Accounting Records in Due Diligence

The most common misconception is that a diligence request list is a fixed document. It is not. We maintain several, pitched at different deal sizes, and the choice between them is made before anything is sent.

The first question we ask is not about the business at all. It is about the seller: how they work, and how comfortable they are with technology. A request list that assumes a controller, a data room and clean exports is the wrong instrument for an owner who runs the business out of one accounting file and a filing cabinet, and sending it produces delay and anxiety rather than documents. What gets asked, and how, follows from that answer.

The formal version of this process, a very large labelled request list returned as a set of labelled responses, does happen on larger mid-market transactions. In the lower middle market it mostly does not. Far more often the work is getting the right information, in a usable format, from someone who has never been through this before.

Diligence follows a rhythm, and knowing it removes most of the anxiety about what is coming next.

1. Access first. If there is a data room, we go through what is already in it before asking for anything. A great deal of what gets requested is often already there, and asking for it again wastes the seller’s goodwill early, when you need it most.
2. A gap request, not a wish list. Whatever is genuinely missing gets asked for, laid out plainly.
3. Two or three days inside the material. General ledgers and bank statements are the main two, with monthly profit and loss statements and balance sheets alongside them. This is where the questions actually come from.
4. The first management call, in the first week. Usually three or four days in.
5. A second, much more specific request list. Generated by what the first pass found.
6. Occasionally a second call, nearer the end. Most engagements need only two request lists in total.

For establishing whether financial statements can be relied on, general ledgers and bank statements come before everything else. They support the reconciliations and tie-outs, and those have to happen before any analysis means anything.

A data room containing polished summaries and no ledger detail cannot be worked with, however well organised it looks. Trial balances are a genuine positive signal about the state of the bookkeeping, and they are still summaries. You cannot trace a deposit to a customer from a summary.

Where a seller is able to grant read-only access to the accounting system directly, it removes a significant portion of the request list on its own, and the difference in pace is substantial. QuickBooks Online versus desktop covers why that access matters more than any feature comparison.

This is the part sellers underprepare for, and in our experience it is the most valuable hour in the whole exercise.

In an owner-operated business, a large part of how the company actually works lives in the owner’s head rather than in any document. They know how they bill, how payroll runs, how the ledger is maintained, how journal entries get made, who does the bank reconciliations. None of that is written down anywhere, and none of it can be inferred reliably from the file. So we push hard to get the owner on the phone, and early, and we get there on the large majority of engagements.

The questions are about process:

• How revenue is recognised, and when deposits are taken
• How progress billings work, and whether overbilling and underbilling are tracked
• Who performs the reconciliations, and whether they are actually being done
• Whether adjusting journal entries are manual, and who makes them
• Who has access to the bank

We also bring specific observations from the ledger into that call rather than saving them. Noticing that a company books month-end reclassifications and asking the owner to walk through why is a far more productive conversation than asking about accounting policy in the abstract.

Then we take the answers back and compare them against what the records actually show. Where those two things agree, the file gets easier. Where they diverge, that is the second request list.

The first list asks for categories. The second asks for individual items, by number and date, because by then we know what looks unusual.

Typically that means invoice sampling: five to ten invoices traced through the ledger and through the bank statement, weighted toward high value and high risk items and anything that looked out of place, rather than selected at random. It is a litmus test on whether transactions are recorded accurately and in the right amount. Alongside it runs the proof of cash, which ties bank deposits to recorded revenue and bank disbursements to recorded expenses.

Worth saying plainly, because sellers often assume their situation is uniquely bad. It is not. We rarely see a deal arrive clean, with everything in one place and organised. Data rooms are usually messy. Working through that is a normal part of the engagement rather than a sign that something is wrong.

What actually causes problems is narrower than general untidiness:

None of those are accusations. All of them are time, and in a transaction time comes out of an exclusivity period. See quality of earnings red flags for the findings that go beyond housekeeping.

The preparation that pays is unglamorous and mostly consists of making your existing records answerable rather than producing anything new.

• Export a full general ledger from any system you are leaving, in a workable format, before access lapses
• Check that customer and vendor names are landing on transactions, particularly after a system change
• Keep bank statements for the full period in one place, in order
• Be able to explain your month-end entries and who makes them
• Know which basis each of your systems holds
• Decide early whether you will grant read-only access, because it changes the pace of everything