Chapter 32: Self-Employed Borrowers: The Most Complex (and Most Rewarding) Files

"The accountant's job is to make the number small. The underwriter's job is to count the number. Nobody's job is to warn the borrower that those are the same number." — constructed; the operating premise of this chapter

Overview

A borrower who owns a business tells you they make about \$9,500 a month, and they are not lying. They know what comes out of the business. They know what goes into their household account. They have an accountant, they pay that accountant well, and the accountant told them the number. If you take that number, issue a pre-approval on it, and let them write an offer, you have a live grenade in your pipeline and roughly three weeks before it goes off.

The reason is not fraud and it is not incompetence. It is a structural conflict that exists in every self-employed file in America, and it is worth stating in one sentence because everything else in this chapter follows from it: a business owner's accountant is paid to minimize taxable income, and an underwriter counts taxable income. Both of them are doing their jobs correctly. The accountant who does not hunt for every legitimate deduction is committing malpractice. The underwriter who counts income the tax return does not report is inventing collateral for somebody else's money. The two professions are pointed in opposite directions by design, and the borrower stands between them holding a number that only one of the two will honor.

You are the only person in the transaction positioned to explain this before it becomes a crisis. The accountant will not, because nobody asked them about a mortgage. The underwriter will not, because underwriters do not talk to borrowers and by the time the file reaches them the offer has been accepted. The real estate agent cannot, because they have no idea this problem exists. That leaves you, on the discovery call, saying a sentence the borrower does not want to hear.

This is the hardest income work in residential lending and it is also, if you learn it, the most durable competitive advantage available to a loan officer. Most of your competitors will quote a rate to a self-employed borrower without ever opening a tax return, discover the problem on day thirty, and blame the underwriter. The borrower will remember which lender told them the truth in the first week. Self-employed households are also disproportionately the households that buy again, refinance, buy investment property, and refer other business owners. A loan officer who can read a K-1 out loud on the phone is not competing on rate.

In this chapter, you will learn to:

• Explain the accountant-underwriter gap to a borrower in ninety seconds, without insulting either party
• Identify the four business structures and the returns and schedules each one produces
• Complete a cash-flow analysis, line by line, and state the qualifying figure it produces
• Reason about an add-back you have never seen using a single principle instead of a lookup table
• Handle declining income correctly, including what an explanation must actually contain
• Frame the business-liquidity question when a borrower wants to fund a down payment from the business
• Run the borrower-and-accountant conversation without ever giving tax advice

Learning Paths

🎓 Exam — §32.2 and §32.5. Know which return each structure files, which produces a K-1, and that a C-corporation's owner receives W-2 wages and dividends rather than pass-through income. 🏠 New LO — §32.1, §32.7, and §32.11. Section 32.1 is the conversation that saves the file; §32.7 is the arithmetic; §32.11 is how you get the documents without a fight. 🤝 Partner — §32.1 and §32.9. An agent who understands why a self-employed pre-approval takes longer will stop pushing you to issue one on a phone call. 📊 Operations — §32.7 and §32.10. These are the two places files sit, and the two places a processor can either save a week or lose one.

32.1 Why self-employed income is different

Start with what a W-2 borrower actually gives you. A paystub and a verification of employment describe a promise: an employer has agreed to pay this person this amount, and unless something changes, will keep doing so. The income is forward-looking by construction. Chapter 11 built the whole apparatus for reading it — base, variable, continuity, the 24-month average — and the reason that apparatus works is that somebody other than the borrower is on the hook for the number.

Now take that away. A self-employed borrower has no employer to make a promise. Nobody has agreed to pay them anything next month. The only evidence that exists about what this household earns is a record of what it already earned, filed with the Internal Revenue Service under penalty of perjury, prepared by a professional whose engagement was to make the resulting tax bill as small as the law allows.

That is the whole of it. Self-employment underwriting is backward-looking because there is nothing else to look at, and the backward-looking record has been deliberately compressed.

The two numbers, and why they never match

Here is the compression, in the plainest terms. A business owner experiences their income as what the business produced minus what the business spent. That is a perfectly reasonable definition and it is roughly how they run their household. The tax return reports something related but different: what the business produced, minus what the business spent, minus a set of deductions the tax code permits that do not correspond to money leaving the business this year, plus or minus a set of timing choices the accountant made on purpose.

Depreciation is the clearest example and the one you will explain a hundred times. A contractor buys a \$60,000 service truck. The cash left the business the day they bought it. But the tax code does not let them deduct \$60,000 of income in the year of purchase and then nothing thereafter — it spreads the deduction across the asset's useful life. So in year three, the return shows a depreciation deduction of some thousands of dollars against income, and not one cent of cash left the business in year three on account of that truck. The return says the business earned less than it did.

That is why the cash-flow analysis exists, and why it adds depreciation back.

But depreciation is the friendly case, because the worksheet fixes it. The unfriendly cases are the deductions that reflect real money genuinely spent: the accountant's suggestion to buy the equipment in December rather than January, to put every legitimate expense through the business, to prepay the insurance, to hire the spouse. Every one of those is legal, sensible, and permanently invisible to an underwriter as anything other than a smaller number.

📞 On the Phone

Borrower: "I clear about ninety-five hundred a month. My accountant told me."

The wrong answer: "Great, that works." You have just accepted a number sourced from a professional who was not asked the question you are asking, and you will be re-explaining this in week four with an accepted contract on the table.

The other wrong answer: "Well, that's not how we look at it." True, useless, and vaguely insulting to a person who has just told you something they believe.

What actually works: "I believe you, and your accountant is right about what you take out of the business. Here's the thing nobody warns business owners about: the number I'm allowed to use comes off your tax returns, and your accountant's entire job for the last two years has been to make that number as small as legally possible. They did a good job. That means the figure I can qualify you on is going to come in under ninety-five hundred, and I'd rather tell you that today than in six weeks. Send me two years of personal returns with every schedule, and two years of the business returns, and I'll have your real number in about a day. Then we go shop."

Three things are happening in that answer. You have validated the accountant — never make a borrower choose between you and their CPA, because you will lose. You have named the mechanism rather than the conclusion, which is what makes it believable. And you have converted a bad feeling into a finite task with a deadline, which is the shape of every good conversation in this book.

The three specific ways this goes wrong

One: the pre-approval issued on a conversation. This is the book's second theme in its purest form — the file is approved when it is documented, not when it is promised — and the self-employed borrower is where loan officers break it most often, because getting tax returns feels like friction and the borrower sounds so confident. Chapter 8 drew the line between pre-qualification and pre-approval. For a self-employed borrower there is effectively no meaningful pre-approval without the returns. You can issue a pre-qualification and label it honestly. You cannot issue a letter that an agent will hand a listing agent as evidence of financing, because you cannot support the income figure inside it.

Two: the calendar. Self-employed files need more documents from more sources, and some of those sources are a third party who has no stake in your closing date. A CPA in early April is not going to prioritize your letter. The Linden Street file took fifty-one days with two W-2 borrowers; a self-employed borrower adds document-gathering at the front, a longer underwriting review in the middle, and a materially higher chance of a condition that requires the accountant. Build that into the timeline you promise the agent, out loud, on day one.

Three: the emotional load. Business owners are used to being the most financially competent person in the room. Being told their income is not what they think it is lands as an accusation even when it is delivered gently. Some of them get angry. Almost all of them get quiet. Respect it. This person took a risk most people do not take, employs other people, and has just been told that the system counts them as smaller than they are. That reaction is reasonable, and the loan officer who treats it as an obstacle to be managed will lose the file and deserve to.

What is not different

One correction, because new loan officers over-learn this chapter. Self-employment does not make a borrower a worse credit risk, and nothing in the agency rulebook treats it as a defect. A documented self-employed borrower with two years of stable returns and reserves is an ordinary conventional file. Chapter 14's eligibility framework applies unchanged: the guidelines are the guidelines, the overlays are the overlays, and layered risk is layered risk. The additional work here is documentary, not moral, and you should never let a borrower hear otherwise.

32.2 The four business structures and their returns

Before you can analyze the income you have to know which pile of paper it lives in. There are four structures in ordinary use, and each one produces a different set of documents. Get this wrong and you will ask for a form that does not exist, which is a fast way to lose a business owner's respect.

The single most useful question on a discovery call is not "how much do you make." It is: "How is the business set up — sole proprietor, partnership, S-corp, or C-corp?" Most owners know the answer instantly. If they do not, ask who prepares the return and whether the business files its own return separate from their personal one. That distinction alone narrows it to two.

WHICH RETURN, AND WHERE THE INCOME LANDS          [constructed teaching example]

  Does the business file its own federal return?
        │
        ├── NO ──► SOLE PROPRIETORSHIP (or single-member LLC by default)
        │            Business activity is reported on SCHEDULE C
        │            inside the owner's personal Form 1040.
        │            Owner's income  = Schedule C net profit
        │            You ask for     = 2 yrs personal 1040, ALL schedules
        │                              + year-to-date P&L
        │
        └── YES ─► which return?
                     │
                     ├── FORM 1065 ──► PARTNERSHIP (or multi-member LLC)
                     │      Pays no income tax itself; income passes through.
                     │      Owner receives a SCHEDULE K-1 (Form 1065),
                     │      reported on Schedule E, Part II of the 1040.
                     │      Owner's income = K-1 ordinary business income
                     │                     + guaranteed payments
                     │      You ask for   = 2 yrs personal + 2 yrs 1065
                     │                      with K-1s, + YTD P&L
                     │
                     ├── FORM 1120-S ─► S-CORPORATION
                     │      Pays no income tax itself; income passes through.
                     │      Owner-employee normally ALSO gets a W-2.
                     │      Owner receives a SCHEDULE K-1 (Form 1120-S),
                     │      reported on Schedule E, Part II of the 1040.
                     │      Owner's income = W-2 wages + K-1 ordinary income
                     │      You ask for   = 2 yrs personal + 2 yrs 1120-S
                     │                      with K-1s and W-2s, + YTD P&L
                     │
                     └── FORM 1120 ────► C-CORPORATION
                            Pays its OWN income tax. Nothing passes through.
                            Owner's income = W-2 wages + DIVIDENDS only
                            You ask for   = 2 yrs personal + (usually)
                                            2 yrs 1120, + YTD P&L

  NOTE: an LLC is a state-law entity, not a tax classification. An LLC may be
  taxed as ANY of the four. Never assume from the name — ask which return it files.

That last note is the one that trips people. "LLC" tells you nothing about the tax treatment. A single-member LLC with no election files a Schedule C. A multi-member LLC with no election files a Form 1065. Either can elect S-corporation treatment and file a Form 1120-S. When a borrower says "I have an LLC," the correct next sentence is "which return does it file?"

Sole proprietorship

The simplest structure and the most common. There is no separate business entity for federal income tax purposes — the business is the person. Business revenue and expenses are reported on Schedule C of the owner's Form 1040, and the bottom line, net profit or loss, flows into the owner's total income. There is no K-1, no separate business return, and often no formal balance sheet.

What that means for you: you can analyze a sole proprietor from the personal returns alone. This is the fastest self-employed file to underwrite and the one where a loan officer can most easily give a borrower a real number the same day. It is also the structure where the borrower is most likely to have no financial statements at all, because nobody ever required them to produce any.

One trap: a sole proprietor pays self-employment tax on the business's net earnings — both halves of Social Security and Medicare. Qualifying income is gross for everybody, and the cash-flow analysis does not deduct it, so a self-employed borrower with \$7,000 a month of qualifying income has meaningfully less spendable cash than a W-2 borrower with \$7,000 a month. The ratios do not see that. It is one of the places where a qualified borrower and an affordable payment are not the same question, and it belongs in your conversation even though it never appears on the worksheet.

Partnership

Two or more owners. The partnership files Form 1065, an information return: the partnership itself generally pays no federal income tax. Instead each partner receives a Schedule K-1 (Form 1065) reporting their distributive share, which they carry onto Schedule E, Part II of their personal return.

The K-1 is a dense one-page document and it is worth learning to read. Among other things it reports the partner's share of ordinary business income or loss, guaranteed payments (compensation to a partner for services, which is real income and counts), the partner's ownership percentage, whether they are a general or limited partner, capital-account activity, and nondeductible expenses.

S-corporation

A corporation that has elected pass-through treatment. It files Form 1120-S and generally pays no federal income tax on its operating income; each shareholder receives a Schedule K-1 (Form 1120-S) and reports it on Schedule E, Part II.

The wrinkle that defines the S-corporation, and the reason it is the structure you will see most often in owner-operated small businesses: an owner who works in the business is required to take reasonable compensation as wages. So the owner is simultaneously an employee and a shareholder, and receives both a W-2 and a K-1 from the same company. Both count. This is the Fulton Avenue file, and it is why §32.7's worksheet starts with a line for W-2 wages paid by the business to the borrower.

C-corporation

A corporation that has not elected pass-through treatment. It files Form 1120 and pays its own federal income tax. Nothing passes through to the shareholder automatically. The owner's income appears on the personal return only as W-2 wages and dividends — and dividends have already been taxed once at the corporate level, which is why small operating businesses rarely choose this structure.

The practical consequence is genuinely counterintuitive: a C-corporation owner can look exactly like a W-2 borrower, right up until somebody notices they own the company. §32.6 works through what happens next.

🎓 NMLS Exam Watch

This section is directly testable and the exam likes it because the distinctions are crisp.

"Which business structure files Form 1065?" The partnership. "Which entity pays federal income tax on its own earnings?" The C-corporation — the other three are pass-through. "A borrower owns 100% of an S-corporation. What income documents should the loan originator expect?" Both a W-2 and a Schedule K-1, plus the Form 1120-S.

The trap in the stem is almost always the LLC. A question that says "the borrower owns an LLC" and offers "Form 1065" as an answer is testing whether you know that an LLC's tax classification is an election, not an entity type. Read for the return, not the name.

A second reliable trap: candidates confuse a distribution with income. A distribution is a transfer of already-taxed earnings out of a pass-through entity. It is not, by itself, qualifying income. §32.5 explains why this one costs real files.

32.3 Schedule C

Schedule C — Profit or Loss From Business (Sole Proprietorship) — is where most self-employed borrowers actually live, and it is the friendliest document in this chapter because everything you need is on two pages of the personal return.

Its structure is a funnel. Gross receipts at the top. Returns and allowances and cost of goods sold come off to produce gross income. Then a long list of expense categories — advertising, car and truck, contract labor, depreciation, insurance, interest, legal and professional, office, rent, repairs, supplies, taxes and licenses, travel, meals, utilities, wages, and a catch-all for other expenses. Those total, and are subtracted, to produce a tentative profit. The business-use-of-home deduction comes off last. The bottom line is net profit or loss, and that figure is the starting point for the cash-flow analysis.

Three habits will make you good at reading these quickly.

Read the expense list before the bottom line. The bottom line tells you the answer; the expense list tells you the story and predicts your add-backs. A large depreciation figure means a big add-back and a strong file. A large "other expenses" figure with no detail means a condition. A "contract labor" figure that jumped 60% in one year means either growth or a change in how the business staffs itself, and the underwriter will want to know which.

Look for the lines that are not there. A landscaping business with \$186,000 of receipts and no insurance expense is either uninsured or paying for it somewhere else. A trucking business with no fuel is not a trucking business. Missing expenses are not always a problem, but they are always a question, and the loan officer who asks it first looks competent.

Compare the two years side by side, category by category. The comparison is where the narrative lives, and §32.9 is entirely about what to do with what you find.

📄 Read the File

```text FIGURE 32.1 — "Two pages of a sole proprietor" [constructed teaching example] THE DOCUMENT Schedule C (Profit or Loss From Business), most recent tax year, attached to a jointly filed Form 1040. Solo residential landscaping and lawn-care business, owner plus two seasonal employees. Six years in operation. THE CONTEXT Discovery call two days ago. The borrower said "the business does a little under two hundred thousand." That is a receipts figure, and it is the first thing on the page. WHAT IT SHOWS Gross receipts $186,400 Cost of goods sold (28,900) mulch, sod, plant material Gross income 157,500 Total expenses (125,200) incl. depreciation $18,700, wages $24,000, contract labor $21,600, meals $1,800 Tentative profit 32,300 Business use of home (3,100) NET PROFIT $29,200

               Run the cash-flow adjustments this document supports:
                  Net profit                          $29,200
                  + Depreciation                       18,700
                  + Business use of home                3,100
                  - Nonrecurring income                      0
                  = Annual cash flow                  $51,000
                  = $51,000 / 12 =                  $4,250.00 per month

WHAT IT DOESN'T It does not show whether the business still exists this morning; a return describes a year that ended months ago. It does not show receivables, debt, or a balance sheet, because Schedule C has none. It does not show the nondeductible portion of meals, which is not reported here at all. It does not show that this is a seasonal business whose cash flow between December and March is near zero. And it does not show a second year — one year is an observation, not a trend. THE DECISION Order the prior year's return today and put a year-to-date profit and loss statement on the document request. Then call the borrower with a range, not a number: "the most recent year supports roughly forty-two fifty a month; I need last year before I can tell you what we actually qualify on, because if last year was higher we may be stuck with an average and if it was much higher we have a conversation." THE LESSON $186,400 and $4,250.00 are both true statements about the same business, and the borrower has been quoting the first one to everybody for six years. Your job is not to correct their sense of themselves. It is to be the first person who tells them which number the mortgage runs on. ```

Constructed for teaching. Work from the borrower's actual return and the current version of the cash-flow worksheet; schedule layouts and adjustment lines are revised.

Notice what happened to the receipts figure. \$186,400 of gross receipts produced \$4,250.00 a month of qualifying income — and the borrower has been describing the business by the first number since the day they started it. That gap is not deception. It is the difference between a business's size and a household's income, and business owners talk about the first because it is the number that measures what they built.

One more Schedule C-specific point, because it will come up. A sole proprietor who uses the standard mileage rate rather than actual vehicle expenses has taken a deduction that includes an embedded depreciation component, and that component is generally addable back — the IRS publishes the per-mile depreciation figure annually, and it changes. If your borrower drove 28,000 business miles, this line is worth real money. Get the current per-mile figure from the source, and get the mileage from the return, and never estimate either.

32.4 Partnerships and Form 1065

A partnership file requires two return sets: the borrower's personal returns and the partnership's Form 1065 with the borrower's Schedule K-1 attached. You cannot do this analysis from the personal return alone, because the add-backs live on the business return.

What the K-1 tells you

The Schedule K-1 (Form 1065) is the bridge between the two returns, and it answers four questions that determine how much work the file is going to be.

How much did this partner earn? Ordinary business income or loss is the partner's distributive share of the partnership's operating result. Guaranteed payments are compensation for services or for the use of capital, paid regardless of profitability — economically much closer to a salary, and treated as income.

What share do they own? The K-1 reports the partner's profit, loss, and capital percentages. This matters twice. First, because add-backs taken from the partnership's return must be prorated to the borrower's share — if the partnership deducted \$40,000 of depreciation and your borrower owns 40%, the add-back is \$16,000, not \$40,000. Second, because ownership percentage generally determines whether a borrower is treated as self-employed at all. The common threshold across agency guidelines is 25% or more ownership — below that, a borrower receiving W-2 wages from the company is often documented as an ordinary wage earner. Verify the current definition in the applicable guide; this is precisely the kind of value that gets revised.

Are they a general or a limited partner? A limited partner is an investor, not an operator. Their share of income may be treated differently, and the question of whether they can actually reach the cash is a live one.

Did they take out more than they earned? The capital account section shows contributions, distributions, and the partner's ending capital. A partner drawing \$180,000 from a partnership that earned \$70,000 is consuming capital, and that is a fact about sustainability, not a source of extra income.

Where the numbers go

The partner reports the K-1 on Schedule E, Part II of the personal return. On the cash-flow worksheet, the partnership section typically pulls ordinary business income and guaranteed payments from the K-1, then takes prorated add-backs from the Form 1065 itself: depreciation, depletion, amortization, and the partnership's nondeductible expenses. Some worksheets also subtract the partner's share of mortgages or notes payable in less than one year, on the reasoning that short-term business debt is a cash demand the business must meet within twelve months. That subtraction usually has a documented exception path — evidence that the obligation rolls, or that the business holds sufficient assets to retire it — and it is worth asking about before you accept a reduced number.

What to ask for, in one message

Send this as a single list on the first call, not in five follow-ups:

• Two years of personal federal returns, all pages, all schedules, signed
• Two years of Form 1065 with all schedules and the borrower's K-1 for each year
• A year-to-date profit and loss statement for the partnership
• If the borrower's ownership is below 100%, a statement of ownership percentage and, where the file needs business funds, the other partners' position on distributions

⚖️ Compliance Check

Three things worth being careful about in this section of the file.

Tax transcripts. Chapter 11 covers Form 4506-C and the tax transcript process, and it applies here with more force than anywhere else in origination: for a self-employed borrower the returns are the income documentation, and a lender that does not validate them against the IRS record is relying on a document the borrower could have produced on a laptop. Expect transcripts on business returns as well as personal ones, and expect timing problems near filing deadlines and around extensions.

Amended returns. A borrower who amends a prior-year return after learning what the underwriter counts has created a serious problem, and a loan officer who suggests it has created a much more serious one. Amended returns invite scrutiny, take months to appear in the transcript record, and — if the amendment was made to qualify for a mortgage rather than to correct a genuine error — sit uncomfortably close to loan fraud. Chapter 27 covers the exposure. Do not go near this. If a borrower raises it, your answer is that any amendment is between them and their tax professional for tax reasons, and that you cannot advise on it.

Evenhandedness. Equal Credit Opportunity Act and Regulation B obligations apply to how you treat applicants, and self-employment is not a protected class — but the practical risk is discouragement. Telling a business owner "self-employed loans are really hard, you probably won't qualify" before taking an application is exactly the behavior Regulation B's discouragement provisions are aimed at. Say what the documentation requires. Do not editorialize about their chances before you have looked.

Requirements change and state law varies. Verify current rules with your compliance department and your regulator.

32.5 S-corporations and Form 1120-S

This is the structure you will see most, in the trades and in professional services, because it lets an owner take part of their income as wages and part as a distribution of profit. It is also where the most expensive misunderstandings in this chapter live.

The S-corporation files Form 1120-S. Operating income generally passes through to shareholders on Schedule K-1 (Form 1120-S), reported on Schedule E, Part II. The owner-employee also takes reasonable compensation as W-2 wages, which appear on the personal return like any other job.

So an S-corporation owner hands you a W-2, a K-1, a personal return, and a business return, and every one of the four matters.

Both halves count, and neither is optional

The Fulton Avenue file is an S-corporation — a six-employee residential HVAC company — and its worksheet shows exactly this shape:

Miss either half and the file is wrong in an obvious direction. A loan officer who takes only the W-2 has understated this borrower by \$1,800 a month. A loan officer who takes only the K-1 has understated them by nearly \$5,900 a month. Both mistakes get made, constantly, usually by someone who did not realize the borrower owned the company that issued the W-2.

That is the first discipline of this section: when a W-2 shows up, ask whether the borrower owns any part of the employer. It costs one sentence on the application call. The paystub does not say "the person receiving this owns the company," and the automated underwriting findings will not either — but the underwriter will notice, on day twenty-six, and will condition for two years of business returns you have not requested.

Distributions are not income

Here is the mistake that kills files.

A borrower says, "I take twelve thousand a month out of the business." They are describing distributions — transfers of the corporation's already-taxed earnings to its shareholder. The money is real, it lands in their account, and it feels exactly like a paycheck. It is not qualifying income, and no amount of bank statements showing it will make it so.

Why not? Because a distribution is not a measure of what the business earned; it is a measure of what the owner withdrew. A corporation that earned \$40,000 can distribute \$150,000 by drawing down cash it accumulated in better years, or by borrowing. Counting distributions as income would mean counting the same dollar twice — once when the business earned it and again when it left — and would let a business fund a mortgage qualification out of its own balance sheet until the balance sheet was empty.

What the underwriter counts is the K-1's ordinary business income plus the W-2, adjusted. What the distributions tell you is something different and also useful: whether the owner is living off earnings or off capital.

⚠️ Where Deals Die

The borrower who "takes twelve thousand a month" and qualifies at seven.

The mechanism is always the same. The owner has been running distributions at a comfortable level for years, funded partly by current earnings and partly by cash the business built up. They quote that number to you. You issue a letter. The returns arrive on day nineteen and the K-1 plus W-2, after add-backs, supports something far smaller. The file is short by thousands of dollars a month of income and there is no fix, because the fix would require a different tax return.

Two things prevent it, both free:

Ask the question in the borrower's own words. Not "what's your income" — say "when you say you take twelve thousand out, is that a paycheck with taxes withheld, or is it a transfer from the business account?" Business owners answer that question accurately and instantly. A paycheck is a W-2. A transfer is a distribution. You have just learned which half of the file you are missing.

Get the returns before the letter. Every single time, without exception, for every self-employed borrower. There is no version of this job where a letter issued on a conversation is worth the four days it saves.

There is also a warning-sign version of the same fact pattern. If the distributions substantially and repeatedly exceed the business's earnings, the underwriter is looking at a company consuming its own capital, and that is a continuity question about the income itself — not just an arithmetic problem. Chapter 14's layered-risk framing applies: a declining business, thin reserves, and a high ratio are not three separate issues.

Losses, and what a K-1 loss does

A K-1 can report a loss, and a loss does not politely sit in the business — it reduces the borrower's qualifying income. An S-corporation owner with \$84,000 of W-2 wages and a \$31,000 K-1 loss does not have \$84,000 of income for underwriting purposes. This surprises borrowers badly, because the wages are real, they were paid, and they were spent.

The reasoning is the one the whole chapter runs on: the borrower owns the business, and if the business is losing money, that loss is a claim on the household. There are documented exception paths where a loss is demonstrably non-recurring or non-cash, and they are worth pursuing — but the starting position is that the loss counts against you.

The rest of the 1120-S

Two schedules on the business return matter beyond the income lines.

Schedule L is the balance sheet: what the corporation owns and owes at year end. It is the source for the liquidity analysis in §32.10 and it is often the fastest read in the file — cash, receivables, payables, and the loans. Note that smaller corporations and partnerships are excused from completing the balance sheet if they fall below a receipts-and-assets threshold, which means the balance sheet you need is sometimes simply absent. Verify the current threshold; when it bites, you need a CPA-prepared balance sheet instead.

Schedules M-1 and M-2 reconcile book income to tax income and track the accumulated adjustments account. You will not usually work them line by line, but a large M-1 reconciliation is a signal that book and tax income diverge meaningfully, and that is worth understanding before you promise anything.

32.6 C-corporations

The C-corporation is the least common structure among owner-operated small businesses and the one most likely to be missed entirely.

Mechanically it is simple. The corporation files Form 1120 and pays its own federal income tax on its earnings. Nothing passes through. The shareholder's personal return therefore shows only two things attributable to the company: W-2 wages, if they work there, and dividends, if the corporation declared any. Retained earnings — profits the corporation kept — are the corporation's money, not the shareholder's, and they are not qualifying income no matter how completely the borrower controls the company.

This is the double-taxation problem that drove most small businesses to elect S-corporation status: the corporation pays tax on its profit, and then the shareholder pays tax again on any dividend paid out of that profit.

Why it looks easy and is not

A C-corporation owner brings you a paystub and a W-2. Every instinct says wage earner, and Chapter 11 has a well-worn path for wage earners. Follow it and you may be fine — or you may be looking at a file where the borrower owns 100% of the employer, the wages are set at whatever the owner decided to set them at last January, and the underwriter is going to require the corporate returns anyway because ownership crosses the self-employment threshold.

The questions that separate the two cases:

Does the borrower own 25% or more of the company? (Verify the current threshold, but this is the usual line.) If yes, treat this as a self-employed file and get the business returns, even though the income arrives as wages.

Are the wages stable and supportable? An owner who paid themselves \$96,000 and then \$102,000 looks like a person with a raise. An owner who paid themselves \$60,000, then \$180,000, then \$70,000 looks like a person tuning a tax outcome, and the underwriter will average or take the lower figure. The wages are only as reliable as the business behind them.

Are the dividends usable? Dividends are treated like other investment income: they generally require a two-year history and evidence of likely continuance. A dividend that appeared once will not count. A dividend that is declining will, at best, be counted at the lower figure. And a dividend that exists only because the accountant chose to declare one this year is a scheduling decision, not an income stream.

Is the corporation healthy enough to keep paying? Some worksheets include a section for analyzing corporate cash flow where the borrower owns the whole company, on the theory that a sole shareholder can direct the corporation's cash. Use it with care and with the balance sheet in front of you. The corporation is a separate taxpayer and a separate legal person, and treating its cash as the borrower's requires evidence, not logic.

That last row deserves a sentence. When a C-corporation transfers money to its owner without calling it wages or a dividend, it is often booked as a loan to the shareholder. That is a liability, not income. If it carries a repayment obligation, it may belong in the debt column.

32.7 The Cash Flow Analysis worksheet, line by line

Everything above has been preparation for one document. Fannie Mae publishes Form 1084, Cash Flow Analysis, and a companion, Form 1088, Comparative Income Analysis; Freddie Mac publishes Form 91. These are real, free, published worksheets, and they are revised. Work from the current version. What follows describes their structure and logic, which is stable, rather than their exact line labels and ordering, which are not.

What the worksheet is for

The worksheet answers one question: given these returns, what monthly figure may be counted as this borrower's stable qualifying income? It does this by walking the personal return and, where applicable, the business return, converting reported taxable income into an estimate of the cash the business actually generated and the borrower can actually reach.

It has three parts.

THE SHAPE OF A CASH FLOW ANALYSIS                  [structure, not a form reproduction]

  PART 1 — THE PERSONAL RETURN (Form 1040 and its schedules)
     W-2 wages from the borrower's own business
     Schedule B  — interest and dividends (recurring only)
     Schedule C  — sole proprietorship, with its adjustments
     Schedule D  — capital gains (recurring only; usually excluded)
     Schedule E  — rentals; and Part II, the K-1 flow-through
     Schedule F  — farm income, with its adjustments
                                  │
  PART 2 — THE BUSINESS RETURN (1065 / 1120-S / 1120)
     ordinary business income (from the K-1, at the borrower's share)
     + guaranteed payments (partnerships)
     + depreciation                     ─┐
     + depletion                         │  prorated to the
     + amortization / casualty / one-time│  borrower's ownership
     - meals and entertainment exclusion │  percentage
     - nonrecurring income               │
     - notes payable in under 12 months ─┘  (documented exceptions exist)
                                  │
  PART 3 — THE AVERAGING DECISION
     Year 1 total ──┐
                    ├──► 24-month average, or most recent year alone,
     Year 2 total ──┘    or the LOWER of the two if income declined
                                  │
                                  ▼
                    QUALIFYING MONTHLY INCOME

The order matters less than the discipline: you go through the returns once, systematically, and you do not skip a schedule because you assume it is empty. Schedule E in particular hides things — a rental property, a second K-1 from a business the borrower forgot to mention, a partnership interest inherited from a parent.

The Fulton Avenue worksheet

Here it is, completed. This is a frozen file; the figures below are the file's canonical numbers.

Check the footing, because you should always check the footing:

• Year 1: \$62,000 + \$38,400 + \$14,200 − \$2,100 − \$3,000 = **\$109,500** ✓
• Year 2: \$71,000 + \$21,600 + \$16,800 − \$2,400 − \$0 = **\$107,000** ✓

And the three monthly figures the worksheet produces:

• Year 1 alone: \$109,500 ÷ 12 = **\$9,125.00**
• Year 2 alone: \$107,000 ÷ 12 = **\$8,916.67**
• Both years: (\$109,500 + \$107,000) ÷ 24 = \$216,500 ÷ 24 = **\$9,020.83**

Income declined from year 1 to year 2 — \$2,500, or 2.3% — so the underwriter uses the lower figure, \$8,916.67. Section 32.9 works through why, and what it costs.

📄 Read the File

```text FIGURE 32.2 — "The number the borrower will not recognize" [the Fulton Avenue file] THE DOCUMENT Completed cash-flow analysis worksheet, two tax years, prepared by the loan officer from the borrower's personal Form 1040s and the corporation's Form 1120-S with K-1s. Six-employee residential HVAC S-corporation; borrower is the sole shareholder and an employee. THE CONTEXT Day 3 of the file. The borrower told you on the discovery call that they "make about $9,500 a month," which is what their accountant told them, and they have a house in mind. WHAT IT SHOWS YEAR 1 YEAR 2 W-2 wages to self $62,000 $71,000 K-1 ordinary business income 38,400 21,600 + Depreciation 14,200 16,800 - Meals and entertainment excl. (2,100) (2,400) - Nonrecurring other income (3,000) 0 ───────────────────────────────────────────────────────── TOTAL $109,500 $107,000 ($9,125.00) ($8,916.67)

               24-month average .................. $9,020.83
               Most recent year alone ............ $8,916.67
               Change, year 1 to year 2 .......... -$2,500  (-2.3%)
               QUALIFYING INCOME USED ............ $8,916.67   (the lower)

WHAT IT DOESN'T It does not show why the K-1 fell $16,800 — that story is in the business return and in the borrower's head, and it turns out to matter. It does not show whether the business is still healthy today; the most recent figure describes a year that has already ended. It does not show the balance sheet, the receivables, or whether the company could survive a $40,000 withdrawal. It does not show self-employment or corporate tax paid, which is real money the household no longer has. And it does not show a third year, which would tell you whether a 2.3% decline is a wobble or a trend. THE DECISION Call the borrower today, before they see a house they cannot buy. Give them $8,916.67, explain the two averaging figures and why the lower one applies, and ask the one question the worksheet raised: what happened to the K-1? Then request the year-to-date profit and loss statement, because if the current year is running ahead, that is the evidence that turns a decline into a story. THE LESSON The borrower is not wrong and the accountant is not wrong. The accountant answered "what does this household have to live on," which is roughly $9,500. The worksheet answers "what may an investor be told this household earns," which is $8,916.67. Those are different questions, and the loan officer is the only person in the transaction who knows both of them are being asked. ```

The Fulton Avenue file is constructed for teaching. Form 1084 is a real, published Fannie Mae worksheet and is revised; work from the current version and your lender's instructions.

Reading the worksheet as a narrative

Compare the two columns and something interesting appears. Between year 1 and year 2:

Verify: \$9,000 − \$16,800 + \$2,600 − \$300 + \$3,000 = **−\$2,500** ✓, which matches \$109,500 − \$107,000 exactly.

That table is not arithmetic for its own sake. It is the question you take to the borrower. The owner gave themselves a \$9,000 raise and the corporation's profit fell \$16,800 — which is what happens when an owner shifts compensation from profit to wages, or when the business spends money. The depreciation add-back rose \$2,600, which is the fingerprint of an asset purchased: on this file, a new service truck. Part of the "decline" is an investment.

That is the sentence that makes a 2.3% decline acceptable to an underwriter, and you cannot write it unless you did the comparison.

32.8 Add-backs and what is not one

New loan officers try to memorize the add-back list. It does not work, because the list is long, it varies by form and by structure, and sooner or later a return will show you a line nobody put on any list. What works is a principle, and there is exactly one.

Add back what was deducted and did not leave the business. Subtract what left the business and was not deducted.

Every line on the worksheet is an application of that sentence. The goal of the whole exercise is to convert taxable income into cash the business actually generated, and the two differ only where a deduction and a cash outflow failed to coincide.

THE ADD-BACK TEST — one question asked twice

                                     DID THE MONEY LEAVE THE BUSINESS?

                              ┌────────────────────┬────────────────────┐
                              │         NO         │        YES         │
   ┌──────────────────────────┼────────────────────┼────────────────────┤
   │                          │  ADD IT BACK       │  LEAVE IT ALONE    │
   │  WAS IT      YES         │    depreciation    │    rent, wages,    │
   │  DEDUCTED                │    depletion       │    fuel, insurance │
   │  ON THE                  │    amortization    │    a vehicle lease │
   │  RETURN?   ──────────────┼────────────────────┼────────────────────┤
   │                          │  DO NOTHING        │  SUBTRACT IT       │
   │              NO          │    it never        │    the nondeduct-  │
   │                          │    touched the     │    ible portion    │
   │                          │    number          │    of meals        │
   └──────────────────────────┴────────────────────┴────────────────────┘

The add-backs, and why each one qualifies

Depreciation. The allocation of an asset's cost across its useful life. The cash left in the year of purchase; the deduction arrives in slices for years afterward. Deducted, no cash out this year: add back. This is almost always the largest add-back in a file and it is the reason equipment-heavy businesses — trucking, construction, medical practices, farming — qualify far better than their returns suggest.

Depletion. The same idea applied to a natural resource being consumed: oil, gas, timber, minerals. A deduction for the exhaustion of an asset, not a payment. Add back.

Amortization. The write-off of an intangible asset — goodwill from a business purchase, a covenant not to compete, organizational or startup costs. Deducted over years, no cash out in the year of deduction: add back. Be careful here, because "amortization" also means paying down a loan, and the two have nothing to do with each other. On a tax return it means the intangible. In Chapter 4 it meant the loan schedule. Same word, different universes.

Casualty loss and one-time expenses. A book loss from a fire or a storm is not the year's operating cash flow, and a genuinely non-recurring expense will not repeat. Add back, with documentation that it was in fact non-recurring.

Business use of home (Schedule C). An allocation of household costs — a share of the utilities, the insurance, the depreciation on the residence — assigned to the business. Much of it is money the household spends whether or not the business exists. The worksheet has a line for it: add back.

Nonrecurring loss. Real cash left, but it will not leave again. Add back — the mirror image of nonrecurring income.

The subtractions, and why each one qualifies

Meals and entertainment exclusion. The tax code allows only a portion of business meals to be deducted. The business paid for all of them. So the nondeductible portion is cash that left and was never deducted — taxable income overstates available cash by exactly that amount. Subtract it.

A practical note that will save you an argument: on a partnership or S-corporation return, the nondeductible portion is reported as a nondeductible-expense item, which is where the worksheet's figure comes from — on Fulton Avenue, \$2,100 and \$2,400. On a Schedule C it generally does not appear at all, because a sole proprietor reports only the deductible amount. Different forms, different availability of the same fact. This is one of several reasons you work from the current worksheet rather than from memory. Note also that Congress has changed the deductible share of business meals more than once; the current percentage is a Tier-2 figure to verify, but the logic of the exclusion does not change.

Nonrecurring income. Money that arrived and will not arrive again: a legal settlement, an insurance recovery, the sale of a piece of equipment, a one-time contract. Real cash, correctly reported, and useless for predicting next year. Subtract it. Fulton Avenue's year 1 carried \$3,000 of it.

Notes payable in less than twelve months. Short-term business debt is a claim on next year's cash. Worksheets commonly subtract it, with a documented exception path where the borrower shows the obligation rolls or that liquid assets cover it. Ask before you accept the reduction.

What is not an add-back, and the reason people get it wrong

A vehicle lease payment. The business writes a check every month and deducts it. Money left, and it was deducted. Both halves of the test are satisfied on the "leave it alone" side. Nothing to do.

Borrowers argue about this one, and their argument is sincere: "but that's a business expense, it's not really my money going out." The answer is that it is money going out — that is precisely why it is not an add-back. Depreciation gets added back because the cash left in a different year. A lease payment leaves this year, every year, and will leave next year.

Rent, wages, fuel, insurance premiums, professional fees, supplies. All ordinary operating expenses. All deducted. All paid in cash. Nothing to do.

Owner health insurance. Cash left the business, and it was deducted. Nothing to do — although borrowers reasonably point out that a W-2 employee's premium comes out of gross pay too, so the comparison is fairer than it feels.

Distributions. Not a deduction at all. A distribution is not an expense of the business; it is a transfer of after-tax earnings. It never reduced the number you started with, so there is nothing to add back, and it is not itself income (see §32.5).

The two hard cases

Section 179 expensing. The tax code permits certain asset purchases to be deducted entirely in the year of purchase rather than depreciated over years. The worksheet generally treats it as depreciation and adds it back — and here the principle strains, because the cash really did leave the business in that year to buy the equipment. A borrower who expensed \$90,000 of trucks under Section 179 gets an add-back that hands them qualifying income the business did not have in cash. The worksheet's answer and the business's reality diverge. A careful underwriter will look at the balance sheet and the business bank statements before getting comfortable, and a careful loan officer will mention it before the underwriter does.

Anything you are about to add back for the second time. This is the most common worksheet error in practice, and the discipline that prevents it is one question: before you add something back, find out whether it is already inside the number you started with. Where a deduction sits — on the business return or on the personal return — determines whether it has already reduced your starting figure. Depreciation reported on the Form 1120-S has already reduced the K-1's ordinary business income, so adding it back is correct. A deduction taken on the personal return's adjustments has not touched the K-1 at all, so adding it back would be inventing money. Trace the line to its source before you touch it.

🧮 Run the Numbers

What one line is worth.

Take Fulton Avenue's year 2 and delete a single entry — depreciation:

	With the add-back	Without it
W-2 wages to self	\$71,000 | \$71,000
K-1 ordinary business income	\$21,600 | \$21,600
+ Depreciation	\$16,800	—
− Meals and entertainment exclusion	(\$2,400) | (\$2,400)
Annual total	\$107,000** | **\$90,200
Monthly	\$8,916.67** | **\$7,516.67

The difference is \$16,800 ÷ 12 = **\$1,400.00 a month**, which is also \$8,916.67 − \$7,516.67.

Now price it. Suppose the file is being run to a 45% back-end ratio (agency automated underwriting has in recent years accepted ratios in this range and higher — verify the current limit and your lender's overlays). That \$1,400.00 of income supports:

$$\$1{,}400.00 \times 0.45 = \$630.00 \text{ per month of additional obligation}$$

At 6.625% over thirty years, each dollar borrowed costs \$0.00640313 a month in principal and interest. So \$630.00 a month buys:

$$\$630.00 \div \$0.00640313 = \$98{,}389$$

One line on a worksheet is worth roughly \$98,000 of loan amount on this file — and that is the generous version, assuming every dollar of capacity goes to principal and interest rather than to taxes, insurance, and mortgage insurance. In reality the house-price effect is smaller and still enormous.

Across both years the add-back is (\$14,200 + \$16,800) ÷ 24 = \$1,291.67 a month.

That is why you never eyeball a self-employed file. A loan officer who "roughly figures" a business owner's income is roughly figuring a six-figure difference in what they can buy. (Rate illustrative; run current pricing.)

32.9 Declining income

Two years of returns give you two totals, and the relationship between them is a decision, not a detail.

The rule, and the asymmetry inside it

If the most recent year is equal to or higher than the prior year, income is stable or rising, and the conservative convention applies: average the two years. That convention was built in Chapter 11 and it applies identically here, including its familiar sting — a borrower with genuinely rising income is qualified on a figure lower than what they now earn.

If the most recent year is lower than the prior year, you do not get to average. Averaging a declining stream produces a number higher than the most recent evidence supports, and the underwriter will not stand behind it. The starting position is the lower figure — normally the most recent year alone — and, if the decline is severe enough, the question stops being which figure and becomes whether any figure is supportable.

The asymmetry is deliberate and it is worth naming for the borrower, because it feels unfair and the reason is straightforward. Somebody else's money is at risk for thirty years, and the most recent observation is the best available evidence about next year. When it points down, the conservative reading is the honest one.

THE DECLINING-INCOME LADDER                        [structure, not a guideline table]

  Year 2 >= Year 1        ──► average the two years (the conservative convention)
                              ask: is a rising trend being penalized? often, yes

  Year 2 < Year 1, small  ──► use the most recent year alone
   (a few percent)            + a written explanation
                              + year-to-date P&L showing the current year holding

  Year 2 < Year 1,        ──► most recent year alone, and expect scrutiny
   material                   + explanation that names a CAUSE, not a feeling
                              + YTD P&L showing stabilization or recovery
                              + business bank statements
                              + possibly a third year of returns

  Year 2 << Year 1,       ──► the agency path may not work at all
   severe or accelerating     do not spend six weeks proving otherwise
                              Chapter 34 covers what exists outside agency
                              documentation, and this is where you point

Where the exact boundaries sit between "small," "material," and "severe" is a matter of guideline, overlay, and underwriter judgment, and it is not a number this book will invent for you. The Selling Guide is the authority and it is updated continuously.

What an explanation has to contain

Borrowers write bad explanation letters because nobody tells them what a good one is. A good one answers three questions in about four sentences: what happened, why it will not keep happening, and what evidence supports that.

Explanations that work, because they name a cause with a shape:

• A large equipment purchase in the most recent year, with the invoice
• The loss of a single large customer, since replaced, with the new contract
• An owner's medical leave, with dates
• A deliberate strategic change — dropping a low-margin service line — with current-year margins
• A one-time expense: a lawsuit settled, a facility moved, a fleet re-roofed

Explanations that do not work, because they explain nothing:

• "Business is fine, my accountant just took more deductions." Possibly true and completely unusable; the underwriter counts the deductions, not the intent behind them.
• "It was a slow year." Which is a description of the decline, not a cause of it.
• "The economy." Unless the borrower can point at something specific to their market and their book of business.

The evidence that carries the most weight is the year-to-date profit and loss statement — an interim statement of revenue and expenses for the current year, prepared by the borrower or their accountant. A P&L showing the current year running at or above the prior year turns a decline into a dip. Get one on every declining file, early, and read it before you send it.

🧮 Run the Numbers

What the accountant's number costs on the Fulton Avenue file.

Three figures are in play, and the borrower has only heard one of them:

Figure	Source	Monthly
What the accountant said	cash the household lives on	\$9,500.00
24-month average	(\$109,500 + \$107,000) ÷ 24	\$9,020.83
Most recent year alone	\$107,000 ÷ 12 | \$8,916.67
What qualifies them	the lower, because income declined	\$8,916.67

The decline itself. \$109,500 − \$107,000 = \$2,500, and \$2,500 ÷ \$109,500 = 2.3%. Small — and the rule does not care how small. Averaging up is not available.

What the averaging rule costs. \$9,020.83 − \$8,916.67 = about \$104 a month of qualifying income, purely from the direction of the trend. At a 45% back-end that is roughly \$47 a month of payment capacity. Real, but not the headline.

What the accountant's number costs. This is the headline:

$$\$9{,}500.00 - \$8{,}916.67 = \$583.33 \text{ per month}$$

Annually, that is \$114,000 versus \$107,000 — a gap of exactly \$7,000 a year, or 6.14% of what the borrower believes they make. Priced at a 45% back-end ratio:

$$\$583.33 \times 0.45 = \$262.50 \text{ per month of obligation capacity}$$

$$\$262.50 \div \$0.00640313 = \$40{,}996$$

About \$41,000 of house, on the generous assumption that every dollar goes to principal and interest. That is the number the borrower discovers on day thirty if you did not tell them on day three: the difference between the house they toured and the house they can buy.

And the story that saves it. Recall §32.7's comparison. The K-1 fell \$16,800 while W-2 wages rose \$9,000 and the depreciation add-back rose \$2,600 — a new service truck. The decline is 2.3%, it has a cause, the cause is an investment in the business, and the year-to-date P&L can be asked to show whether the truck is earning. That paragraph, written by the borrower and supported by an invoice, is worth more to this file than any argument about the arithmetic. (Rate and ratio illustrative; run current pricing and verify current guidelines.)

The trap on the other side

One more, because it catches people who have learned this section well. A rising trend can also raise a question. Income that jumps 60% in one year is not automatically good news — an underwriter sees a spike that may not repeat, and may average it or ask what caused it. The instinct that "declining is bad, rising is fine" is only half right. What underwriting wants is predictable, and a large move in either direction is less predictable than a flat line.

32.10 Business liquidity and the funds-from-the-business question

Sooner or later a business owner says the sentence: "I'll just take the down payment out of the business."

They can, sometimes. It is not automatic, and it is not a documentation problem — it is an underwriting question with a real answer that depends on the business.

Two things have to be true. First, the borrower has to be an owner of the business, and where there are other owners, the file needs to deal with them. Second — and this is the part that surprises people — the lender generally has to confirm that withdrawing the money will not have a negative impact on the business. Agency guidelines require this analysis; the Selling Guide is the authority and is updated continuously, so verify the current requirement and your lender's overlays before you promise a borrower anything.

Note the logic, because it is not arbitrary. The business is the source of the borrower's income. If taking \$40,000 out of it cripples the company, the lender has just funded a down payment by damaging the thing that repays the loan. Chapter 12 owns asset sourcing and seasoning, and every one of those requirements still applies — business funds must be verified, sourced, and seasoned like any other asset. What §32.10 adds is a question Chapter 12 does not ask: can the business afford to lose this money?

How you actually answer it

There is no formula that settles it, and any loan officer who tells you there is a magic ratio is selling something. There are two ratios that structure the question, both computed from the business's balance sheet — Schedule L on the business return, or a CPA-prepared statement where Schedule L was not required.

Current ratio = current assets ÷ current liabilities. Above 1.0 means the business's short-term assets cover its short-term obligations.

Quick ratio = (current assets − inventory) ÷ current liabilities. The same test with inventory removed, on the theory that inventory may not convert to cash quickly.

Both are snapshots on one date. Neither is the answer. The answer is a sentence, and here it is:

After this money leaves, can the business meet its next payroll, its next tax deposit, and its next equipment payment, without borrowing?

Everything else is a way of getting to that sentence with evidence.

🧮 Run the Numbers

The \$40,000 withdrawal, on the Fulton Avenue file.

The borrower wants to take \$40,000 out of the S-corporation for a down payment. The corporation's year-end balance sheet shows [constructed teaching example]:

Current assets		Current liabilities
Cash	\$48,000 | Accounts payable | \$37,000
Accounts receivable	\$61,000 | Accrued payroll | \$18,000
Inventory (parts, materials)	\$23,000 | Current portion of equipment notes | \$19,000
Total	\$132,000** | **Total** | **\$74,000

Before the withdrawal:

• Current ratio: \$132,000 ÷ \$74,000 = 1.78
• Quick ratio: (\$132,000 − \$23,000) ÷ \$74,000 = \$109,000 ÷ \$74,000 = 1.47

Comfortable. Short-term assets cover short-term obligations nearly twice over.

**After a \$40,000 withdrawal**, current assets fall to \$92,000:

• Current ratio: \$92,000 ÷ \$74,000 = 1.24
• Quick ratio: (\$92,000 − \$23,000) ÷ \$74,000 = \$69,000 ÷ \$74,000 = 0.93

Both still positive. The current ratio still exceeds 1.0. On the ratios alone, a reasonable underwriter might approve this.

Now ask the sentence. Cash — not current assets, cash — falls from \$48,000 to **\$8,000**. This company runs six employees at roughly \$5,200 a month each fully loaded:

$$6 \times \$5{,}200 = \$31{,}200 \text{ per month of payroll}$$

$$\$31{,}200 \div 30 = \$1{,}040 \text{ per day}$$

$$\$8{,}000 \div \$1{,}040 = 7.7 \text{ days}$$

The business has about a week of payroll in the bank, and everything after that depends on \$61,000 of receivables arriving on time. If a large customer pays in sixty days instead of thirty, this company misses payroll — and the income that qualifies this borrower is the income of a company that just missed payroll.

The quick ratio at 0.93 was quietly telling you the same thing. Below 1.0 means that without selling inventory, short-term assets no longer cover short-term obligations.

What a good loan officer does with this: do not decline it and do not promise it. Go back with a smaller number. Ask what \$20,000 does instead of \$40,000, ask when the receivables actually land, and ask whether the seasonal shape of an HVAC business means the December balance sheet is the best month or the worst. Then get the borrower's accountant on the phone. (All figures constructed for teaching.)

The four things that decide it in practice

Seasonality. A December 31 balance sheet is one day of a business's year. An HVAC company in July and the same company in April are different companies financially. Ask when the balance sheet was taken and what the trough month looks like.

Receivables quality. \$61,000 of receivables is an asset if customers pay and a fiction if they do not. A business whose receivables are ninety days old has a collections problem disguised as a current asset.

Whether the withdrawal is actually a loan. If the corporation books the transfer as a loan to the shareholder rather than a distribution, the borrower now has a debt with a repayment obligation, and it may belong in the debt column of the ratio you just improved.

Whether there is a simpler path. Business funds are the hardest asset in residential lending to document. Before you spend three weeks on a liquidity analysis, ask what else exists: personal savings, a gift, a retirement account, a smaller down payment with mortgage insurance. Sometimes the right structural answer is to leave the business's money in the business, and Chapter 13's comparison framework is how you show the borrower the trade.

32.11 The conversation with the borrower and their accountant

Everything in this chapter is arithmetic except the part that actually decides files.

The first call

Say the number early and say it as a mechanism, not as a verdict.

📞 On the Phone

The verdict version, which fails: "So, running your returns, you only qualify for about eighty-nine hundred." Accurate, and it lands as a judgment on the borrower's life's work delivered by someone who has known them for four days. Expect silence, then an argument about a number that is not arguable.

The deflection version, which also fails: "It's what the underwriting guidelines require." True, unfalsifiable, and it tells the borrower there is a wall somewhere that neither of you can see. They will go call another lender to find out whether the wall is real.

What actually works:

"Here is what I am going to do. I am going to take your two years of returns and run the exact same worksheet your underwriter will run. It adds back things like depreciation — a deduction where no money actually left your business — and it takes out things like the nondeductible half of your meals, where money left and you never got to deduct it. The number that comes out is almost always lower than what you take home, because your accountant spent two years making it lower on purpose and they were right to. I'll have it for you in about a day, and then we'll know what we're shopping for."

Then, when you call back with \$8,916.67: "Here's the worksheet. Here's the line that moved. Here's what it means for price range. What questions do you have?"

Three things are working in that version. It never asks the borrower to distrust their accountant. It never asks them to accept a number on your authority — it describes a procedure they could check themselves, which is exactly what business owners respond to. And it puts the number in their hands with the arithmetic attached, so the conversation is about a document instead of about their competence.

What to ask for, and when

Ask once, in a single list, at the beginning:

• Two years of personal federal returns, all pages, all schedules, signed
• Two years of business returns with all schedules and K-1s, where the structure files one
• A year-to-date profit and loss statement (and a balance sheet, if the business has one)
• Two to three months of business bank statements, on files where they will matter
• Form 4506-C, as Chapter 11 describes, for the tax transcript validation

The single most common process failure in self-employed lending is asking for these in five separate messages over three weeks. Each request resets the borrower's clock and costs a day. Send the list on day one and follow up on the list, not on individual items.

The accountant letter — what it can and cannot do

An accountant letter is a signed statement from the borrower's tax professional. Used well, it confirms facts that are otherwise expensive to establish: that the accountant prepares the returns, that the business has existed since a particular date, that the borrower's ownership percentage is what the K-1 says, and sometimes that the accountant is not aware of anything that would prevent a withdrawal from the business.

Used badly, it becomes a wish. Here is what it cannot do:

• It cannot substitute for the returns. No letter makes an unfiled or unverifiable return unnecessary.
• It cannot create income. A CPA writing "the borrower's true income is \$12,000 a month" is offering an opinion about a number the return does not support. It changes nothing.
• It cannot obligate the accountant to opinions they will refuse to give. Many tax professionals will not sign a statement about a business's viability or the impact of a withdrawal, because it is an assurance opinion carrying real professional liability, and they were engaged to prepare returns. This is not obstruction. Have a plan B before you request it: business bank statements, a CPA-prepared balance sheet, or a smaller withdrawal.

Ask for exactly the facts you need, in writing, with the language you want. "Please confirm the business has operated continuously since March 2019 and that the borrower owns 100% of it" gets signed. "Please confirm the borrower can afford this loan" does not.

The business narrative

Not a required form and not in anyone's guideline. A business narrative is a short document — half a page — in which the borrower describes what the business does, who its customers are, how it has changed over two years, and what explains anything unusual in the returns.

It is the cheapest underwriting tool in this chapter. An underwriter reading a set of returns cold is reconstructing a business from tax lines. Half a page in the borrower's own voice turns a \$16,800 K-1 drop into "we bought a second service truck in March and put a fourth crew on the road." Write the prompt for them, read the draft, and put it in the submission package.

The timing conversation nobody has

The most valuable thing you will ever say to a self-employed borrower is said to someone who is not buying a house yet.

Once a return is filed, the number is the number. There is no correcting it and no arguing with it, and the only remedy anyone will suggest is one you must never suggest. But a business owner who knows eighteen months before they buy that mortgage qualification runs off the tax return can have a genuine conversation with their accountant about the trade-off: the tax saved by a deduction versus the borrowing capacity it costs.

That conversation is theirs and their accountant's, not yours. You do not give tax advice, ever. You do not tell a borrower to take fewer deductions, to change how they file, to reclassify compensation, or to amend anything. What you may do — and should — is describe the mechanism accurately and hand them the question:

"I am not your tax advisor and I am not going to tell you how to file. What I can tell you is how the qualification works, and it works off the taxable income on your return. If buying is eighteen months out, that is worth a conversation with your accountant about the trade-off, because they can price the tax side and I can price the mortgage side. Do you want me on that call?"

Three of the best referral relationships a loan officer can build are with CPAs, because a CPA has twenty clients with exactly this problem and no way to solve it alone. Chapter 38 turns that into a business plan. The reason it works is the book's fourth theme: the relationship outlasts the transaction, and this is a conversation that has no transaction in it at all.

When the agency path does not work

Sometimes the worksheet produces a number that will not buy the house, the decline is real, the business is fine, and nothing about the file is dishonest. Say so plainly, do not spend six weeks proving it again, and do not manufacture an add-back you cannot defend.

There is a documented market for exactly these borrowers — loans underwritten to the ability-to-repay standard using alternative documentation rather than agency cash-flow analysis. Chapter 34 covers it in full, including who it serves, what it costs, and where its risks sit. That chapter owns the subject, and this one stops here: when the agency analysis fails, the honest next sentence is "there is another path and it prices differently — let me show you what Chapter 34's products actually cost," not a creative reading of a K-1.

🗂️ The Loan File

Chapter 32 contribution: re-run the qualification with Borrower 2 self-employed.

⚠️ This is a counterfactual, and the Linden Street file's canonical facts do not change. Borrower 2 is a W-2 outside sales representative; the file's frozen income is \$10,500.00, its housing ratio is 28.89%, and its back-end ratio is 42.66%. What follows asks a different question: what would this same file look like if Borrower 2 did the same work as a business owner instead of an employee? Call it the Fulton Avenue overlay. It is the fastest way to feel what this chapter costs.

The setup. Everything else holds. Borrower 1 is unchanged at \$6,300.00 a month. The house is still \$385,000, the loan is still \$365,750, and PITI plus mortgage insurance is still \$3,033.72. Monthly debts are still \$1,446.00. Only Borrower 2 changes: instead of a W-2 base of \$2,400.00 and a 24-month commission average of \$1,800.00, they operate the same outside sales book through an S-corporation, taking wages plus a K-1.

The worksheet [constructed overlay for teaching]:

Footing: \$30,000 + \$15,200 + \$3,600 − \$4,200 = \$44,600 ✓ · \$30,000 + \$19,100 + \$3,200 − \$4,800 = \$47,500 ✓

The meals exclusion is large because this is an outside sales business, and the deductible-half rule bites hardest on the people whose job is buying customers lunch.

The averaging decision. Income rose: \$47,500 − \$44,600 = \$2,900, or 6.50%. Rising income takes the conservative convention — the 24-month average:

• 24-month average: (\$44,600 + \$47,500) ÷ 24 = \$92,100 ÷ 24 = **\$3,837.50**
• Most recent year alone: \$47,500 ÷ 12 = **\$3,958.33**

The rising trend costs this borrower \$120.83 a month, exactly the way the canonical file's rising commission trend costs them \$150.00. Same rule, same sting, different form.

**Qualifying income for Borrower 2: \$3,837.50** — against \$4,200.00 as a W-2 earner.

The recomputed ratios:

$$\text{Total income} = \$6{,}300.00 + \$3{,}837.50 = \$10{,}137.50$$

$$\text{Housing} = \frac{\$3{,}033.72}{\$10{,}137.50} = 29.93\%$$

$$\text{Back-end} = \frac{\$3{,}033.72 + \$1{,}446.00}{\$10{,}137.50} = \frac{\$4{,}479.72}{\$10{,}137.50} = 44.19\%$$

Does the file still work? Honestly: yes, but with less room, and one thing gets worse.

The case for yes. A 44.19% back-end is above the 43% figure that anchors a lot of manual underwriting, but conventional automated underwriting has in recent years returned approvals at ratios above that and up to roughly 50% — verify the current limit and your lender's overlays, since this is exactly the kind of value that gets revised. The compensating factors are real and unchanged: a 706 representative score with no lates in twenty-four months, 4.16 months of reserves after closing, and a documented business. Chapter 15's automated underwriting decision is where this actually gets answered, and Chapter 14's eligibility framework is what it gets answered against.

The case for caution, and it is specific. Recall the crisis. On day 44 a pre-closing credit refresh found a furniture account at \$611.00 a month. Re-run it on the overlay:

$$\frac{\$4{,}479.72 + \$611.00}{\$10{,}137.50} = \frac{\$5{,}090.72}{\$10{,}137.50} = 50.22\%$$

In the canonical file that shock produced 48.48% — over the approved ratio and a violation of condition 11, but arguably still inside the outer envelope of what automated underwriting tolerates. On the overlay it produces 50.22%, which is past that envelope entirely. The same mistake, on the same day, on a file whose only difference is how Borrower 2's employer is organized, moves from fixable in four business days to possibly not fixable at all.

And the file gets slower. Two years of business returns, K-1s, a year-to-date P&L, business bank statements, and a business tax transcript are added to a file that already took fifty-one days with a lock that was three days short of the contract's closing date on the day it was taken. The documentation would land at the front of the file, which is the only good place for it — but any condition that requires the borrower's accountant is a condition you do not control.

What this settles: the arithmetic of the overlay, and a concrete feel for what self-employment costs a file that was otherwise identical — 3.45% less income, 1.53 points more back-end ratio, and substantially less tolerance for a day-44 surprise.

What it does not settle: whether the underwriting system approves it. That is Chapter 15's question, run against Chapter 14's rulebook, and neither of them will answer it from a table in this chapter.

Open questions carried forward:

• Q32-a. If Borrower 2 were self-employed, would the \$4,900 quarterly commission deposit still be an asset-sourcing condition, or would it become a business-funds question? (Chapters 12 and this one; it is still not income either way — it is already inside the average.)
• Q32-b. At 44.19% back-end, which compensating factor is doing the most work — the reserves or the score? (Chapter 15)

Your task. In Appendix C's workbook, complete the overlay worksheet and both ratios yourself before reading the table above. Then write two sentences you would say on the phone to a self-employed Borrower 2 on day one, naming the number and the mechanism, without using the word "unfortunately."

Conclusion

A self-employed borrower's income is not hidden and it is not exaggerated. It has been compressed, legally and deliberately, by a professional doing exactly what they were hired to do. The cash-flow analysis is the industry's attempt to decompress it — adding back the deductions where no money left the business, subtracting the money that left without a deduction, and producing a figure an underwriter can put in front of an investor.

The arithmetic is learnable in an afternoon. The four structures determine which returns you need. The worksheet walks the personal return and the business return in order. One principle governs every adjustment: add back what was deducted and did not leave; subtract what left and was not deducted. A declining trend takes the lower year and needs a cause, not a feeling. Business funds require a question the balance sheet cannot answer on its own — can the company make its next payroll after this money goes.

What is not learnable in an afternoon is the conversation. Every self-employed file contains a moment where a competent, successful person is told their income is smaller than they believe. That moment happens on day three, from you, for free — or on day thirty, from an underwriter, with a contract on the table and a house at risk. The entire professional difference between loan officers, on this kind of file, is which of those two days it lands on.

Fulton Avenue's borrower makes about \$9,500 a month and qualifies at \$8,916.67. Both numbers are true. Only one of them buys a house, and only one person in the transaction is in a position to say so out loud.

Next: the opposite problem. Chapter 33 turns to first-time buyers and down-payment assistance, where the income is straightforward and the obstacle is the cash — and where the Harlow Street file's ratios of 41.48% and 51.00% are approvable in a way that would be impossible on the file you just re-ran.

Key Terms

Self-employed borrower — a borrower whose qualifying income derives from a business they own, generally at an ownership threshold of 25% or more; documented from tax returns rather than from an employer's promise. (Ch.32)

Sole proprietorship — an unincorporated business owned by one person, with no separate federal return; business activity is reported on Schedule C of the owner's Form 1040. (Ch.32)

Schedule C — Profit or Loss From Business, the schedule of the personal return where a sole proprietor reports receipts, cost of goods sold, expenses, and net profit or loss. (Ch.32)

Partnership — a business with two or more owners that files Form 1065 and passes income through to partners; generally pays no federal income tax itself. (Ch.32)

Form 1065 — the partnership information return, which produces a Schedule K-1 for each partner. (Ch.32)

S-corporation — a corporation that has elected pass-through treatment, files Form 1120-S, and normally pays its owner-employee both W-2 wages and a K-1 share of profit. (Ch.32)

Form 1120-S — the S-corporation return, which produces a Schedule K-1 for each shareholder. (Ch.32)

C-corporation — a corporation taxed as a separate entity, filing Form 1120 and paying its own federal income tax; the owner's income appears only as W-2 wages and dividends. (Ch.32)

Form 1120 — the C-corporation return. Its retained earnings are the corporation's money, not the shareholder's income. (Ch.32)

K-1 (Schedule K-1) — the statement issued by a pass-through entity reporting an owner's share of income, deductions, credits, ownership percentage, and capital activity. (Ch.32)

Cash Flow Analysis (Form 1084) — Fannie Mae's published worksheet for converting a self-employed borrower's tax returns into qualifying monthly income; Freddie Mac's counterpart is Form 91. Revised periodically — work from the current version. (Ch.32)

Add-back — an amount deducted on a tax return that did not represent cash leaving the business, restored to income on the cash-flow worksheet. (Ch.32)

Depreciation — the deduction allocating a tangible asset's cost across its useful life; the largest and most common add-back. (Ch.32)

Depletion — the deduction for the consumption of a natural resource; a non-cash deduction and therefore an add-back. (Ch.32)

Amortization — on a tax return, the write-off of an intangible asset such as goodwill or startup costs; a non-cash deduction and an add-back. Not to be confused with loan amortization. (Ch.32)

Meals and entertainment exclusion — the nondeductible portion of business meals: money that left the business and was never deducted, so it is subtracted on the worksheet. (Ch.32)

Nonrecurring income — income that arrived once and will not repeat, such as a settlement or an asset sale; subtracted from qualifying income. Its mirror, a nonrecurring loss, is added back. (Ch.32)

Business liquidity — the business's capacity to meet its short-term obligations, measured from the balance sheet by the current ratio and the quick ratio, and the analysis required before business funds may be used for a down payment. (Ch.32)

Profit and loss statement (P&L) — an interim statement of revenue and expenses for the current year, used to show whether a declining or unproven trend has stabilized. (Ch.32)

Accountant letter — a signed statement from the borrower's tax professional confirming specific facts about the business; it cannot substitute for returns or create income. (Ch.32)

Business narrative — a short borrower-written description of what the business does and what explains anything unusual in the returns; not required by guideline, and one of the cheapest underwriting tools available. (Ch.32)

Spaced Review

1. (Ch.11 + Ch.32) Chapter 11 established the conservative 24-month averaging convention for variable income, and this chapter applied it to a self-employed borrower whose income declined. State the rule for each direction of the trend, and explain in one sentence why the treatment is asymmetric.

2. (Ch.14 + Ch.32) A self-employed borrower's file produces a 44% back-end ratio, a 706 representative score, and four months of reserves. Using Chapter 14's vocabulary, identify which of those three facts is a guideline question, which is a compensating factor, and which might be an overlay question — and say which of them your lender, rather than the agency, controls.

3. (Ch.32) A borrower's Form 1120-S K-1 shows \$18,000 of ordinary business income. Their business bank statements show \$11,000 a month transferred to their personal account. What is the second number, what may you count, and what does the gap tell you about the business?

4. (Ch.11 + Ch.32) Both a W-2 borrower and a self-employed borrower qualify at \$7,000 a month of gross income. Name one reason the self-employed borrower has less cash to spend, and say whether any ratio in this book measures it.

5. (Ch.32) Apply the add-back principle, without looking at any list, to these four lines and defend each answer in one clause: a \$9,400 depreciation deduction; a \$7,200 annual vehicle lease; a \$4,000 insurance settlement received; the nondeductible half of \$6,800 of business meals.