Two EB-5 deals can cost the same $800,000 and end at the same green card while sharing almost nothing else. A Regional Center investor buys a unit in a pooled fund and counts indirect jobs an economist modeled years before construction finished. Management belongs to somebody else. A direct investor buys into one operating company and personally carries the risk that ten real people are on payroll when Form I-829 is filed. Below are two composite deals, assembled from ordinary market terms rather than from any single offering, set against each other on the four things that decide outcomes: job math, the business plan, fees and exit.
Deal A: a $60 million Regional Center tranche
Picture a $210 million mixed use development in a high unemployment urban census tract. The developer raises $60 million of EB-5 money from 75 investors at $800,000 each, through a new commercial enterprise that lends the pooled capital to the project company at around 6 percent. Investors hold limited partnership units with no vote over construction decisions or refinancing. Functionally you are a lender.
The economic report allocates 12 to 14 jobs per investor, giving the offering a cushion above the required 10. Construction spending and hotel operations both feed that model. In deals of this shape the construction component usually supplies the largest share of the job number, which is exactly where the 2022 rules bite hardest.
Your money moves before you do. The Regional Center files Form I-956F, the application for approval of an investment in a commercial enterprise, and only after that can you file Form I-526E, the immigrant petition by regional center investor. If the I-956F is denied, or the project was described inaccurately in it, every investor in the tranche is hit at once. Seventy five people, one document.
Deal B: an $800,000 stake in one operating business
Now a specialty food manufacturer in a rural county. Total capitalization is $2.4 million, of which one EB-5 investor supplies $800,000 for a 30 percent membership interest plus a seat on the management committee. There is no fund and no economic model.
Ten people have to be on payroll. Not modeled, not projected. Ten workers at 35 hours a week or more, each one a US citizen, a permanent resident or another immigrant authorized to work here, and none of them your spouse or children. Somebody on a nonimmigrant visa does not count toward the ten. The investor files Form I-526, the standalone immigrant petition by investor, then carries the hiring plan personally through two years of conditional residence.
Job math forces the choice
Direct EB-5 counts only jobs inside the new commercial enterprise. Regional Center investors may also count indirect and induced jobs generated outside it, estimated with input output models such as RIMS II or IMPLAN, subject to limits the 2022 Act added. Construction activity lasting less than two years is capped at 75 percent of the jobs counted. Estimates built on prospective tenant occupancy no longer count at all. USCIS describes the methodology it accepts in Volume 6, Part G of the USCIS Policy Manual, covering immigrant investors.
Run the arithmetic before you fall in love with a business. Ten full-time staff at 35 hours a week and $16 an hour cost about $291,000 a year in wages alone, before payroll taxes or benefits. Across the two year sustainment period that is roughly $580,000 of an $800,000 investment, which means the business has to generate real revenue rather than burn capital on payroll. Deal B works only if the enterprise sells something.
Our guide to judging whether a project can really deliver ten jobs covers the questions to put to an economist's report.
Who writes the business plan, and what it has to survive
Both deals need a plan meeting the Matter of Ho standard, the 1998 administrative decision that still defines what USCIS treats as comprehensive. Hiring schedules by position and year. Market analysis. Cost breakdowns tied to the actual construction or equipment budget. Financial projections an adjudicator can trace back to stated assumptions.
Who commissions it differs, and that difference matters more than most investors expect. In Deal A the Regional Center hires the business plan writer and the economist, folds the cost into the administrative fee, and files both documents with the I-956F. You are auditing a document you had no hand in. In Deal B you hire the EB-5 business plan writer yourself, so the quality of one vendor becomes a direct immigration risk. Template plans draw Requests for Evidence, and an RFE response usually costs more than a properly researched plan would have. Buy the good one.
One warning about Deal A. An economic report can be methodologically clean and still describe a project whose senior financing has not closed. Approval of an I-956F tells you the job model was acceptable to USCIS. It tells you nothing about whether the senior lender will fund the remaining $150 million of the capital stack, and a stalled project with an approved I-956F still leaves your money locked in a hole in the ground.
Fees, returns and what comes back
Deal A charges an administrative fee on top of the $800,000, $60,000 in this composite, and pays a nominal preferred return, half a percent a year here. Both figures move from offering to offering and both are stated in the documents, so read those rather than the brochure. Exit means repayment of the loan after a refinancing or a sale. It normally arrives later than the offering suggested. Redeployment clauses let the fund move your capital into a second investment if the loan repays before your immigration process finishes, which can extend the hold by years.
Deal B has no administrative fee and no preferred return. You take distributions if the business earns them. Getting your capital out means finding someone willing to buy a 30 percent minority stake in a small food plant, which is a search that can run for years and end at a discount you did not plan for. There is no queue of buyers.
Both are illiquid for a long time. Our page on how and when an EB-5 investment is repaid goes through the mechanics of each.
Deadlines sitting over both structures
Capital has to stay invested for at least two years from the date it is made available to the new commercial enterprise, a clock that can finish before conditional residence even starts. Conditional residence is a separate two years. Rural projects get priority processing under the 2022 Act, so a rural version of either deal can move faster than an urban one at the same price. Regional Center authorization currently runs to 30 September 2027, and petitions filed by 30 September 2026 are grandfathered against a lapse. That deadline matters to Deal A investors far more than to Deal B, because a standalone direct petition does not depend on Regional Center authorization at all.
Both minimums rise with the first inflation adjustment on 1 January 2027.
Which structure suits which investor
Pick Deal A if you want passivity, want the job cushion an economic model provides, and can accept that your outcome rests with people you will never meet. Pick Deal B if you have run a business in that sector and intend to live near it. Watching your own payroll is the entire point of the structure. An investor sitting in another country running a US restaurant by video call is the worst version of the direct path. The regulations require you to be engaged in management, either through day to day control or through policy formation, and a title on an org chart does not satisfy that.
Read what USCIS expects from a direct investor day to day before assuming Deal B is passive, and how a Regional Center actually works before assuming Deal A is safe. Neither structure protects you from a bad sponsor.