Why Feasibility Studies Get Sent Back: What Reviewers Check

No agency publishes a list of the reasons a feasibility study gets rejected. What they publish is the instruction given to the person reviewing it, and that turns out to be more useful, because it tells you what the reviewer is looking for before they find it.

There is no deficiency list

For the regulation itself, section by section, see our 7 CFR Part 5001 Handbook.

Search for an official catalogue of common feasibility study defects and you will not find one. Neither SBA, nor USDA, nor HUD publishes a returned items list, a scoring rubric or a set of acceptance thresholds for a market or feasibility study. State appraiser boards publish lists of the most frequent standards violations they encounter, and USDA publishes application completeness checklists, but nothing in the federal material catalogues why a study itself fails.

That absence is not a dead end. Every one of these programmes tells its reviewers what to check, some of them in considerable detail, and those instructions are public. Read from the reviewer's side, they amount to a specification. A study written against them tends not to generate correspondence. A study written without knowing they exist generates a great deal.

Who actually reviews the study

The first thing to establish on any file is which desk the report lands on, because the answer differs by programme and the reviewers apply different standards.

On an SBA file, the reviewer is the lender. Nothing in the operating rules gives SBA a role in reviewing a feasibility study, because nothing in those rules requires one. The second reviewer, if there is one, appears only after an early default, when the guaranty purchase package is examined and the file has to show the documentation relied on to justify cash flow and loan approval. That is a bad moment for a first review.

On a USDA guaranteed file, the state office reviews the application for completeness against the regulation, and the National Office concurs where the loan or the applicant requires it. The published lender checklists frame the mechanism plainly: a complete application is required for the request to be considered, and the checklists exist to help the lender get there. They are informational guides that are not themselves submitted.

On a HUD multifamily file there are two reviewers in sequence. The lender's underwriter reviews and accepts the third party reports, and the guide is explicit that this responsibility requires the lender's oversight, acceptance and recommendation to HUD of the conclusions of those reports, except where modified, explained and justified in the underwriting. HUD then performs its own technical review. HUD's review appraiser conducts that review in compliance with the appraisal review standard, documents it on HUD's own appendix, and signs a certification.

On a conventional bank file, the reviewer is whoever the institution's collateral valuation programme puts in the seat, subject to the independence rules that keep loan production out of it, and the examiner behind them.

USDA's own review instruction, and the tiering nobody talks about

The most specific published review instruction in this field is not in the regulation. It is a Rural Development unnumbered letter dated 14 March 2018, on the requirements for financial feasibility evaluations in Community Facilities applications, which states that it replaces all previous guidance on the subject.

It sets three levels of report. Community Facilities accepts a financial feasibility analysis, a financial feasibility study with review report, or a financial feasibility study with examination opinion, and the level required depends on the loan amount, the applicant's experience and the security. The analysis describes the applicant's present situation, analyses alternatives and outlines the proposed project. The review report provides a moderate level of assurance. The examination opinion provides a high level of assurance on the reliability of the applicant's financial statements and management's underlying assumptions.

The tiering runs on dollar bands and on a test of the applicant. The letter states that for small loans, less than $500,000, a financial feasibility analysis without an examination opinion will likely be appropriate, and its decision tree escalates the required assurance through bands at $500,000, $1,000,000 and $5,000,000, with the answer also turning on whether the loan is secured by a general obligation bond or assured income.

The applicant test is the part worth memorising. It is called the five and five criteria: the operation must have been in existence for at least five years and have operated on a financially successful basis for the five years immediately before the application, with financially successful defined as positive net income before interest, depreciation and amortisation and a debt service coverage ratio of 1.1 or greater. An applicant that does not meet it must be referred to the National Office for concurrence. A borrower who fails that test by one year, or by one weak year, is on a different track entirely, and the difference is not visible anywhere in the regulation.

The letter also tells the processing office what to verify. Where a review report or examination opinion is required, it must be prepared by a certified public accountant with the necessary expertise and backed by professional liability insurance, and the processing office will verify that the insurance is in place. It then lists what the study will typically include: a signed and dated opinion or review letter; five years of historic and five years of forecast financial statements; a schedule of ratios pertinent to the industry; a summary of the significant financial forecast assumptions and accounting policies; a summary of the significant demand forecast assumptions, meaning the market assessment; sensitivity analyses; other information as appropriate; and information on the governing body and management team.

That is as close to a specification as this field gets, and it is eleven years old, published outside the regulation, and largely unread outside the state offices that apply it.

For the guaranteed programmes the position is thinner. Beyond the five enumerated components in the regulation and the completeness checklists, USDA publishes no sentence level rubric for reviewing a feasibility study. The review is a completeness review against the components, which is why an unaddressed component is the most predictable reason a study comes back.

What a bank reviewer is told to test

The most quotable instruction to any reviewer of a feasibility study sits in the supervisory handbook for commercial real estate lending, in a subsection on determining project feasibility:

"Prudent underwriting includes determining the project's feasibility. Feasibility describes the likelihood that the project as proposed will be economically successful. Feasibility studies can be included as part of an independent appraisal or as a separate analysis; however, feasibility studies commissioned by the borrower may be biased and should be critically reviewed. While studies and appraisals can be helpful in providing useful information and analysis, the bank should conduct its own analysis of the project."

Three tests follow it, and each maps directly to a category of defect.

On projections: credible pro forma projections are a key determinant of a project's feasibility, and prudent underwriting includes reviewing the pro forma statement to determine whether the underlying assumptions and related projections are reasonable based on knowledge of the market and of income and expenses for similar properties. The reviewer is not asked whether the numbers add up. They are asked whether the assumptions are reasonable against comparable properties, which is a question a study either answers on the page or does not.

On sensitivity: the analysis should include stress testing to analyse sensitivity to changing economic conditions under a variety of scenarios, and the handbook names absorption rates, interest rates and capitalisation rates. Elsewhere it adds that capitalisation rates, interest rates and coverage ratios should be stress tested to determine whether a property will likely remain viable during a period of economic stress. A single point forecast invites the reviewer to build the stress case themselves, and they will build it less generously than you would have.

On market analysis: construction lending is particularly sensitive to market conditions, so a thorough market analysis is a critical component of underwriting, and it may include effective rental rates, sales prices, vacancy rates, building starts and absorption. Then the sentence that ties the whole thing together: the results of this review should support the revenue assumptions relied on in the pro forma financial statements. A study whose market section and financial section do not visibly connect has failed the test the reviewer is applying, however good each section is on its own.

The examination material adds a note on how much deference an outside analyst's judgement gets. Assumptions made by qualified appraisers that are appropriate to the property should be given a reasonable amount of deference, and examiners should not challenge assumptions that deviate slightly. But examiners may adjust an estimated value for credit and classification purposes where they can establish that facts or assumptions provided by management are irrelevant or inappropriate, or where they can support alternative assumptions from available information. Deference attaches to the qualified professional's judgement, not to the borrower's inputs.

Depth of review, and what happens when a report is found wanting

The nearest thing in federal text to a deficiency letter procedure sits in the interagency appraisal guidance, which devotes a section to reviewing appraisals and evaluations. It is written for appraisals, but it is the template every reviewer in this field has absorbed.

On who reviews: the reviewer should have competence commensurate with the complexity of the transaction, the type of property and the market, and should be capable of assessing whether the report contains sufficient information and analysis to support the institution's decision.

On how deeply: the institution should implement a risk focused approach to determining the depth of review, differentiating high risk from low risk transactions so the review is commensurate with the risk, and the depth should be sufficient to ensure that the methods, assumptions and conclusions are reasonable. A larger or more unusual project draws a harder look by design.

On what happens next: the institution should have procedures for resolving inaccuracies or weaknesses identified through review, including communicating the deficiencies to the preparer and requesting their correction, and those communications must not result in coercion or undue influence. That is the deficiency letter, described in general terms. And the consequence of failing to resolve one is specific. Significant deficiencies that cannot be resolved with the original preparer must be addressed by obtaining a second report, or by relying on a review that complies with the appraisal review standard and is performed by an appropriately qualified professional, before the final credit decision.

Finally, the review is documented in the credit file at a level addressing the type, risk and complexity of the transaction. Which means that on any file that later goes wrong, there is a written record of what the reviewer thought of the study.

The reconciliation trap on HUD files

HUD multifamily files carry a defect category that exists nowhere else, because HUD is the only programme that puts two independent analysts on the same question.

The market analyst projects the rents. The appraiser conducts an independent rent analysis. The guide requires the analyst's projected rents to be compared with the appraiser's independent analysis and reconciled by the lender, and where the two differ, HUD's own briefing material is explicit that the discrepancies must be described and detailed by the lender, which must explain why the differences exist.

The practical consequence is that a market study on a HUD file is not reviewed on its own merits alone. It is reviewed against another professional's conclusions on the same building, by a lender who has to write down why they diverge. A study that reaches a defensible rent conclusion by an undocumented route is much harder to reconcile than one that shows its comparable selection and its adjustments, because the reconciliation is a comparison of reasoning, not of numbers.

Two further HUD rules bear on rejection. The lender hires the third party contractors, and involvement by a consultant or broker in any of those functions is unacceptable under the programme, with the contractors barred from any identity of interest with the borrower, sponsor or an affiliated entity. And HUD staff check the limited denial of participation list and the consolidated list of suspended and debarred contractors, with the guide requiring lenders to run those checks every six months before employing or contracting anyone for this work. Where there are problems with the qualifications or past performance of a lender's reviewer, the team leader should halt the HUD review.

Appraisal review as the model

Feasibility work has no review standard of its own. Appraisal does, and it is worth understanding because it is the discipline every reviewer in this field was trained in.

Appraisal review is developed under one standard and reported under another, following the retirement of the old consulting standards. The reviewer must identify the work under review and the characteristics relevant to the intended use, identify the physical, legal and economic characteristics addressed in it, identify the extraordinary assumptions and hypothetical conditions it relies on, and form an opinion about the quality of the work. Where the review assignment includes the reviewer's own opinion of value, that portion is itself an appraisal and carries the full development and reporting obligations.

The reporting standard requires each analysis, opinion and conclusion to be communicated in a way that is not misleading, sets mandatory content for the review report, and requires a signed certification. Anyone who signs any part of the review, including the letter of transmittal, signs the certification too.

Two things follow for a feasibility study. First, extraordinary assumptions and hypothetical conditions are a named object of review, which is an argument for stating them explicitly in the study rather than leaving them implicit in a stabilised year. Second, the reviewer's job is to form an opinion about the quality of the work, not to re-perform it. Quality here means whether the reasoning is visible and sound. A study that shows its working is being reviewed on what it did. A study that presents conclusions is being reviewed on whether the reviewer happens to agree.

What the record shows actually fails

The oversight record is consistent across three decades and it points somewhere slightly unexpected. The failures are rarely wrong forecasts. They are unverified inputs.

When the Government Accountability Office examined 24 loss generating Business and Industry borrowers in 1999, it found the agency had not followed its own requirements on 18 of them. Eleven of those loans were missing feasibility studies that officials should have obtained. Three carried studies with significant flaws. Four were questionable given the business risk and the potential loss of collateral. Field officials had not obtained studies where the lender and borrower information appeared to provide an adequate basis, and internal quality control reviews across thirteen states in the same period had found the same thing.

The SBA inspector general's early default reviews name their defect categories repeatedly, and the categories are documentary. Business valuations. Affiliates. Equity injection. Repayment ability. Underneath those headings sit seller financial statements never verified against tax records, affiliate obligations never brought into the analysis, add backs that did not survive scrutiny and injection funds that could not be traced. In a 2019 consolidated review, deficiencies justified denial of the guaranty on five of eight early defaulted loans, approximately $8.7 million.

USDA's equivalent is quieter and just as instructive. A 2001 audit of a guaranteed loan to a forest products company in continuous operation since 1894 found that the state office had waived loan covenants two months before the lender analysed the borrower's financial statements, at a point when the borrower was technically insolvent, with a significant decrease in net worth and an insufficient repayment ratio. Four million dollars was questioned. The analysis was not wrong. It had not been done yet.

The lesson for a study is direct. The parts most likely to be attacked are the parts resting on somebody else's unverified figures: revenue supplied by the borrower, competitor performance taken from a listing rather than from the operator, cost inputs carried from a bid nobody tested, market data whose vintage is not stated. A conclusion the reviewer can trace to a source survives. A conclusion that traces back to the applicant does not.

What happens to the preparer

Very little, in the two programmes where most of this work is done, and that is worth saying plainly rather than implying otherwise.

SBA publishes no mechanism that reaches a feasibility study preparer. Its supervision and enforcement machinery reaches lenders, development companies and agents, and general debarment reaches persons in covered transactions, but nothing addresses the author of a deficient study as such. The consequence runs to the lender instead, through repair or denial of the guaranty. USDA is the same in shape: it can decline to accept a study, or decline to accept the consultant as acceptable to the Agency, and it participates in government wide debarment, but it publishes no preparer specific sanction and no removal from an approved list, because there is no approved list.

HUD is the exception. A limited denial of participation may be issued against a person for causes including irregularities in past performance in a HUD programme or failure to honour contractual obligations, excluding that person from a specific programme within a field office's geographic jurisdiction for a period of up to twelve months. And the enforcement rules for programme lenders name third party report conduct directly, listing as grounds for sanction both lender collusion with or influence upon third party contractors to modify reports affecting the contractor's independent evaluation, and a violation of programme procedures by a third party contractor which the lender knew or should have known was occurring and which would be a sanctionable ground if the lender had done it itself.

So the real consequence of a deficient study, in the programmes where most studies are written, is not a sanction. It is a rejected report, a delayed application, a lender who does not call again, and, if the loan defaults early, a purchase reviewer reading the file.

Twelve things that get a study sent back

None of this is published as a list, so the following is derived from the review instructions above rather than quoted from any agency. Every item corresponds to something a reviewer is told to check.

A missing component. On USDA files the study is reviewed against the five enumerated components. An unaddressed component is a completeness defect and returns the application.

The wrong level of report. On Community Facilities files, delivering an analysis where the loan size, the security or the five and five test called for a review report or an examination opinion produces a resubmission, not a revision.

No sensitivity analysis. It is a named factor in the codified components and a named expectation in supervisory guidance. Its absence is the easiest defect for a reviewer to identify without reading closely.

A market section that does not support the financial section. The supervisory instruction is that the market analysis should support the revenue assumptions in the pro forma. Two internally sound halves that never meet will fail that test.

Undated data. A reviewer assessing whether conclusions are reasonable in current market conditions cannot do so where the vintage of the underlying data is not stated.

Unverified competitor information. Supply findings taken from aggregators rather than from the operators or the public record are the first thing a reviewer familiar with the market will test.

Borrower supplied inputs presented as findings. The supervisory instruction assumes a borrower commissioned study may be biased and should be critically reviewed. Anything traceable only to the applicant is read in that light.

Unstated extraordinary assumptions. Assumptions and hypothetical conditions are a named object of review. Leaving them implicit in a stabilised year invites the reviewer to identify them for you.

Conclusions without visible reasoning. The reviewer forms an opinion on the quality of the work, which requires the work to be visible. Comparable selection and adjustment logic belong on the page.

A capacity assumption that breaks a codified cap. Residential care projections above 90 per cent occupancy fail a rule rather than a preference on USDA files.

An unreconciled rent conclusion. On HUD files the analyst's rents are compared against the appraiser's independent analysis and the lender must explain any difference. A conclusion that cannot be traced cannot be reconciled.

An interested preparer. Identity of interest with the borrower or sponsor is disqualifying on HUD files, and independence is codified for USDA consultants. This one is not a defect in the report. It is a defect in who wrote it.

The wrong report for the file is a related failure, and on acquisitions it is avoidable: our SBA acquisition report stack planner returns the reports a file carries on the SOP edition that governs its loan number.

To check whether a specific loan needs a study, which other reports the file carries and how each section should be prepared, use the Feasibility Study Methodology Navigator.

For the cross-asset framework, demand regimes, valuation methods and reviewer tests, see the 2026 Feasibility Methodology Atlas.

Frequently asked questions

Does any agency publish a list of common feasibility study deficiencies? No. Neither SBA, USDA nor HUD publishes a deficiency list, a returned items list or a scoring rubric for a feasibility or market study. What they publish is the instruction given to reviewers, from which the recurring defects can be derived, and USDA publishes application completeness checklists that frame the return mechanism.

Who reviews a feasibility study on an SBA loan? The lender. SBA has no review role at origination, because SBA does not require the study. The second reader, if there is one, is the guaranty purchase reviewer after an early default, who examines the documentation relied on to justify cash flow and loan approval.

What does USDA look for when it reviews a study? Completeness against the five components enumerated in the regulation, at state office level, with National Office concurrence where the loan or the applicant requires it. On Community Facilities applications there is a further layer, an agency letter setting three levels of report keyed to loan size, applicant experience and security, with a five year operating and performance test that sends applicants who fail it to the National Office.

What is the difference between a financial feasibility analysis, a review report and an examination opinion? They are three levels of assurance in the Community Facilities framework. The analysis describes the applicant's situation, analyses alternatives and outlines the project. The review report provides moderate assurance. The examination opinion provides a high level of assurance on the reliability of the financial statements and management's underlying assumptions, is prepared by a certified public accountant with the necessary expertise, and must be backed by professional liability insurance that the processing office verifies.

What happens if a reviewer finds the study deficient? The ordinary path is a request to the preparer to correct it, made without coercion or undue influence. Where a significant deficiency cannot be resolved with the original preparer, the supervisory expectation on a bank file is a second report, or a compliant review by another qualified professional, before the final credit decision. On a USDA file an incomplete application is returned in writing for the missing items.

Can a consultant be barred for a deficient study? In the SBA and USDA programmes, no published mechanism reaches the preparer. The remedy is that the report or the consultant is not accepted, and the consequence of a weak study falls on the lender through repair or denial of the guaranty. HUD is different, with a limited denial of participation available against a person for up to twelve months within a field office's jurisdiction, and enforcement rules that name third party report conduct as grounds for sanctioning a programme lender.

Why do studies fail on documentation rather than on analysis? Because that is what the oversight record shows. Reviews of failed loans across three decades return to unverified seller financials, affiliates left out of the analysis, untraceable equity and studies that were never obtained, rather than to forecasts that turned out wrong. A reviewer can test whether a figure has a source. Testing whether a forecast is right takes years.

What single change makes a study harder to reject? Sourcing every load bearing figure visibly, with its date. The reviewer's task is to decide whether the methods, assumptions and conclusions are reasonable. Everything traceable to a named source survives that test. Everything traceable only to the applicant is read as potentially biased, because the supervisory guidance instructs the reviewer to read it that way.

Sources

USDA Rural Development Unnumbered Letter, Requirements of Financial Feasibility Evaluations for Community Facility Applications, 14 March 2018
7 CFR 5001.303, Applications for loan guarantee
7 CFR 5001.304, Specific application requirements for Community Facilities projects
7 CFR 5001.306, Specific application requirements for Business and Industry projects
Appendix A to Subpart D of 7 CFR Part 5001, Feasibility Study Components
USDA RD Instruction 5001
USDA Rural Development, Business and Industry Lender Complete Application Checklist
USDA Rural Development, Community Facilities Lender Complete Application Checklist
HUD Multifamily Accelerated Processing Guide, Handbook 4430.G, Revision 19 March 2021, Chapter 7, including Sections 7.4, 7.5 and 7.6
HUD Multifamily Accelerated Processing Guide, Chapters 2 and 11, lender and underwriter responsibilities
HUD Exchange, MAP Guide Briefing Webinar Series, Session 2, Chapter 7
HUD Handbook 4232.1 REV-1, Section 232 Handbook, Chapter 5
24 CFR Part 200, Subpart Y, MAP Lender Quality Assurance Enforcement
24 CFR Part 25, Mortgagee Review Board
2 CFR Part 180, government-wide nonprocurement debarment and suspension
2 CFR Part 2424, including Subpart J, HUD limited denial of participation
Uniform Standards of Professional Appraisal Practice, 2024 edition, appraisal review development and reporting standards
Interagency Appraisal and Evaluation Guidelines, 75 FR 77450, 10 December 2010, Sections III, XV and XVI
OCC Comptroller's Handbook, Commercial Real Estate Lending, Version 2.0, March 2022
FDIC Risk Management Manual of Examination Policies, Section 3.2, Loans, and the commercial real estate lending examination module
SBA SOP 50 10 8, Lender and Development Company Loan Programs, effective 1 June 2025
SBA SOP 50 53, Supervision and Enforcement, Office of Credit Risk Management
13 CFR 120.524, When SBA is released from liability on its guarantee
SBA, Regular 7(a) Guaranty Purchase Package Tabs
SBA Office of Inspector General Report 19-22, Consolidated Results of the OIG High Risk 7(a) Loan Review Program, 26 September 2019
SBA Office of Inspector General Report 26-07, SBA's Screening of 7(a) Loan Applications Under its Risk Mitigation Framework, 11 March 2026
GAO/RCED-99-249, Rural Development: Rural Business-Cooperative Service Business Loan Losses, 25 August 1999
USDA Office of Inspector General Report 34601-9-Te, Business and Industry Guaranteed Loans, Rural Development State Office, Bangor, Maine, November 2001
GAO-03-404, Regulatory Programs: Opportunities to Enhance Oversight of the Real Estate Appraisal Industry, 14 May 2003