How to Spot Weak Books Before You Sign an LOI

The seller slides a P&L across the table. Clean columns, round numbers, a healthy bottom line. It looks great, and it's a presentation, not an audit.

Mix the two up and you pay for the diligence, burn the time, and put your name on a number the books can't support. The red flags worth catching are sitting on those first financials. You just have to know what to look for.

What weak books look like

Weak books are financials that present well but won't survive a buyer checking them. Before an LOI the signs are readable: round numbers that never tie to a tax return, add-backs that swell the earnings, cash-heavy revenue with no paper trail, margins that jump for no clear reason.

None of these proves fraud by itself. Each flag marks a spot where the number on the page is an assumption dressed up as a fact, and your job before the LOI is to tell the two apart.

The audit comes later. Right now you're deciding whether the earnings are real enough to justify the diligence spend and the price you're about to anchor, so it's a screen. Buyers who skip it sign on the seller's framing, and buyers who run it sign on a number they've started to test.

And it's cheap. A flag caught before the LOI costs you an afternoon. The same flag caught in diligence costs you the fees and a month, and after you sign, a soft number resells later at the same discount you should've caught.

Flags you can read without diligence access

You don't need a data room for most of these. They live on the P&L, the tax returns, and the bank summary any serious seller will share early.

• Round numbers everywhere. Revenue at $1,200,000 and "owner expenses" at $80,000 are estimates. It tells you the books were rebuilt for the sale, so you can't trust any one line to be sourced yet.
• A P&L that doesn't tie to the tax return. The statement shows more profit than the filed return. You don't know which one's true, only that the seller swore the lower number to the IRS and showed the higher one to you.
• Add-backs doing too much work. When a third of the earnings is add-backs the seller wants you to trust, the real number's buried under adjustments. You can't verify it until every add-back produces a receipt.
• Cash-heavy revenue with no trail. A business running on undeposited cash can't prove its top line. If you can't confirm the revenue exists, you can't finance it or insure it either.
• Margins that jump for no reason. Gross margin climbing five points in the year before a sale usually means deferred costs or revenue pulled forward. You can't tell engineered earnings from real ones until the trend's normailzed.
• One customer carrying the place. If one account is 30% of sales, the earnings are one phone call from falling apart. You don't know yet if that revenue transfers, so it's concentration risk being sold as stability.

Go through the actual statements line by line, the way you'd read any small business P&L before you buy. Once you're looking, none of these are subtle.

One flag is a question. Three or four together is a pattern, and the pattern's usually that the books got dressed up for the sale.

A clean P&L proves almost nothing

A clean P&L tells you the seller can format a spreadsheet. It doesn't tell you the earnings are real, repeatable, or transferable. It's a picture of the business drawn by the person with the most to gain from your offer, and treating it as verified earnings is the most common way buyers overpay.

It helps to think about who made the books. Usually it's an owner who spent years keeping taxable profit low and is now, for exactly one transaction, trying to push it high. The same person who rounded expenses down for the IRS rounds add-backs up for you, and that's the gap you're reading.

So ask the buyer's question instead: what's left after the seller leaves? Two businesses can post the same clean earnings and the one that runs without its owner is worth a lot more. The statement looks identical either way. Only the cash-flow read shows the difference.

That's also why the seller's number is usually just an estimate to start with. 86% of small business owners have no professional valuation or only a rough estimate, so the books you're handed were rarely built to be checked. A formal quality-of-earnings review comes later, and the pre-LOI screen is the cheaper version you run yourself first.

Stress-testing the books before you sign

You can pressure-test the earnings with what a serious seller will already share. Go in order, since each step tells you whether the next is worth doing.

1. Tie the P&L to the tax returns. Pull three years of filed returns and match them to the statements. The return is the number the seller swore to, and a gap here ends most deals on its own.
2. Rebuild the SDE from the source. Take the bank deposits and the returns and rebuild earnings yourself, because the seller's recast is an argument for a price. You want to value the business on numbers you built.
3. Pull the owner's labor out of the profit. Subtract the market cost of replacing every hour the owner works. Earnings that depend on unpaid owner time don't transfer, and what's left is closer to what you'd actually inherit.
4. Normalize the trend. Look at three to five years, since a single strong year right before a sale is often engineered. A flat or rising margin you can explain beats a spike you can't.

None of that needs a paid provider or signed access. Three years of returns, the bank summary, and a few hours of your own attention.

If the books pass, you sign the LOI on a number you've started to verify. If they don't, you've saved the diligence spend and walked before the price got anchored.

Pass, and the formal work starts. The signed LOI opens the full diligence request list, where a quality-of-earnings provider confirms what your screen flagged. Your screen just decides whether the company earns that review, and what it's worth going in, which is the core of how to value a small business in the first place.

FAQ

What are the financial red flags in a small business before you sign an LOI?

The clearest red flags are a P&L that does not tie to the filed tax returns, round-number "estimates" instead of records, add-backs carrying a large share of the earnings, undocumented cash revenue, and margins that jump without a cause. Each one is a question the seller has not yet answered, and three together is usually a pattern.

Can you spot manipulated books before due diligence?

You can spot most manipulation before formal diligence by reading the statements the seller shares early. Match the P&L to three years of tax returns, check the add-backs, rebuild the number from bank deposits, and manipulation shows up as gaps between those sources.

What financials should you review before signing an LOI?

Before signing, review three years of P&Ls, three years of filed tax returns, the bank deposit summary, and the add-back schedule, then tie them together and rebuild the SDE yourself. If the seller will not share enough to do that, treat the refusal as a red flag of its own.

You cannot price a business on books you have not tested.

Keystone is closed while the build finishes, so this check waits for now. When it opens, the diagnostic will give you three scores and your value range with the most likely figure locked, calibrated against 10 years of BizBuySell Insight Reports and 1.6M+ SBA 7(a) loan records, and show what the earnings are actually worth and what is discounting them.

Join the waitlist at https://app.trykeystone.io.

For deeper work, Keystone Command ($199/mo, $1,990/yr) runs the Deal Analyzer a buyer uses to screen targets, and Keystone Operator ($129/mo, $1,290/yr) pressure-tests your own books if you are the operator preparing to sell.