sold, less payables not yet paid. That trapped amount is working capital, and a buyer expects to receive a normal level of it at close, the way a house is expected to come with its plumbing.

The peg is the number that makes the expectation enforceable. The parties agree what a normal level of working capital looks like, and the purchase price adjusts dollar for dollar against it at completion. Deliver less than the peg and the price comes down by the difference. Deliver more and it goes up.

Setting the peg sounds mechanical and is not. The usual starting point is average net working capital over the trailing twelve months. Then the arguments begin. A seasonal business has no single normal level, so the peg has to reflect the season the deal closes in. A growing business needs more working capital every month, so a backward-looking average understates what the buyer will actually need. One-off items in the history distort the average and get stripped out.

The quieter fight is definitional, and it is usually worth more than the arithmetic. Whether a given item counts as working capital, as net debt, or as neither is decided in a definitions schedule, and each placement moves real money. An accrued bonus classified as working capital dilutes into an average. Classified as debt, it cuts the price in full.

Two mechanisms carry the peg's logic. Completion accounts measure the balance sheet after close and true the price up. A locked box fixes price on a historical date and bars value from leaking out afterwards. Both stand on the diligence team's view of normal.

Sellers also manage toward a peg. Stretch payables, drain inventory, chase collections hard, and the closing balance sheet flatters. The defense is simple to state and slow to execute: compare closing-date behavior against the trailing pattern, and treat departures from that pattern as adjustments rather than luck.

Net Debt and the Items That Behave Like Debt

Enterprise value is the headline. Equity value, the amount the seller actually receives, is that headline minus net debt. So the definition of net debt is not an accounting detail. It is price.

The bank facility is the easy part. The contested territory is the list of items that behave like debt without being called debt: obligations that exist at close, belong to the past, and will be paid with the buyer's cash.

The recurring candidates:

• Accrued bonuses and commissions earned before close but not yet paid
• Income tax owed for pre-close periods
• Overdue payables stretched beyond normal terms
• Customer deposits and, in some deals, deferred revenue
• Earn-outs and deferred consideration from the target's own past acquisitions
• Finance leases and equipment obligations
• Unfunded pension commitments
• Capital expenditure committed but not yet paid
• Provisions for litigation or restructuring already underway

Each item moves equity value in full, which is why a dry-looking schedule can swing a deal by more than the entire EBITDA argument. It is also why sellers argue placement rather than existence. Nobody denies the bonus accrual is real. The argument is whether it is debt-like, working capital, or an ordinary cost of doing business.

Deferred revenue is the classic gray zone. The seller says it is not debt because no cash ever leaves. The buyer says it is an obligation to deliver services after close, funded by cash the seller already banked. Both positions are defensible, which is exactly why the item gets negotiated early or fought over late.

The practical discipline is to build the enterprise-to-equity bridge in the first weeks of diligence and share it. A debt-like item surfaced in week three is a negotiation. The same item surfaced at signing is a crisis.

Cash Conversion and the Quality of Free Cash Flow

EBITDA is a set of judgments. Cash is an arrival. The distance between the two is cash conversion, and it is one of the most revealing measures a diligence team computes.

The arithmetic is plain. Start with adjusted EBITDA, subtract capital expenditure, adjust for the working capital the period consumed or released, and compare the result to where you started. A business converting most of its EBITDA into free cash explains itself. A business converting little needs a reason, and the reason is usually one of three.

Growth that eats cash. Rising receivables and inventory absorb money even when the growth is healthy. This is the benign explanation, and it still matters, because the buyer funds that appetite after close.

Capital intensity the income statement hides. Equipment refresh cycles, fit-outs, tooling. The split that matters is maintenance versus growth spending: what it costs to stand still versus what was chosen for expansion. Sellers habitually present all of it as growth. History usually disagrees.

Capitalization policy. Costs moved to the balance sheet leave EBITDA immediately and return quietly as depreciation. Development labor is the common case. Two otherwise identical software companies can report materially different EBITDA on this choice alone, so diligence reads the policy, quantifies the capitalized costs, and re-runs the numbers with those costs expensed.

Conversion also disciplines the earnings work. An adjustment that raises EBITDA but never shows up as cash invites a harder look at the adjustment.

Free cash flow quality is the second question, and it is about repeatability. One good year of conversion can be manufactured: collect hard, stretch suppliers, defer spending. So the test is conversion measured across several years, with the working capital games backed out. That is the figure worth pricing. A single strong year recorded in the months before a sale is close to meaningless, and experienced buyers treat it that way.

The Red Flags That Recur

Most financial red flags are not fraud. They are strain, and strain shows up in the books in repeatable ways. The patterns below recur across deals of every size.

• Receivables growing faster than revenue for several quarters. Customers may be paying slower, or revenue may be booked ahead of reality. Either way, cash will eventually say so.
• Margins improving with no operational story. Real margin gains have causes someone can narrate: price, mix, automation, scale. Unexplained improvement before a sale usually lives in an accounting policy.
• One-time adjustments that happen every year. A restructuring charge in each of four consecutive years is not one-time. It is the cost of running the company.
• Revenue spiking in the final weeks of each period. Repeated cut-off spikes suggest the calendar, not the customer, is driving recognition.
• Deferred revenue falling while reported revenue grows. The backlog is being consumed faster than it is being replaced.
• The books closing later each month. Close discipline decays before the numbers do. A close that has drifted from day five to day twenty is a finding about the finance function itself.
• Auditor changes without a clean explanation, or an audit firm out of proportion to the company's size.
• Related-party transactions at terms no outsider would accept.
• A change of accounting policy inside the deal window, whatever the stated reason.

A red flag is a question, not a verdict. Any one of these can carry an innocent explanation, and most do. The pattern that should genuinely change behavior is convergence: three or four flags pointing the same direction in the same period. That is when a diligence finding stops being a bullet in a report and becomes a price adjustment, a restructured deal, or a reason to walk.

The Timeline, the Mesh, and Reading the Report

Financial due diligence typically runs four to eight weeks, and the spread is mostly data quality. Clean monthly closes and a well-kept ledger sit at the short end. A shoebox of spreadsheets sits at the long end.

The arc is consistent. Early weeks: the data request, the tie-out, and the first management sessions. Middle weeks: the analysis itself, quality of earnings first because everything else keys off it. Final weeks: the draft report, a factual-accuracy pass with management, and the final deliverable with its databook.

The workstream does not run alone, and the mesh points are where deals are quietly won or lost. The earnings bridge feeds the valuation case. The working capital analysis becomes the peg and its definitions schedule in the purchase agreement. Tax findings become specific indemnities. Legal diligence needs the customer concentration table to decide which contracts to read first for change-of-control clauses. Commercial diligence tests the forecast; financial diligence tests whether forecasts here have ever come true. Sequence these handoffs deliberately and the deal timeline pays for it twice.

For the non-accountant reading the final report, a short discipline:

• Read the scope and limitations pages first. What the firm did not examine is as informative as what it did.
• Go to the adjusted EBITDA bridge before the executive summary. The bridge is the finding. The summary is the interpretation.
• Treat every adjustment as a claim. Ask which ones management contested.
• Search for the word "unable". It marks the places where data was requested and never arrived.
• Ask the team what concerned them but did not clear the bar for print. In many engagements the best finding is delivered verbally.

The report is not a verdict on the deal. It is the evidence table. Judgment stays with the buyer.