SBA Loan Process

How to Get an SBA Loan Approved: The 2026 Document List, the Four Numbers, and When a Feasibility Study Is Actually Required

By David Jones, MAI, CCIM

Last reviewed: August 2026. Rulebook of record: SOP 50 10 8, effective 1 June 2025, as amended by SBA Policy and Procedural Notices.

1. The four tests that actually decide the file

Most guides to SBA approval present a checklist. A checklist is useful for assembly, but it does not tell you what the underwriter is doing with the pile once it arrives.

Here is what is happening on the other side of the submission.

Test one: eligibility. Is the applicant a for profit small business operating in the United States, in an eligible industry, with eligible ownership, using proceeds for an eligible purpose? Eligibility is binary and it is checked first, because a file that fails here is dead regardless of how strong the credit is. Ineligible business types, passive real estate holdings without an operating company, ownership that does not meet current citizenship and residency rules, and prior federal debt delinquency are the usual disqualifiers.

Test two: injection. Has the borrower put in the required equity, and can the source of that money be traced? SBA lenders do not accept a balance on a statement. They trace deposits, question anything that arrived recently, ask whether a "gift" carries an obligation to repay, and confirm that borrowed funds are not being dressed up as equity.

Test three: repayment. Can the business service the proposed debt with a margin, and can the guarantors meet their own obligations without stripping the business? The SBA benchmark is a debt service coverage ratio of at least 1.15x. Note that the standard applies to the underwriting case the lender documents, not to the most optimistic tab in your spreadsheet.

Test four: evidence. Is every assertion in the file supported by something outside the borrower's own narrative? Historical tax returns are evidence. A contractor's stipulated sum contract is evidence. An appraisal is evidence. An independent feasibility study is evidence. A confident paragraph in a business plan is not.

The reason this framing matters more than a checklist: it tells you where to spend money. A borrower with three years of clean tax returns buying an established business has an evidence problem that is already mostly solved. A borrower building a special purpose facility that has never operated has no historical evidence at all, which is precisely the situation the third party report exists to fix.

2. The 2026 rulebook: what SOP 50 10 8 changed

The operating instructions that govern 7(a) and 504 lending are published in SBA Standard Operating Procedure 50 10. The version currently in force is SOP 50 10 8, effective 1 June 2025, as subsequently amended by Policy Notices. Confirm the current edition and any successor notice before you rely on any specific rule below, because SBA amends between SOP releases and the amendments are where the surprises live.

The 2025 edition was, in substance, a tightening. Several liberalizations from the 2023 to 2024 period were rolled back. The changes that most often affect whether a file gets approved:

Equity injection was restored as a hard minimum. A minimum ten percent equity injection applies to start-ups and to complete changes of ownership. Under the prior edition this had been left more substantially to lender discretion. If your capital stack was modeled on 2024 guidance, remodel it.

Seller financing counts, but only on standby. Seller debt may be applied toward the required injection only where it is on full standby, and only for part of the requirement. Structure the standby period into the term sheet before you sign the purchase agreement, not after.

Tax transcript verification was reinstated across the board. IRS verification of the financial information submitted is required. Every discrepancy between what you filed with the IRS and what you handed the lender becomes a question, and questions become weeks.

Ownership eligibility narrowed. The current edition requires that the applicant business be wholly owned by United States citizens, United States nationals, or lawful permanent residents. Files with partial foreign ownership that were structured under earlier guidance need to be re-examined.

Franchise review returned to a directory model. The SBA Franchise Directory was reinstated effective 1 June 2025. Franchise brand eligibility is now a listing question rather than an agreement-by-agreement control analysis.

Hazard insurance and life insurance requirements were restored. Collateral insurance is required, and life insurance on a sole owner whose loss would materially impair repayment is again the expectation rather than a lender option.

The combined program cap. SBA has moved to permit a borrower to carry up to five million dollars of 7(a) exposure alongside up to five million dollars of 504 debenture exposure, for a combined ten million dollars, per SBA Policy Notice 5000-879058, effective 4 July 2026. This is the single most consequential structural change for larger owner-user projects in several years, and it is recent enough that you should confirm current guidance and your lender's adoption of it before designing a capital stack around it.

The practical point for an applicant: SBA policy is not stable furniture. Any consultant, checklist, or article that describes SBA requirements without naming the edition and the date it is working from is describing a rulebook that may no longer exist. That includes this page, which is why the review date is at the top.

3. 7(a) or 504: choosing the wrong program costs six weeks

Program selection is made before documents are assembled, and it is the cheapest decision to get right and the most expensive to reverse.

Two structural traps worth naming.

The 504 goodwill trap. A buyer purchasing an operating business that includes real estate frequently wants 504 pricing on the whole transaction. The debenture cannot fund goodwill. The usual resolution is a split: 504 on the real property and long life equipment, 7(a) on the goodwill, working capital and closing costs. That is two files, two sets of conditions, and one closing that has to be coordinated.

The special purpose escalation. If your property is on SBA's special purpose list and the operating business is new, the 504 injection requirement moves from ten percent to twenty percent. Borrowers routinely discover this after the land is under contract. See the list in section nine; if your asset is on it, price the injection at twenty percent from day one and be pleasantly surprised if it is fifteen.

4. The four numbers, and how each one is computed

Underwriting is not vibes. Four numbers are calculated, and the file lives or dies on them.

4.1 Debt service coverage ratio

The SBA benchmark is 1.15x. That is a floor, not a target. A conventional lender on the same asset typically wants 1.20x to 1.25x, and USDA Business and Industry underwriting typically sits at 1.25x, which is why the same project can pass one program and fail another.

The computation that matters:

DSCR = (Net operating cash flow available for debt service)
       ÷ (Total annual principal and interest on all debt, including the proposed loan)

Where files break:

• The wrong numerator. Owner compensation has to be normalized to a market salary for the work being done, not to zero and not to the seller's tax-optimized draw.
• Missing debt. Global debt service includes existing business debt, affiliate debt, and in most credit policies the guarantors' personal obligations. A 1.28x business DSCR can become a 1.06x global DSCR once a personal residence, two auto notes and a second business are added.
• The ramp. A start-up that achieves 1.35x in year three but 0.62x in year one has not met the test. It has met the test in year three. Interest during construction, pre-opening expense and a funded working capital reserve are how that gap gets bridged, and they belong in the sources and uses schedule.
• The stress case. If a ten percent revenue reduction takes you below 1.0x, expect the file to be conditioned or declined, whatever the base case says.

4.2 Equity injection

Ten percent is the headline for start-ups and complete changes of ownership. What underwriting actually tests is the source and seasoning of the money.

Prepare, for every dollar:

• Which account it came from, with statements covering at least the last two to three months
• What produced it, with documentation for asset sales, retirement distributions, or gifts
• Whether it is borrowed, and if borrowed, against what and on what terms
• Whether it is still there, because injections spent before closing are not injections

Seller notes on full standby can carry part of the requirement. Home equity lines can be a problem if the repayment obligation is not covered in the global cash flow. Investor money changes the ownership schedule and therefore the guarantor list.

4.3 Collateral and loan to value

SBA is a cash flow program, not a collateral program, but the file still gets a collateral analysis. For 7(a), the lender is generally expected to collateralize to the extent of available business assets and, where the loan is not fully secured, to take available equity in personal real estate. An independent appraisal is generally required for commercial real estate transactions above the SBA threshold, currently five hundred thousand dollars.

For special purpose property the appraisal is a different animal. Where the value includes the business as a going concern, the appraisal has to allocate value between real property, machinery and equipment, and intangible or business value. That allocation is where feasibility work and appraisal work meet, and it is the most common place for the two reports to contradict each other. See section nine.

4.4 Total project cost

Not the loan amount. The total project cost. Underwriting reconciles the sources and uses schedule against the construction contract, the equipment quotes, the purchase agreement, the appraisal and the fee estimates, and total sources must equal total uses to the dollar.

The line items borrowers most often omit, in order of frequency:

1. Working capital for the ramp period
2. Interest carry during construction
3. Pre-opening expense: hiring, training, marketing, initial inventory, licensing
4. Furniture, fixtures and equipment beyond the headline machinery
5. Construction contingency at a defensible percentage
6. Professional fees, including the third party reports themselves
7. Closing costs, guaranty fees and prepaid items

A budget missing any three of these does not read as aggressive. It reads as inexperienced, and it invites the underwriter to re-underwrite the whole cost side.

5. The document matrix, by transaction type

Generic checklists list every document that any SBA loan has ever required, which is why they run to two hundred bullets and help nobody. What you actually need depends on what you are doing with the money.

Required in every file, regardless of transaction:

Then add, by transaction type:

6. The timeline: where files actually stall

A complete, uncomplicated 7(a) file can move from submission to closing in roughly thirty to sixty days. A 504 construction file with a feasibility requirement, an environmental escalation and a CDC review realistically runs ninety to one hundred fifty days. The variance is almost entirely caused by four things, and none of them is SBA.

The single highest leverage scheduling decision is when the third party reports are ordered. They run concurrently with underwriting if they are ordered at submission, and sequentially after it if they are not. A feasibility study commissioned after the credit committee asks for one adds its entire production time to the calendar, typically three to six weeks, on top of a file that has already been sitting.

The second highest leverage decision is resolving inconsistencies before submission rather than after. Each condition raised in underwriting costs a response cycle: the lender writes, you gather, the lender reviews, the analyst re-runs. Files that arrive internally consistent frequently close in half the time of files with the same credit quality and six reconciliation questions.

7. Why projections fail underwriting

A spreadsheet that balances is not a projection. A projection is an argument about the future with evidence attached.

The failure is almost always in the revenue build. Here is the difference.

A projection that fails:

Year one revenue: $2,400,000, growing 8% annually.

A projection that survives:

Year one revenue is built from 44 rentable units at an average achieved rate of $1,340, at 71% average occupancy across the year, reflecting a six month lease up from 22% at opening. The rate is set at the mid point of the four comparable properties surveyed within 4.2 miles, adjusted downward 4% for the absence of covered parking. Occupancy stabilizes at 89% in month 19, below the submarket average of 92% reported for the trailing four quarters.

The second version can be argued with, which is exactly why it is credible. Every number in it points at something the underwriter can check.

The units to build revenue from, by asset type:

The expense side fails more quietly. Payroll is the usual culprit: a staffing plan with fourteen positions and a payroll line that supports nine. Then payroll taxes and benefits, typically eighteen to thirty percent on top of wages depending on the benefit package, omitted entirely. Then insurance, property tax reassessed at the new basis rather than the seller's assessment, repairs and maintenance, and a replacement reserve for anything with a service life shorter than the loan.

And the working capital gap. A business can be profitable on the profit and loss and insolvent on the cash flow statement, because receivables, inventory build and the ramp period consume cash before the profits arrive. The projections need a monthly cash flow for at least the first twenty four months, and the low point of the cash balance across those months is a number the underwriter will find whether or not you present it.

8. Business plan, feasibility study, market study, appraisal: the four way distinction

Most published comparisons stop at business plan versus feasibility study. That is the easy half. The confusion that actually costs money in SBA files is between the feasibility study, the market study and the appraisal, because all three touch market data and only one of them is an opinion of value.

Three practical consequences.

One: the appraisal cannot substitute for the feasibility study. For a special purpose property with a going concern component, the appraiser is estimating value, often using a projected income stream. Where does that income stream come from? On a strong file, it is reconciled to an independent feasibility forecast. On a weak file, it is the borrower's own pro forma passed through, which means the appraisal and the projections share a single unexamined assumption set, and the file has one source of evidence pretending to be two.

Two: the feasibility study cannot substitute for the appraisal. A feasibility consultant who is not a certified appraiser cannot render an opinion of value on the real property, and SBA appraisal requirements are specific about who may. If a report offers both without the credentials for both, that is a defect a reviewer will find.

Three: independence has a definition. A feasibility study that reaches a favorable conclusion in every engagement is a marketing document with a different cover. The value of the document to the lender comes entirely from the possibility that it could have concluded otherwise. Any consultant whose fee, retention or referral flow is contingent on the conclusion has an impairment that should be disclosed and, in most cases, disqualifies them from the assignment.

9. When a feasibility study is required, and who requires it

This is where most published guidance goes vague, because the honest answer has three parts and only one of them is in the SBA rulebook.

Source one: the agency

SBA program requirements drive third party analysis where repayment depends on projections rather than history, most commonly for start-ups and for projects at special purpose properties. The specific triggering language changes between SOP editions, so the correct move is to have the lender identify the requirement in writing, citing the current SOP section, before anything is commissioned. A study ordered against a guessed scope is a study that gets rewritten.

Source two: the lender or CDC credit policy

This is the more common source in practice. A bank or CDC may impose a feasibility requirement that the SOP does not compel, because its own credit committee, its regulator, or its loss history on that asset class demands it. Credit policy requirements are usually narrower and more specific than SBA's, and they are the ones that actually define your scope.

Source three: the appraisal itself

An underappreciated trigger. When a special purpose property must be appraised with a going concern component, the appraiser needs a defensible income forecast and a value allocation between real property, equipment and business value. Where the appraiser is not independently qualified to develop the operating forecast for that asset class, a feasibility study becomes a practical prerequisite to a usable appraisal. Ordering the appraisal first, then discovering this, is how files lose a month.

The special purpose property list

SBA maintains a list of property types treated as special purpose, which drives both the 504 injection requirement and the appraisal treatment. It includes, among others:

amusement parks, bowling alleys, car washes, cemeteries, clubhouses, cold storage facilities where more than half the space is refrigerated, dormitories, farms including dairy and poultry facilities, funeral homes with crematoriums, gas stations and convenience stores with fuel, golf courses, hospitals, surgery centers and urgent care facilities, hotels and motels, marinas, mines, museums, nursing homes and assisted living facilities, oil and gas wells, quarries including gravel pits, railroads, sanitary landfills, service centers with pits and lifts, sports complexes, tennis clubs, theaters, and wineries.

Confirm the current list against the SOP appendix in force at the time of your application, because it is periodically revised.

The trigger checklist

Count the ones that apply to your transaction:

• The business has less than two years of operating history
• The property is on the special purpose list
• The project involves ground up construction
• The business model is materially changing under new ownership
• Repayment in years one and two depends on projected rather than historical revenue
• There are multiple revenue centers with different demand drivers
• The market is thin on comparable operating data
• The projection assumes market share or capture materially above the observable comparable set
• Absorption or lease up risk sits between opening and stabilization
• The brand or concept is emerging rather than established
• The lender has already asked what supports the revenue assumptions

Zero to one: a study is unlikely to be required, and probably should not be ordered.
Two to four: raise it with the lender before submission and get a written answer.
Five or more: budget for it now, commission it at submission rather than after, and expect it to be read closely.

10. What a lender grade feasibility study has to prove

Published descriptions of feasibility study contents are usually topic lists: market analysis, site analysis, financial analysis. A topic list tells you what is in the table of contents. It does not tell you what the document has to establish.

Here is the same territory expressed as the propositions a study must actually support, and the evidence each requires.

The single test that separates a lender grade study from an expensive brochure: can you trace a line from a named source, through the demand arithmetic, to a revenue line in the projection, to the DSCR? If any link in that chain is missing, the document is a collection of facts rather than an analysis, and an experienced reviewer will find the break in about twenty minutes.

11. Why feasibility studies get rejected

Studies are returned, discounted or ignored for a small number of recurring reasons. Ranked by how often they surface in lender review:

1. The scope did not match the requirement. The consultant wrote a market study, the lender needed a feasibility study, or the CDC needed both plus a specific section that was never included. This is a preventable failure and it accounts for a large share of rework.
2. Secondary data with no primary work. National industry averages applied to a specific site, with no competitor visits, no rate survey, no local verification. The report contains nothing the lender could not have generated with a subscription.
3. The conclusion does not follow from the analysis. The competitive section describes an oversupplied submarket and the forecast assumes above market capture, with no bridge between the two.
4. No sensitivity, or theatrical sensitivity. A downside case that never breaches coverage is not a downside case.
5. Numbers that contradict the rest of the file. Study project cost of 8.4 million, construction contract of 9.1 million, application of 8.9 million. Once one contradiction is found, everything else in the study gets re-checked.
6. No named sources or no dates. "Industry data indicates" is not a citation. A reviewer who cannot verify a number treats it as unsupported.
7. Stale data. A study leaning on a five year old survey in a market that has changed.
8. Evident advocacy. No risk that could actually impair repayment, no scenario that could fail, no recommendation the borrower would not have wanted. Independence that is asserted but not demonstrated.
9. Wrong author for the asset class. Generalist analysis of an asset with specific operating physics: a marina, a cold storage facility, an ambulatory surgery center.
10. The consultant will not take the call. A study whose author cannot or will not answer the credit committee's questions has a fraction of its intended value.

12. Hiring a consultant: the questions, and the answers that pass

Published lists of questions to ask a feasibility consultant are common. They are also useless on their own, because the borrower asking them does not know what a good answer sounds like. Here are both.

On price. Price is a legitimate factor and anyone who tells you it is not is selling. The right frame is not cheap versus expensive, it is what does a rejected report cost. A study that fails scope costs the fee, the replacement fee, and three to six weeks on a rate lock, a purchase agreement deadline, or a construction start. On a five million dollar project, three weeks of delay usually exceeds the entire difference between the cheapest and the most expensive proposal on the table.

The correct question is not "what does it cost" but "what is the probability this document survives review", and the way you assess that probability is the table above. See also our feasibility study cost breakdown.

13. Failure modes ranked by the time they cost

Not all mistakes are equal. Ranked by the delay each typically causes:

Fatal, restart required

• Applying under the wrong program and discovering it after underwriting has begun
• Ownership structure that does not meet current eligibility rules
• An ineligible use of proceeds discovered late
• Prior delinquency on federal debt, undisclosed

Costs four to eight weeks

• Third party reports ordered late, or ordered against a scope that was never confirmed
• Environmental escalation to Phase I or Phase II discovered mid underwriting
• Equity injection that cannot be traced, or that was spent before closing
• A business valuation that does not meet the independence requirement
• Discovering that the special purpose designation moves the injection from ten to twenty percent

Costs one to three weeks each, and they compound

• Project cost understated, requiring the budget and projections to be rebuilt
• Working capital omitted from sources and uses
• Projections inconsistent with the business plan narrative
• Staffing plan inconsistent with the payroll line
• Ownership percentages that differ between the entity documents, the application and the personal financial statements
• Unsigned or expired purchase agreement
• Interim financial statements prepared on a different basis from the tax returns
• Affiliate businesses omitted from the disclosure and found later

Costs credibility, which is worse than time

• A credit event that the lender discovers rather than being told about
• Overstated management experience that the resumes do not support
• A market share assumption with no derivation
• Any number that appears differently in two documents in the same file

The last category deserves emphasis. Disclosure is cheap and discovery is expensive. A bankruptcy discharged nine years ago, explained in a paragraph, is a footnote. The same bankruptcy found by the lender in a credit report is a re-examination of everything else you said.

14. The pre submission audit

Run this before the package leaves your hands.

1. Step 1. Confirm the program and the scope in writing. Ask the lender or CDC to state the program, the equity requirement, the collateral expectation, which third party reports are required, and the scope of each. Get it in an email. This one step prevents the two most expensive failure modes on the list above.
2. Step 2. Build the sources and uses to the dollar. Every cost, every source. Include contingency, interest carry, pre-opening, working capital, professional fees and closing costs. Total sources must equal total uses exactly.
3. Step 3. Reconcile the historical financials. Tax returns to interim statements to the debt schedule. Prepare a written explanation for every unusual item before anyone asks.
4. Step 4. Build the projections from operating assumptions. Units, rates, utilization, seasonality, ramp. Monthly for twenty four months minimum. Never select a target profit and work backward, because underwriters can see the reverse engineering from the shape of the expense lines.
5. Step 5. Order the third party reports. Appraisal, environmental, valuation, feasibility, whichever apply, all at submission, against the written scope from step one.
6. Step 6. Write the business plan against the projections. Not before them. The narrative should explain the assumptions the model already contains.
7. Step 7. Run the consistency check. Take these items and confirm each one is identical everywhere it appears: legal entity name, ownership percentages, loan amount, total project cost, equity injection amount, purchase price, opening date, square footage, unit count, staffing count, and the first year revenue figure. Ten minutes on this saves ten days.
8. Step 8. Stress it yourself. Compute the DSCR with a ten and a fifteen percent revenue reduction and with a six month delay to opening. If you cannot survive those, the lender will find out anyway, and you would rather find out first.
9. Step 9. Respond completely. When conditions come back, answer each question in a numbered response document that mirrors the lender's numbering, with the supporting exhibit attached to each answer. Partial answers create new questions.

15. Frequently asked questions

16. Method, sourcing and limitations

What this article is built from. SBA program requirements as published in SOP 50 10 8, effective 1 June 2025, and subsequent SBA Policy and Procedural Notices; SBA program descriptions for the 7(a) and 504 programs; and the pattern of conditions, rework and delay observed in lender facing feasibility engagements across the asset classes named above. Our full data methodology is published separately.

What it is not. It is not legal, accounting, tax or lending advice, and it is not a substitute for the requirements your lender or CDC gives you in writing. SBA policy changes between SOP editions through notices, and any specific figure here, injection percentages, thresholds, caps, coverage ratios, should be confirmed against the guidance in force on the date you apply.

What we will state plainly. No document, however well prepared, guarantees SBA approval. A complete, internally consistent, independently supported file does not make a weak project fundable. What it does is remove the avoidable reasons for decline, so that the credit decision is made on the merits of the project rather than on the quality of its paperwork. That is the entire claim, and it is the only honest one.

Sources and further reading

1. U.S. Small Business Administration, SOP 50 10 8, Lender and Development Company Loan Programs, effective 1 June 2025 (sba.gov)
2. U.S. Small Business Administration, 7(a) loan program overview (sba.gov)
3. U.S. Small Business Administration, 504 loan program overview (sba.gov)
4. U.S. Small Business Administration, Policy and Procedural Notices archive (sba.gov)
5. U.S. Small Business Administration, business plan guidance (sba.gov)
6. Uniform Standards of Professional Appraisal Practice, The Appraisal Foundation
7. Financial Institutions Reform, Recovery and Enforcement Act appraisal requirements, as implemented by the federal banking agencies
8. SBA Weekly Lending Report, 7(a) and 504 approval volumes (sba.gov)
9. SBA Feasibility Study LLC engagement records and lender review outcomes

About the author

David Jones, MAI, CCIM
Principal, SBA Feasibility Study LLC

SBA Feasibility Study LLC prepares independent, lender facing feasibility studies relied on by SBA lenders, Certified Development Companies, USDA lenders and conventional banks. Engagements span hospitality, self storage, car wash, medical and dental, manufacturing and agricultural processing, outdoor hospitality, and special purpose real estate. Our duty in every engagement runs to the lender and the agency reviewing the file, not to the outcome the borrower would prefer, which is what makes the document worth including in the file at all.

Disclaimer: This article is provided for general informational purposes and does not constitute legal, accounting, financial or lending advice. SBA policies, thresholds, forms and underwriting standards change. Confirm current requirements with your SBA lender, Certified Development Company, attorney and accountant. Neither professional document preparation nor an independent feasibility study guarantees loan approval.