EB-5 Method: What USCIS Tests in the Business Plan and the Job Model, and How a Study Answers

EB-5 is an immigrant visa category, not a loan program, yet the file USCIS reads for an EB-5 project looks a great deal like a credit file. It holds a business plan that has to survive scrutiny, an operating projection, a construction budget with a schedule, and a count of the jobs the money is expected to create. This article walks through what the statute, the regulation, the leading precedent decision and the USCIS Policy Manual test in the business plan and in the job model, and what a feasibility study can put in front of each test. It uses a hotel as its running example, because a hotel's job model turns on exactly the figures a market and operating study produces: what the building costs, how long it takes to build, what it will earn and whom it will employ. It also says plainly where a study stops. USCIS decides eligibility, designates high unemployment areas and credits jobs. A study supplies evidence; it never forecasts an immigration result.

EB-5 Is Not a Loan, but Its Evidence Reads Like Underwriting

The classification rests on 8 U.S.C. 1153(b)(5), which makes visas available to investors whose capital goes into a new commercial enterprise that will create full-time employment for not fewer than 10 qualifying workers ^[1]. Nothing in that sentence mentions a lender. The link to lending comes through the evidence. In Matter of Ho, the 1998 precedent USCIS still cites for business plans ^[2], the Associate Commissioner pointed out that the Small Business Administration and private financial institutions routinely ask for detailed business plans before they extend credit ^[3], and then drew the comparison outright: "Permanent resident status is no less significant a matter than a loan." ^[4]

The same decision ties the count of jobs to feasibility in so many words. Without knowing whether a business is feasible and can last, the decision reasons, neither the petitioner nor the agency can reasonably conclude that it will create permanent full-time employment ^[5]. That is the question a lender-grade feasibility study is built to answer, and it explains why the document an EB-5 sponsor files can be the same document a senior lender reads when it finances part of the same building.

Three cautions frame the rest of this article. The first is that a study is evidence, not a decision. The standard USCIS applies is a preponderance of the evidence: relevant, probative and credible material that makes a claim more likely than not true ^[6]. The officer weighs the whole record, and a study is one exhibit in it. The second is that a study does not value the property; it is not an appraisal, and nothing in an EB-5 file turns on an appraised value. The third is that a study does not decide how many jobs USCIS will credit. Whether the job model sits inside the study or in a separate economic report, its inputs come from the study, and the study's task is to make those inputs defensible.

What the Statute Fixes: Capital, Place and Ten Jobs

Four elements of the statute frame every EB-5 business plan. The first is capital. The investor must have invested, or be actively investing, at least the required amount in a new commercial enterprise, and the capital must be expected to stay invested for not less than 2 years ^[7]. For petitions filed on or after March 15, 2022, the standard amount is $1,050,000 ^[8], and an investment in a targeted employment area or an infrastructure project needs $800,000 ^[9]. Both figures move on the same clock. The standard figure adjusts automatically from January 1, 2027 and on a five-year cycle after that ^[10], and on those same dates the lower figure resets to 75 percent of the adjusted standard amount ^[11]. A plan likely to be filed near an adjustment date should therefore state which amounts it assumes and the filing date on which that assumption depends.

The statutory meaning of capital matters for the financing picture a study draws. Capital invested in exchange for a note or other debt arrangement with the new commercial enterprise does not count, and neither does capital invested with a guaranteed rate of return ^[12]. The Policy Manual adds that capital is at risk only where there is both a risk of loss and a chance for gain ^[13]. Whether a given investor's money qualifies is a legal question, and a study does not answer it. Its contribution is to describe each layer of the project's funding without contradiction. The Policy Manual's own illustration has a regional center's new commercial enterprise lending pooled capital to the job-creating entity that builds the project ^[14], and the Policy Manual says the project's business plan should also identify the form of the investment from the new commercial enterprise into the job-creating entity, if there is one, whether an equity stake, a loan or another arrangement ^[15]. A study that calls the EB-5 layer equity in its sources and uses chapter and then models it as a fixed-rate loan in its cash flow hands both the officer and the senior lender an inconsistency they will notice.

The second element is place. A targeted employment area is, at the time of investment, either a rural area or a high unemployment area the Secretary of Homeland Security has designated ^[16]. Rural means outside any metropolitan statistical area defined by OMB and outside the outer boundary of any city or town of 20,000 or more people ^[17]. A high unemployment area is a census tract, or a set of contiguous tracts, where the enterprise principally does business and whose labor force weighted unemployment rate reaches at least 150 percent of the national average ^[18]. Congress also set aside visas by place and project type: 20 percent for rural investment, 10 percent for designated high unemployment areas and 2 percent for infrastructure projects ^[19].

The third element is the jobs themselves: full-time employment for not fewer than 10 qualifying people ^[1], a group from which the statute leaves out "the immigrant and the immigrant's spouse, sons, or daughters" ^[20]. The fourth is the regional center program, which lets investors pool their capital and count indirect jobs, and which the statute authorizes through September 30, 2027 ^[21]. A study prepared for a regional center offering should carry that date on its timeline beside the construction and hiring dates, because the program's authorization is a dated fact like any other the plan relies on.

What 8 CFR 204.6 Asks the Petition to Prove

The regulation turns the statute into a list of evidence. A petition must show lawfully obtained capital invested, or being invested, in a new commercial enterprise that will create full-time positions for not fewer than 10 qualifying employees ^[22]. It must also show the capital placed at risk for the purpose of generating a return; a bare intention to invest, or arrangements with no present commitment, will not do ^[23]. For the jobs, the regulation offers two routes. Where the employees have already been hired, tax records, Forms I-9 or similar documents show them. Where they have not, the petition needs a comprehensive business plan showing that the nature and projected size of the enterprise will create the need for not fewer than 10 qualifying employees within the next two years, with approximate hiring dates ^[24]. A hotel that has not yet been built has no employees, so for a development project the second route is the only one, and the business plan carries the whole weight of the job evidence.

Three definitions decide what the plan may count. Full-time employment is a position requiring at least 35 working hours a week ^[25]. Part-time positions cannot be combined into a full-time one, even when their hours add up ^[26], although two or more qualifying employees who share a single full-time position do count, as long as the weekly hours are met ^[27]. And the regulatory definition of employee excludes independent contractors ^[28]. For a hotel these definitions bite at once. BEA's own ratio of full-time equivalent to total employment for accommodation is 0.89 ^[29], a sign that part-time work is common in the industry. A staffing plan that counts heads rather than full-time positions therefore overstates the direct jobs. The study's staffing schedule should list each position with its weekly hours and hiring date, and it should leave contracted services such as outsourced laundry or security out of the direct count.

Where an existing business is expanded rather than created, the regulation asks for a substantial change, defined as a 40 percent increase in net worth or in the number of employees, so that the new figure is at least 140 percent of the pre-expansion level ^[30]. A study of a hotel expansion should report the pre-expansion headcount and net worth from records rather than estimates, because the test is measured from them.

For investors in a regional center, the regulation lets reasonable methodologies show the indirect jobs, and it names multiplier tables, feasibility studies and "analyses of foreign and domestic markets for the goods or services to be exported" among them ^[31]. The paragraph was written for the original pilot program, but it remains the plainest statement in the regulation that a feasibility study is an accepted forecasting device in an EB-5 record.

One trap waits for anyone who reads the printed regulation on its own. The 1-1-25 edition of 8 CFR 204.6(f)(2) still prints a $900,000 targeted employment area amount for petitions filed on or after November 21, 2019 ^[32], which is the wording of a 2019 rule. The Policy Manual records that a federal district court vacated that rule in 2021 and that USCIS applies the regulations and policies in force before it ^[33]. For petitions filed on or after March 15, 2022, the amounts are the ones the statute now sets, $1,050,000 and $800,000 ^[34]. A study that states an investment amount should take it from the statute and the Policy Manual, and should name the edition of each source it read.

Matter of Ho: The Standard a Business Plan Is Read Against

Matter of Ho is the source of the business plan standard, and the Policy Manual still cites it for the elements a plan should contain ^[2]. Its central sentence is a test of inference: a comprehensive plan has to be detailed enough to let the agency draw reasonable inferences about the enterprise's potential to create jobs ^[35]. The decision then names what fails that test. Bare assertions, it says, leave the agency unable to tell whether the job projections are "any more reliable than hopeful speculation" ^[36].

The decision goes on to list what a plan should hold. It should describe the business, its products or services and its objectives ^[37]. It should contain a market analysis that names competitors and their strengths and weaknesses, compares their products and prices, and describes the target market ^[38]. It should list the permits and licenses obtained and, where relevant, describe the production process and supply sources ^[39], and it should detail any contracts executed for supply or distribution ^[40]. It should discuss the marketing strategy and set out the organizational structure and the experience of the people who will run the business ^[41]. And it should explain staffing needs with a hiring timetable and job descriptions, together with sales, cost and income projections and the basis for each ^[42]. Ho itself warns that the list is not exhaustive, because every business needs different information ^[43], and the Policy Manual adds that a plan need not contain every element, although more detail makes a finding that the plan is comprehensive and credible more likely ^[44]. USCIS reads the plan as a whole and asks whether it is more likely than not comprehensive and credible ^[45]. Above the whole list sits one line of the decision: "Most importantly, the business plan must be credible." ^[46]

Table 1 sets each element beside the part of a feasibility study that answers it for a hotel.

Ho also sets the bar for jobs that already exist. Forms I-9 show at best that someone checked work authorization, not that the people named ever started work ^[47], and without paystubs and payroll records showing hours, full-time employment was not established ^[48]. Where a hotel is already operating, a study's history section should therefore rest on payroll registers by position and hours rather than on a headcount.

How USCIS Counts Jobs: Direct, Indirect, Induced and the Limits

The counting rules depend on the filing route. A standalone investor counts only jobs the new commercial enterprise creates directly, as the employer ^[49]. A regional center investor filing on or after March 15, 2022 may rely on indirect jobs for up to 90 percent of the full-time positions ^[50], which the statute states as a limit on the share of the 10 job requirement that indirect jobs may meet ^[51]. For those petitions, direct jobs include employees of both the new commercial enterprise and the job-creating entity, including direct jobs estimated by valid methodologies ^[52].

Indirect jobs are held outside the enterprise but caused by it: at the job-creating entity when it is a separate company, and at the producers of the materials, equipment and services the project buys ^[53]. Induced jobs are a subset, created when the new direct and indirect workers spend what they earn ^[54]. Indirect jobs estimated by reasonable methodologies can count even when they fall outside the regional center's boundaries ^[55]. Because nobody can count these jobs one by one, USCIS relies on reasonable economic models to find them more likely than not, and may ask for evidence that they are full-time ^[56].

Construction gets its own rules. Where the estimated jobs come from construction activity lasting less than 2 years, indirect jobs may satisfy only up to 75 percent of the requirement ^[57], and the Policy Manual's chapter on project applications applies that limit to the jobs the construction creates ^[58]. Direct jobs from such construction are counted by multiplying the estimate by the fraction of the 2-year period the work lasts ^[59]. Jobs expected to last at least 2 years are generally not treated as temporary ^[60], and at removal of conditions USCIS does not exclude construction or tourism jobs by industry; the question is whether the position as described is continuous full-time employment ^[61]. Each of these rules makes the construction schedule in a study more than a planning detail. The schedule decides which limit applies and how any direct construction jobs are discounted.

Tenant jobs were once a separate trap. USCIS rescinded its tenant occupancy guidance as of May 15, 2018 for filings made before March 15, 2022 ^[62], reasoning that putting up standard office or retail space does not by itself connect tenant jobs to the investment closely enough to credit them ^[63]. The 2022 statute now lets the Secretary count jobs attributed to prospective tenants of space the investment creates or improves, if an economically and statistically valid methodology estimates them and they are not existing jobs that have moved ^[64]. For a hotel with leased retail or restaurant space, the study's demand evidence for that space is what such a method would rest on, and its tenant interviews are where the relocation question is answered.

Two last counting rules close the gaps. Investors may not, in general, claim credit for the same job ^[65], so a project with several offerings needs one job ledger across all of them. And for regional center filings, an employee of the new commercial enterprise or the job-creating entity may be treated as holding a directly created job ^[66], which is why the operating company's staffing plan, not only the multiplier, carries part of the count.

The Multiplier Source: What RIMS II Measures and Where It Overstates

The regulation and the Policy Manual speak of multipliers and input-output models without naming a product. One public source is the Bureau of Economic Analysis's Regional Input-Output Modeling System, known as RIMS II, which BEA describes as producing multipliers that economic impact studies use to estimate a project's total effect on a region ^[67]. BEA released a new set on May 5, 2026, built on 2024 regional data and the 2017 national benchmark input-output accounts ^[68], and it supplies them for a region made of one or more contiguous counties ^[69]. Other input-output models exist; this article relies on RIMS II because BEA publishes both its method and its warnings.

Two kinds of RIMS II multiplier matter for a job model. A final-demand employment multiplier converts a change in spending into a change in local jobs ^[70]. A direct-effect employment multiplier converts a change in jobs in the industry first affected into total jobs across the region ^[71]. Type I multipliers cover the direct and indirect effects ^[72], and Type II multipliers add the induced effect of household spending ^[73], which is where the induced jobs USCIS counts come from.

Whichever multiplier is used, its output counts jobs, not full-time positions. In RIMS II, employment is the number of full- and part-time jobs ^[74], and BEA says plainly that its employment multipliers run higher in industries that use more part-time workers, because they are not built on full-time equivalents ^[75]. BEA adds that the mix of full-time and part-time work can heavily influence the value of Type II multipliers ^[76]. For accommodation, BEA's own table puts the ratio of full-time equivalent to total employment at 0.89 ^[29], computed from national accounts tables for 2010 ^[77]. Set beside the 35 hour floor in the regulation ^[25], that gap is the first thing a reader of a hotel job model should look for. A model that neither adjusts for it nor explains why no adjustment is needed overstates the count, and the overstatement grows with the share of part-time work in the industry being modelled.

BEA's guide lists six assumptions behind the framework: backward linkages, fixed purchase patterns, industry homogeneity, no supply constraints, no regional feedback and no time dimension ^[78]. Three of them map onto questions a feasibility study answers. Fixed purchase patterns mean that if a business can raise its output without hiring as many workers as the model expects, the estimate is inflated ^[79]; a hotel that opens lean and grows into its rooms is that case, and the study's staffing ramp is the evidence. The absence of supply constraints means the model assumes workers are available at the going wage, and if they are not, the estimate is inflated again ^[80]; the study's labor market section speaks to it. Industry homogeneity means every business in an industry is assumed to use the same production process ^[81], so a resort with a large food and meeting operation and a select-service hotel with neither are treated alike unless the inputs separate them.

The sharpest warning in the guide concerns the input rather than the multiplier. BEA says a final-demand change "should account for the net purchases of goods and services produced by the initially affected industry" ^[82], and its own example sets the sales at a new facility against the sales it takes from existing businesses ^[83]. It also says that an impact study is only as accurate as its choice of spending change, industry and region ^[84]. It adds that giving Type II multipliers a spending change that includes purchases by workers who already live in the region counts the same spending twice ^[85], and that applying Type II multipliers to a region much larger than the one where workers spend their pay often inflates the result ^[86]. For a hotel, the net change is the room, food and meeting revenue the market study finds to be new to the region after the share the new hotel will take from hotels already operating. That share is a finding of the market study. No multiplier can produce it.

USCIS's own test points the same way. The Policy Manual says USCIS asks whether the multipliers and the geographic assumptions are reasonable ^[87], weighing among other things the area's demographic structure, its place in the project's supply chains and its socioeconomic connectivity ^[88]. RIMS II multipliers cannot be added together across regions or split into their parts ^[89], so the region chosen for the multipliers has to match the region the evidence describes. The study's market area and its labor shed are the natural evidence for that choice.

A Hotel Job Model Built From the Study's Numbers

The Policy Manual sets out how each kind of model input is tested. If the model starts from direct jobs, the investor must show that the direct jobs input is reasonable ^[90]. If it starts from spending, the spending must be shown to be reasonable, with receipts for what has been spent and projections for what will be ^[91]. If it starts from revenue, the revenue must be shown to be reasonable, with records for revenue already earned or detailed projections of sales, costs and income ^[92]. A hotel job model can draw on all three: construction spending, operating revenue and the direct staff. Each of them is a number a feasibility study produces.

The illustration below makes those links concrete. It is not a model of any real project, and its two multipliers are labelled assumptions chosen to show the arithmetic, not values published by BEA. The offering has 40 investors, so the requirement is 400 jobs. Construction spending in the region is $34,000,000 over 20 months, the stabilised hotel earns $10,500,000 a year, a quarter of that revenue is taken from existing hotels, and the staffing plan holds 62 full-time positions. The inputs for both illustrations in this article are in Table 2.

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

Table 3 counts the jobs under the two limits and then moves one study input at a time.

Read the base column from the top. The construction budget and the illustrative multiplier produce 357.0 construction jobs, but because the work lasts less than 2 years the construction-based indirect jobs may meet no more than 75 percent of the requirement, so only 300.0 count. Operations add 110.3 jobs on net new revenue, of which the 62 staff are direct and the rest indirect. After the overall limit on indirect jobs, the project counts 410.3 jobs against 400, a surplus of 10.3. That is a thin cushion, and the shock columns show where it goes. The model applies the 75 percent limit to the construction-based indirect jobs, which is how the Policy Manual's chapter on project applications reads it ^[58]; how the limits apply to a particular offering is for the economist and counsel to settle, and the study's part is to make the inputs sound.

A revenue line cut as in the second column leaves the project 6.3 jobs short. The displacement column, where the market study finds more of the hotel's business coming from competitors, leaves it 11.8 short. Restating the modelled operating jobs on BEA's full-time equivalent ratio for accommodation, with no other change, leaves it 1.9 short. Each of those three failures traces to a number the feasibility study supplies: the revenue projection, the demand analysis that measures displacement, and the staffing and hours analysis that decides how part-time the operation will be.

The construction cost column shows the opposite effect. Cutting the budget lowers the construction estimate to 321.3 jobs, yet the count does not change, because the 75 percent limit was already binding. Once that limit binds, extra construction spending adds nothing, and the offering rests on operations. The last column shows what moves the limit itself: a schedule of 2 years or more takes the construction jobs out of the short-construction rule, and the count rises to 422.0, where the overall limit on indirect jobs now binds. A study should not read that column as an option. The construction period is a fact the study evidences from the contractor's schedule and comparable projects, and a schedule stretched on paper would meet the actual completion date at removal of conditions. The bottom row shows the second consequence of the schedule: if any construction jobs were direct, a 20 month build would discount them to 0.83 of their estimate.

The lesson for a lender or a regional center reading such a model is that a job count is a stack of study outputs multiplied together. When the count clears the requirement by a narrow margin, the reader should test the revenue projection, the displacement estimate and the hours assumption before accepting the multiplier's result.

Where the Project Is: Targeted Employment Areas and the Tract Arithmetic

The place test decides the investment amount, so it belongs in the study's location chapter. An enterprise principally does business where it regularly, systematically and continuously provides the goods or services that support job creation ^[93]. The factors include where its direct jobs are, where it spends capital to create jobs, where it runs day to day and where its job-creating assets sit ^[94]. In a regional center investment, the job-creating entity's location decides the question, and that entity must principally do business in the area for the lower amount to apply ^[95]. For a hotel, all four factors usually point at the parcel, but a project with an off-site laundry, a remote reservations office or a management company elsewhere should say where each function sits.

For high unemployment areas, the statute uses census tracts and a labor force weighted average ^[18], and only the Secretary of Homeland Security or a designee may make the designation; neither a state nor a local official may ^[96]. For a standalone petition filed on or after March 15, 2022, the investor describes the tracts and the unemployment statistics so that USCIS can make a case-specific designation ^[97], and USCIS does not issue separate designation notices ^[98]. The printed text of 8 CFR 204.6(j)(6)(ii)(B) is the 2019 rule's wording, yet the Policy Manual still cites it for this point ^[99]; it asks for the boundaries, the statistics and the reliable method by which the statistics were obtained ^[100]. A designation lasts 2 years from the investment date for a standalone investor, or from the project application filing for a regional center investor, and it can be renewed ^[101].

The arithmetic matters because it is easy to get wrong. Table 4 runs the test on a site tract alone and on the site tract combined with one adjacent tract, using the tract inputs in Table 2.

The site tract alone, at 5.50 percent, fails at either comparator. Adding the adjacent tract lifts the labor force weighted rate to 7.23 percent, which clears the threshold at comparator A by 0.78 points and misses the threshold at comparator B by 0.72 points. The simple average of the two tract rates would be 8.25 percent and would clear both, but the statute asks for the weighted figure, not the simple one ^[18]. The choice of national comparator is the open question. For petitions filed before March 15, 2022, the Policy Manual recognised Census Bureau data such as the American Community Survey and BLS data such as Local Area Unemployment Statistics as acceptable sources ^[102]; the post-2022 text read for this article names no national series for the tract test. The dataset's own probe of TEA computation found the answer for one large county swinging widely with the comparator and with how adjacency is read, and its verdict was to build a location layer only as a covariate, because USCIS makes the designation itself.

A study's place chapter should therefore present the tracts, the source and vintage of each statistic, the comparator it used and the result under the other credible comparator. It should present a candidate area and the calculation behind it, and leave the designation to USCIS.

Regional Center Filings: The Project Application and Its Record

Since the 2022 statute, a regional center files an application for each investment offering before any investor files a petition based on it ^[103]. The application must include a comprehensive business plan for a specific capital investment project and a credible economic analysis of estimated job creation, based on economically and statistically valid and transparent methodologies ^[104]. The Policy Manual repeats both points in its chapter on project applications: the plan must be credible and comprehensive and describe at least the business, its products or services and its objectives ^[105], and the economic analysis must be credible, valid and transparent ^[106]. The offering documents filed with it must refer to all material investment risks of the new commercial enterprise and the job-creating entity ^[107].

The word transparent carries weight for a study. An input-output model can be transparent only if its inputs can be traced, and the inputs are the study's construction budget, operating revenue and staffing plan. A study that states each input's source, date and method lets the economist show where each number in the model came from.

The approval of a project application then binds the adjudication of later petitions by investors in the same offering ^[108], subject to exceptions that include a material change affecting eligibility ^[109]. That makes the study's assumptions part of a record that outlives the filing. A redesign that changes the room count, a new brand or a revised budget may raise the question whether the change is material, and a study that recorded its original assumptions clearly makes that question easier to answer. Regional centers also face an audit at least once every 5 years ^[110], and a proposal for the center itself must rest on reasonable predictions supported by valid and transparent forecasting tools ^[111]. Studies that keep their workpapers, data vintages and calculation files make those reviews shorter for everyone.

From the I-526 Business Plan to the I-829 Proof

The business plan filed with the petition is tested twice. At filing, it must establish that the jobs are likely within the next 2 years ^[112], and the Policy Manual deems that period to begin 6 months after the petition is adjudicated, so the plan should show the jobs in place by the end of it ^[113]. At removal of conditions, the investor must show that the capital was continuously maintained over the two years of conditional residence ^[114] and that ten full-time jobs were created, or can be expected within a reasonable time ^[115]. Jobs projected more than 3 years after the investor becomes a conditional resident usually fall outside a reasonable time ^[116].

The Policy Manual softens two points. USCIS does not require the jobs to still exist when the removal petition is decided, as long as they were created and were permanent when created ^[117], and it does not deny a removal petition solely because the business departed from the plan filed with the original petition ^[118]. But the clock is real. A construction schedule that slips, a ramp-up that runs slower than projected or an opening date pushed back by a franchise approval all move jobs toward the edge of the window. A study that sets its hiring dates against a dated construction schedule, and shows what a delay of a few months does to the count, gives the sponsor an early warning rather than a surprise at the second filing.

The Lending Record Nearest to an EB-5 Hotel

The Loan Analytics dataset holds no EB-5 outcomes. Its probe of USCIS's regional center lists found a registry of sponsors by state, with no projects, amounts, property types, job counts or outcomes, and its verdict was not to build on it. There is no public record of how EB-5 hotels perform, so the nearest public lending record for the same kind of property is SBA lending for hotels and motels, which Table 5 lays out after the next two paragraphs. These are owner-operated small business loans, a different capital source with different borrowers, and they are shown for context, not as a measure of EB-5 risk.

From FY2017 to FY2026, SBA approved 8,942 ^[119] hotel and motel 7(a) loans with a median size of $2,200,000 ^[119], and 2,885 ^[120] hotel and motel 504 loans. A franchise code appears on 64.69 percent ^[119] of the 7(a) approvals, which says that a branded hotel is the common case in the SBA record. Among hotel 7(a) loans approved in FY2017 to FY2021, charge-offs to date stand at 0.99 percent ^[119]; the whole 7(a) program for those approval years stands at 6.33 percent ^[121]. Both are approval-cohort floors: charge-offs to date over disbursed loans, with open loans still able to fail. The cancellation rate for hotel 7(a) approvals in the FY2017 to FY2024 cohort, also a floor, is 8.76 percent ^[119], a reminder that an approved hotel loan does not always close. The hotel class is built from one industry code, so limited-service and full-service hotels cannot be separated in it.

For an EB-5 study the record answers a narrow question. It shows that hotel lending is a large and settled part of the SBA record, with a body of outcomes behind it that EB-5 lacks. It does not show how a hotel financed partly with investor capital will perform, and a study should never present it that way. Readers who want the full hotel pages can find them at hotel lending data, SBA 7(a) hotel data and SBA 504 hotel data. The floors themselves are explained in the approval-cohort charge-off floor.

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.

What a Study Must Not Claim, and What Readers Should Check

A feasibility study for an EB-5 project has a defined lane. It supplies the market analysis, the construction cost and schedule, the operating projection and the staffing plan, and it tests them; those are the elements Matter of Ho lists ^[42]. It may also present a candidate targeted employment area and the calculation behind it. It must not claim that a petition will be approved, that an area is a targeted employment area, that a given number of jobs will be credited, that an investor's funds qualify as capital, or that the offering complies with securities law. Each of those is a determination for USCIS, for counsel or for securities regulators. Nor does the study value the hotel, and it should not report an appraised value it did not make.

A lender, a regional center or an investor reading an EB-5 study can work through a short set of checks drawn from the sources above.

1. Does the investment amount come from the statute and the current Policy Manual, not from the printed 2019 rule text ^[34]?
2. Does the staffing plan count full-time positions of 35 hours or more, with hiring dates, rather than heads ^[25]?
3. Is the revenue fed into the job model net of the business the hotel takes from existing hotels, and does the market study measure that share ^[83]?
4. Is the job model's region the same region the market and labor evidence describes, and are its multipliers for that region alone ^[89]?
5. Does the model address the gap between the multiplier's full- and part-time jobs and the regulation's full-time positions ^[75]?
6. Does the construction schedule rest on evidence, and does the model apply the construction limits that schedule triggers ^[57]?
7. Does the place chapter state its tracts, its statistics, their vintage and its comparator, and does it leave the designation to USCIS ^[98]?
8. Does the financing chapter describe the investor capital the same way the offering documents do ^[15]?

Readers working on the senior debt for the same building will find the bank's side in the conventional bank loan method, and readers comparing job tests across programs will find the SBA 504 version in the SBA 504 method. For hotel market evidence, the archived guides on how to determine demand for a new hotel development and on determining project costs for a hotel project cover the two inputs that drive most of the job count, and a full worked hotel study under another program is in the USDA Business and Industry hotel case.

Related Pages on Loan Analytics

• The Loan Analytics methodology series (all articles)
• How a feasibility study is done: the Loan Analytics methodology
• Loan data by state
• Loan data by asset class
• Worked examples (case studies)
• Conventional Bank Loan Method: Appraisal Rules, Supervisory LTV Limits and the Study a Bank Reads
• SBA 504 Method: The Job Test, the Project Split and What a Study Must Show the CDC

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