Cancellation Is Not Charge-Off: The Two Ways an SBA Approval Can End Without Repayment

An SBA approval starts a loan; it is not yet one. Between the day SBA approves a 7(a) or 504 request and the day the last installment is paid, the approval can stop short in two ways that sit side by side in the loan file and mean nearly opposite things. It can be cancelled before any money moves, or it can be disbursed, fall into default and be charged off. The first is a transaction that did not happen. The second is money that was lent and not repaid. Anyone searching for SBA loan cancellation versus charge-off is usually trying to decide which of the two numbers matters for a project. The honest answer is that each matters for a different question, and that the two can never be folded into one figure.

This article explains how the SBA loan file records each ending, how the Loan Analytics dataset turns each into a rate, why the cancellation rate of a young cohort is still climbing, what each rate can and cannot tell a lender or a borrower reading a feasibility study, and why adding the two produces a number that describes no group of loans at all. Two neighbouring pieces of this series carry the rest of the story: one works through the mechanics that keep a charge-off rate a floor, the other maps the file's wider blind spots, and both are linked at the foot of this page. This article stays with the two endings.

An approval is counted before anything is lent

SBA's own statistics count approvals, not loans that reached a business. The agency's performance tables report gross approval year by year, and the page describing them explains that gross approval takes in loans cancelled in the year of approval or in any later year ^[1]. The FOIA loan file works the same way. Each row is an approval, dated to the fiscal year in which SBA approved it, and the row stays in the file after the approval is cancelled. The Loan Analytics dataset keeps that convention for every count it prints, so a line reading 78,078 ^[2] 7(a) approvals for FY2025 means approvals, cancelled ones included, and says nothing yet about how many loans were funded.

That convention is sound for what SBA uses it for, which is measuring the commitments a program made in a year. It becomes a trap when an approval count is read as lending. An approval opens a sequence of steps that still has to run its course. For a 7(a) loan, the lender has to satisfy the conditions attached to the approval, close the loan and disburse it. For a 504 loan, a Certified Development Company originates the loan together with a senior lender ^[3], the project may be carried on interim financing while it is built or bought ^[4], and the SBA-guaranteed debenture arrives at the end. Only after first disbursement is there a loan that can be repaid or lost.

From approval onward the possible endings are few. A loan can be repaid in full, which is one of the events on which SBA is released from liability on its guarantee ^[5]. It can be disbursed, default and be charged off. Or it can be cancelled before any disbursement, which ends the commitment with nothing lent. Until one of those three things happens, the approval sits in one of two waiting states: approved but not yet disbursed, or disbursed and still open. The regulation that governs the end of a guarantee lists several ways it can stop without any loss claim. 13 CFR 120.524(a) sets out the events that release SBA from liability ^[6], and among them are the lender asking SBA in writing to end the guarantee and a guarantee fee left unpaid past its deadline ^[7]. The file does not say which of these routes, if any, ended a particular cancelled approval. It records the outcome and nothing about the path.

What the status field records, value by value

The FOIA files carry one status field per approval, and its values are the raw material for both rates. The 7(a) files use five values and the 504 files six. One value is spelled differently between the programs: the 7(a) files write paid in full as P I F, with spaces, while the 504 files write PIF. The Loan Analytics dataset maps every value it meets to a closed list, and a value outside the list stops the build, so a new status cannot slip quietly into either rate.

Table 1 sets out what each value records, whether the dataset treats the loan as disbursed, and the source behind that reading.

The 504 files carry one more value, CLOSED, which the dataset files with disbursed loans and keeps out of both numerators; the file offers no further reading of it.

Three readings in that table carry everything that follows. First, a cancelled approval was never disbursed, so it can never become a charge-off. The two endings exclude each other by construction, and no loan can appear in both counts. Second, the waiting states differ by program. A 7(a) approval that has not yet funded shows COMMIT, while a 504 approval waiting on its debenture shows NOT FUNDED, and the 504 rules make the debenture wait for the project: before it is issued, the interim lender has to certify how much it advanced, and the CDC has to certify that the project was built as its final plans and specifications called for, subject to one exception the section cross-references ^[10]. Third, the EXEMPT value hides the condition of every open loan. Under SBA's FOIA rules every loan status except the two final ones, charged off and paid in full, is generally exempt from disclosure ^[11], so a disbursed loan that is current, late or in liquidation looks the same in the file until it is either repaid or charged off. That third reading is the reason a charge-off rate taken from this file can never be more than a floor, a point the series takes up at length in its own piece on that subject. The same rules treat the kinds, amounts and terms of loans, and statistical data on loans and defaults, as generally disclosed ^[13], which is why the file can carry amounts, dates and charge-offs while saying nothing about how an open loan is doing.

Cancellation: approved and never disbursed

In the Loan Analytics dataset a cancellation is an approval whose status is CANCLD: approved, then cancelled before any disbursement. The cancellation rate divides those approvals by every approval in the cohort. The cohort is fixed by approval year, FY2017 to FY2024, so at the data's as-of date of 2026-09-30 its approvals are between 2 and 10 years old. On that basis the 7(a) cancellation rate is 11.24 percent ^[14] and the 504 rate is 13.60 percent ^[15]. Both are approval-cohort floors: approvals in the cohort that are still waiting to disburse may yet cancel, and each one that does pushes the rate up.

The rate counts approvals, not dollars, and it counts every approval once, whatever its size. The dollar side can be read for a single year. Of the FY2024 7(a) approvals, 7,702 ^[16] had been cancelled by the as-of date, carrying $3.62 billion ^[17] of the year's $31.12 billion ^[18] in gross approvals. Those dollars were committed and counted in the year's volume, and none of them was ever advanced to a borrower.

Who ends an approval? The regulations show several hands that can, and the file names none of them. On the 7(a) side, the events that can release SBA from its guarantee, in whole or in part and at SBA's discretion, include the lender's written request and a guarantee fee not paid within the required period ^[7]. On the 504 side, SBA may, at its sole discretion, refuse to close the debenture, shift the loan to a different CDC, or withdraw its guarantee of the debenture ahead of the sale when one of a list of failures occurs, from a CDC's material breach of program requirements to a party's failure to disclose a material fact in time ^[19]. Among the certifications required once the project is complete and before a 504 loan closes, the borrower has to certify to the CDC that no substantial adverse change in its finances or its capacity to repay has arisen since it applied and been left unremedied ^[20]. A borrower whose business has turned down between application and closing meets that requirement at exactly the moment the approval is supposed to become a loan.

An approval can also end for reasons that sit outside the regulations, in the transaction itself. A purchase price that did not hold, an appraisal that came in below the contract, the owner's equity that never arrived, a construction bid that rose past the budget, a lease or a franchise consent that fell through, a borrower who found other money or changed its mind: any of these can end an approval, and so can many others. The file cannot say which of them did. Under SBA's FOIA rules, the non-statistical details of a canceled application, like those of a declined or a withdrawn one, are generally exempt from disclosure ^[21], and the public file carries no field for the reason, the party that initiated the cancellation, or its date. A reader who attributes a class's cancellation rate to any one cause is guessing, however plausible the guess.

What a cancellation is not matters as much as what it is. It is not a default, because no installment was ever due. It is not a loss in the sense a charge-off is, because no principal was advanced and there is nothing to recover. And it is not evidence that the project was unsound, because the same status covers a deal that collapsed under its own numbers and a borrower that walked away from a good one. An archived Loan Analytics post on approvals that never close, linked below, follows the gap between approval and funding for FY2025 in detail, including the explanations in circulation. The point here is narrower and structural: a cancellation is a transaction outcome, and it belongs in a column of its own.

The ceiling, and why a young cohort's rate keeps rising

A cancellation rate measured on a cohort that still holds undisbursed approvals has a lower bound and an upper bound, and the dataset prints both. The floor counts only the approvals already cancelled. The ceiling counts the cancelled approvals plus every approval still waiting to disburse, as if all of them were going to cancel. For the FY2017 to FY2024 cohort, the 7(a) floor of 11.24 percent ^[14] sits under a ceiling of 12.24 percent ^[22], with 4,463 ^[8] approvals still undisbursed. The 504 floor of 13.60 percent ^[15] sits under a ceiling of 16.90 percent ^[23], with 1,852 ^[9] still undisbursed. The final rate for each program will land somewhere between the two, and nobody can say yet where.

The gap between floor and ceiling is wider for 504, a pattern consistent with the program's structure, and the file itself offers no explanation of it. Interim financing may cover all of a 504 project's costs other than the borrower's own contribution, provided its source and terms meet the conditions the section sets ^[4], and no debenture goes out until the interim lender has put its advances in writing and the CDC has signed off on completion ^[10]. A 504 approval for a building under construction can therefore sit as NOT FUNDED for as long as the construction takes, while the project itself is well underway. One year makes the contrast sharp. Among the approvals SBA made in FY2024, the part still waiting on a first disbursement at the as-of date came to 17.30 percent ^[24] in the 504 program and only 3.42 percent ^[25] in 7(a). For 504, a large undisbursed share in a recent year can reflect construction time as much as trouble, and reading all of it as cancellation in waiting would overstate the program's attrition. The ceiling does exactly that on purpose, which is why it is a bound and not an estimate.

Why does a recent cohort's rate keep rising? Because of the way an undisbursed approval resolves. It has two exits and no others. If it disburses, it leaves the waiting pool and joins the disbursed loans, and the cancellation numerator does not change. If it cancels, the numerator rises by one. The denominator, the cohort's gross approvals, does not move either way, because a cancelled approval stays in gross approval ^[1]. So as long as the statuses already recorded stand, a cohort's cancellation rate can only hold or climb as it ages, and it can climb no higher than its ceiling. Each quarterly release of the files ^[26] is a fresh reading of the same cohort, and a young cohort's floor tends to move up from one reading to the next.

The FY2025 7(a) cohort shows how much room there is. At the as-of date its approvals were one to two years old; 14,124 ^[27] of 78,078 ^[2] had been cancelled, and 3,191 ^[28] were still waiting. Its floor is 18.09 percent ^[29] and its ceiling 22.18 percent ^[30]. The archived post on approvals that never close relied on an earlier quarterly reading and counted fewer FY2025 cancellations than the September 2026 file does; that difference is the rise at work, not a disagreement between sources. Table 3 shows where the FY2025 rate can settle, depending on how many of the pending approvals end cancelled. The share is an assumption, set at zero in Table 2 and raised column by column, while the counts come from the dataset. The last three inputs in that table serve a later illustration.

Rounded as shown; lines may not foot by one unit in the last place.

Two rules follow for anyone comparing cancellation rates. The first is to compare a floor with a floor and a ceiling with a ceiling, and to state each cohort's age beside it. The FY2024 7(a) cohort, two to three years old at the as-of date, has a floor of 10.96 percent ^[31] and a ceiling of 14.39 percent ^[32]. The pooled cohort's floor is higher than that single year's floor and its ceiling is lower. Its range is narrow because few of its approvals are still pending, while the FY2024 range is still wide. A single young year read against a pooled cohort can look better or worse than it will end up, and only its ceiling says how much worse it can get. The second rule is that a count net of cancellations is not a count of funded loans. The dataset's FY2025 7(a) count net of cancellations, 63,954 ^[33], still includes the 3,191 ^[28] approvals that had not disbursed by the as-of date.

Charge-off: disbursed, defaulted and written off

A charge-off is the other ending, and it can only happen to a loan that was disbursed. In the Federal Reserve's banking statistics, a charge-off is a loan's value taken off the books and set against the bank's loss reserves, and the rates are annualized and stated net of recoveries ^[12]. The SBA path to that point runs through default. For a 7(a) loan approved since May 2007, the lender can call on SBA, in writing, to make good on its guarantee once an installment has gone more than 60 calendar days unpaid, or less if SBA agrees, and the default is uncured ^[34], and the same paragraph ties that demand to the liquidation of the business personal property securing the loan. What cannot be collected is eventually written off. For an SBA Supervised Lender, the rules call for the unguaranteed share of any loan judged uncollectible to be charged off without delay ^[35].

A charge-off is an accounting recognition of loss, not the end of collection. SBA reports, program by program and year by year, what it has recovered after charging loans off ^[36], which means money can still come back after the write-off. The FOIA file records the charge-off, its date and its gross amount, and the dataset's charge-off figures are gross in the same way. Measured in dollars, the 7(a) loans disbursed from FY2017 to FY2021 approvals have so far produced gross charge-offs equal to 1.95 percent ^[37] of the amounts approved on them, an approval-cohort floor like the count, since that cohort is still accruing charge-offs. Neither the count nor the dollar figure says what the lender or SBA finally lost once recoveries were in.

The dataset's charge-off rate counts loans whose status is CHGOFF and divides them by the disbursed loans of the cohort. The cohort is FY2017 to FY2021, so every loan in it was approved 5 to 10 years before the as-of date. On that basis the 7(a) rate is 6.33 percent ^[38] and the 504 rate is 0.35 percent ^[39]. Both are approval-cohort floors. The denominators still hold open loans whose status the file withholds, 68,103 ^[40] in the 7(a) cohort and 22,353 ^[41] in the 504 cohort, and any of them can still charge off. The companion floor piece explains why such a rate can only understate, and how a second rate, counting only the charge-offs that came inside five years of approval, lets cohorts of different ages be set against each other.

The name charge-off rate is shared by numbers that cannot be compared with each other. SBA's own published charge-off rates are a percent of unpaid principal balance at the end of each fiscal year ^[42]. The Federal Reserve's bank charge-off rates are annualized flows, net of recoveries ^[12]. The dataset's rate is a count of loans charged off to date over the loans a cohort disbursed. The archived Loan Analytics post that ranks default rates by industry, also linked below, works from an older and mostly resolved cohort, which makes a fourth basis. A reader who sets one of these against another without naming the basis is comparing different things, however similar the label.

Two rates, two populations: the national figures side by side

With both endings defined, the two rates can stand side by side, each with its population named. The 7(a) cancellation rate is 11.24 percent ^[14] of approvals made in FY2017 to FY2024, and the 7(a) charge-off rate is 6.33 percent ^[38] of the loans disbursed from approvals made in FY2017 to FY2021. The 504 cancellation rate is 13.60 percent ^[15] on the first basis, and its charge-off rate 0.35 percent ^[39] on the second. All four are approval-cohort floors, and none of them is a probability for any single loan.

Read in pairs, these say something a single figure could not. The 504 program cancels a larger share of its approvals than 7(a) does, and it loses a far smaller share of the loans it disburses. A combined figure would put the two programs only a few points apart and hide the fact that one of them has charged off a small fraction of the share the other has. The file does not say why 504 approvals cancel more often. The program's mechanics described above, a debenture that waits on a finished project and a closing that requires a fresh certification of no adverse change ^[20], give a 504 approval more steps at which it can stop, but that is a description of the process, not a finding about these cancellations. Why 504 loans charge off so much less often is a question about collateral and loan structure that belongs to another article; the point here is only that the two rates move independently.

Table 4 gathers every figure this article leans on, with its basis and its key. For each cell rate the basis column prints the numerator and the denominator, so the rate can be recomputed from the line itself, and the denominators show at a glance that the two rates of one cell stand on different groups of loans.

Loan Data Behind This Article

Every figure in this table comes from the Loan Analytics loan dataset, analytics-loan-data freeze-3 (commit 360c856), and carries its key. SBA rates are approval-cohort floors and are withheld when fewer than 30 loans stand behind them.

Asset classes rank differently on each measure

The same split appears inside the national program by asset cells, where the two rates can rank classes in opposite orders. A cell's asset class is assigned from the loan's industry code, a step the series documents in its separate piece on class assignment, and every rate below is an approval-cohort floor read from one cell.

Hotels and motels financed under 7(a) come first. Out of the cell's 7,364 ^[43] hotel approvals dated FY2017 through FY2024, 8.76 percent ^[43] ended in cancellation; out of the 4,753 ^[43] hotel loans that actually funded from FY2017 through FY2021 approvals, 0.99 percent ^[43] have been charged off so far. The two denominators differ by more than two thousand loans, because the charge-off cohort covers fewer years and leaves out everything that never disbursed. Hotels sit low on both measures.

Fitness centers financed under 7(a) look different. Their cell's cancellation rate is 11.56 percent ^[44] and its charge-off rate 9.48 percent ^[44]. Fitness centers cancel somewhat more often than hotels and charge off close to ten times as often. On cancellation the two classes are a few points apart; on loss, they are an order of magnitude apart.

Self storage financed under 504 reverses the picture again. So far the cell has lost 19.15 percent ^[45] of its cohort's approvals to cancellation, and it would reach 23.76 percent ^[45] if every approval still pending fell away, the highest figures of the three cells on either reading. Against that, its charge-off rate is 0.00 percent ^[45] on 412 ^[45] disbursed loans. No charge-off has been recorded in that cohort so far. A floor of zero is still a floor, since open loans remain in the denominator, but it is a very different record from the fitness cell's.

Ranked on cancellation, self storage under 504 comes first, fitness centers second and hotels third. Ranked on charge-off, fitness centers come first, hotels second and self storage third. A study that read the cancellation rate as a stand-in for credit risk would rank the self storage cell as the riskiest of the three, and the charge-off record contradicts it. Added together, the two rates would put the self storage cell close to the fitness cell, even though one has recorded no charge-off in its cohort and the other carries the highest charge-off rate of the three.

Each cell rate stands on its own denominator, and in these three cells the denominators are large enough for the rates to be printed. That is not always so. The national dialysis center cell under 7(a), for example, holds 28 ^[46] approvals over FY2017 to FY2026, its cancellation cohort holds fewer still, and its cancellation rate is withheld: fewer than 30 loans. The reasoning behind that cut-off gets its own treatment in the series piece on thin samples. A cell with a withheld rate is not a cell with a rate of zero.

Why the two rates must never be added

It is tempting to report one number for how often an SBA approval ends badly, and the shortest route to one number is to add the cancellation rate to the charge-off rate and present the total as a failure rate. The total describes no group of loans. Four separate reasons make that so, and any one of them would be enough.

The denominators differ. The cancellation rate is a share of approvals. The charge-off rate is a share of disbursed loans, a smaller group that leaves out every cancelled approval and every approval still waiting. Two shares of different wholes do not add up to a share of either whole, any more than a share of applicants and a share of hires add up to a share of anything.

The cohorts differ. The cancellation cohort runs FY2017 to FY2024 and the charge-off cohort FY2017 to FY2021. A sum mixes approval years that the charge-off rate does not cover at all, and it mixes approvals that are 2 to 10 years old with loans that are 5 to 10 years old. For 7(a), the pooled floor and ceiling sit close together, consistent with most of a cohort's cancellations arriving within a few years of approval, while charge-offs keep arriving long after that, so the two halves of the sum are also at different stages of completion.

The two endings exclude each other. A cancelled approval can never charge off, so a higher cancellation rate leaves fewer disbursed loans that can be lost. Table 5 uses an illustrative cohort with nothing left pending, an assumed cancellation rate and an assumed charge-off rate on disbursed loans, all labelled as assumptions in Table 2, to show what that does to the arithmetic. At the base assumptions the sum of the two rates is 18.00 percent, but the approvals that actually ended in either way make up 17.28 percent of the cohort, and two thirds of the sum is made of approvals that never disbursed a dollar. Double the cancellation rate and the sum climbs to 30.00 percent, while charge-offs as a share of approvals fall, because fewer loans were disbursed that could fail. A combined figure that rises when fewer loans are lost is measuring the wrong thing.

The events differ in kind. A cancellation means no principal was advanced; a charge-off means principal was advanced and written off. One is a measure of whether a transaction reaches funding, the other a measure of credit loss after funding. A sum treats a buyer who walked away before closing as the same event as a borrower who stopped paying in year four, though a study should treat them as two different risks.

Even the share of approvals that ended either way, the row in Table 5 that is computed correctly on one denominator, is not a failure rate. It is a count of outcomes, and it still mixes events that cost no principal with events that did. Presented properly, the two figures sit in two columns and never in one. The cancellation column carries its approval years, the age of the cohort, the floor and the ceiling. The charge-off column carries its own approval years and age, plus the number of open loans that can still fail. Each column rests on its own denominator, and each says plainly that it is a floor.

What each rate tells a feasibility study

A cancellation rate reads as transaction risk: how often approvals in a class stop between commitment and first disbursement. A charge-off rate reads as credit risk: how often loans in a class that did fund were later written off. A feasibility study touches both, but through different parts of its work, and neither rate is a forecast for the project in front of the lender.

The cancellation side of a study is about whether the conditions that have to hold for the approval to close are documented. That means the market support behind the price or the projected revenue, a cost budget that a contractor's bid can meet, a construction schedule that fits the financing, the source of the owner's equity, and the permits and consents the project depends on. Putting a value on the property is the appraiser's job, and the study leaves it alone; where an appraisal is a condition of the loan, a study should set out its market and cost evidence so that it can be read beside the appraiser's figure. The 7(a) and 504 rules list a feasibility study among the items SBA is free to ask for ^[47], and once SBA has asked, the study sits among the conditions that stand between approval and closing. For a 504 project, the study's construction schedule bears directly on the stage the rules require before the debenture is issued ^[10], and its view of the business's condition bears on the certification of no adverse change at closing ^[20].

The charge-off side of a study is about whether the project's cash flow can carry the debt it is asked to carry, with debt service coverage tested against the program floor and against the lender's own requirement.

Conventions. This article follows the conventions of the Loan Analytics worked examples. Coverage is tested against the program floor the site's program page states, where the program has one, and against a stated 1.25x lender assumption, and the higher of the two binds. Equity is measured at the loan the cash flow supports, with fees recomputed on that loan. An acquisition is tested on the seller's historical earnings, not on projections.

A class's charge-off floor gives that coverage test its context. It is a lower bound on how often loans like this one have been written off within the seasoning the cohort allows, and a study can cite it only as that; the dataset's own definition of the national 7(a) figure of 6.33 percent ^[38] calls it a floor. It is never a probability of default for one project, for the reasons the floor article sets out, and a cancellation rate is even less of one. A study that says a class cancels a given share of its approvals has described the class. It has said nothing about whether this borrower's approval will close.

The same discipline applies to cause. When a class shows a high cancellation rate, a study can say so, print the floor and the ceiling, and point the reader to the conditions that the project has to meet before it can close. It cannot say that approvals in the class cancel because of appraisals, or prices, or equity, or because borrowers walk away, since the file cannot show any of these ^[21]. The useful sentence is about the project: which of its closing conditions are documented and which are still open, not about why other approvals failed to close.

Three misreadings are easy to make and should be ruled out in any study or loan memorandum that cites these figures. A class's cancellation rate is not the chance that a particular approval closes. A 504 undisbursed share is not a count of failing projects, since one reason an approval stays undisbursed is construction that has not yet finished. And the sum of the two rates is not a measure of risk at all.

What to check before quoting either rate

Before either rate goes into a study, a loan memorandum or a pitch to a borrower, it is worth putting eight questions to it. A line that quotes the rate well answers each of them without the reader having to look elsewhere.

• Which ending. Is the figure a cancellation, an approval that never disbursed, or a charge-off, a disbursed loan written off? The status field separates them cleanly, and SBA's own disclosure rules release the charged-off and paid-in-full statuses while withholding the rest ^[11].
• Which population. Is the denominator every approval in a cohort, or only the loans that disbursed? The dataset's 7(a) cancellation rate of 11.24 percent ^[14] and its charge-off rate of 6.33 percent ^[38], both approval-cohort floors, stand on different groups of loans.
• Which cohort and how old. Name the approval years and the seasoning at the as-of date: 2 to 10 years for the cancellation cohort, 5 to 10 years for the charge-off cohort, and one to two years for a single recent year such as FY2025, whose floor of 18.09 percent ^[29] is still moving.
• Floor and ceiling together. A cancellation rate printed without its ceiling hides how much is still pending; the 7(a) cohort's ceiling is 12.24 percent ^[22] and the 504 cohort's 16.90 percent ^[23].
• Which file. The FOIA files are updated quarterly ^[26], and as long as recorded statuses stand, a young cohort's rate read from an older file can be no higher than the same rate read from a newer one.
• Which basis, for an outside charge-off rate. SBA's published rate is a percent of unpaid principal balance at fiscal year end ^[42], and a bank rate is annualized and net of recoveries ^[12]. Neither is an approval-cohort rate.
• Enough loans. Any rate that would rest on fewer than 30 loans is left blank rather than shown as zero, and a cell that shows only a count, such as the dialysis center cell with 28 ^[46] approvals, offers no rate to quote.
• No cause, no sum. The file records that an approval was cancelled, not why ^[21], and the two rates are never added into one figure.

A figure that passes those checks can be quoted with confidence about what it means, even though both rates remain floors. A figure that fails one of them is not necessarily wrong, but it is not yet saying what its reader thinks it says. The difference between a deal that never happened and a loan that was lost is the first distinction a study should draw about any SBA outcome, and the file, read carefully, keeps the two apart.

Related Pages on Loan Analytics

• The Loan Analytics methodology series (all articles)
• How a feasibility study is done: the Loan Analytics methodology
• SBA feasibility studies (7(a) and 504)
• Loan data by state
• Loan data by asset class
• Worked examples (case studies)
• Worked Example: SBA 7(a) Assisted Living Feasibility Study in Utah
• Worked Example: SBA 504 Express Car Wash Feasibility Study in Maryland
• Worked Example: SBA 7(a) Gas Station Feasibility Study in Tennessee
• Worked Example: SBA 7(a) Gym Feasibility Study in Hawaii
• Worked Example: SBA 504 RV and Boat Storage Feasibility Study in New Jersey
• Worked Example: SBA 504 Owner-Occupied Retail Feasibility Study in Michigan
• The SBA Loans That Never Close: Inside the Gap Between Approval and Funding
• SBA Loan Default Rate by Industry: Charge-Off Rates for 416 NAICS Codes
• The Approval-Cohort Charge-Off Floor: Why a Published SBA Loss Rate Can Only Understate
• What the SBA Loan File Cannot Tell You: The Limits of FOIA Data in Underwriting

Commission a Feasibility Study

Loan Analytics prepares lender-grade feasibility studies for SBA 7(a), SBA 504, USDA Business and Industry, conventional and EB-5 financing, built on the methods this series sets out. A study is not an appraisal: it tests whether a project's market, costs and cash flow can carry the debt a lender is asked to approve. To discuss a study for your project, contact Loan Analytics.

Daniel Smith, MAI, Loan Analytics