Cold Storage Feasibility Study
The SBA loan record for cold storage projects nationally and by state, and what a feasibility study for one has to test.
Daniel Smith, MAI
Cold storage feasibility study consultant
Cold storage reaches an SBA credit officer as one of the thinnest classes in the loan record. From FY2017 to FY2026 the SBA approved 49 refrigerated warehousing loans through 7(a), worth $66.1 million, whose median is $615,000, and 39 through 504, worth $102.3 million, whose median is $2.1 million. In no year of that span did 7(a) approve more than 10 of them or 504 more than 7, and in FY2024 the 504 count was 0. A lender reading a cold storage request therefore has very little history of its own to set it against, and the feasibility study becomes the main evidence that the rooms will fill and the debt will be paid. Loan Analytics is the cold storage feasibility study consultant that writes that evidence for the credit file: a report from an author with no stake in the deal, carrying the signature of Daniel Smith, MAI, which places the project against every table on this page and then proves its own case from the trade area.
The record also shows the buyer of a study where a lender's doubt will sit. The cancellation floors are 20.9% for 7(a) cold storage approvals and 15.2% for 504, and neither program prints a charge-off rate, because neither funded cohort reaches 30 loans. The measured risk is the approval that never closes, and the loss history is simply missing. Our study is built around what keeps a refrigerated warehouse approval alive and the loan performing afterwards: the processors, distributors, grocers and growers who will commit to pallet positions or dedicated rooms, at which temperatures and for how long; the public cold space already serving them and how full it runs; the full cost of the insulated envelope, the refrigeration plant, freezer floors, docks and standby power; a utility budget run again at higher power prices; debt coverage with the building only partly occupied; and the value the structure keeps for a buyer who needs no refrigeration at all.
Where the project sits and how large the loan is decide which part of the record a lender compares it with. California leads the states with 18 loans; Texas carries the most dollars, $28.6 million on 11 loans; Wisconsin (10 loans, $23.7 million) and Massachusetts (7 loans, $20.5 million) follow, and no state has enough loans to print a median or a floor of its own. By size, the 504 middle runs from $1.2 million to $4.9 million, while 7(a) divides between small credits and building-scale loans, with 13 of its 49 in the band just below the top. Loan Analytics places the proposed loan in that structure, explains where it departs from it, and writes the answers in terms a credit committee can check. Describe the site, the rooms and the financing on the scope form directly below; our answer is a written scope and timeline for this building.
Request Scope & Timeline
A thin record with large 504 tickets
Cold storage is a thin class in the SBA files. From FY2017 through FY2026, 7(a) approved 49 cold storage loans for $66.1 million, at a median of $615,000, and 504 approved 39 for $102.3 million, at a median of $2.1 million. In several years only a handful of loans were made, and in FY2024 the 504 program approved no cold storage loans at all. The 504 median reflects the price of a refrigerated building more closely than the 7(a) median does, and it is the better benchmark for a borrower sizing a project.
Floors, and what is withheld
Cancellation after approval is high: the 7(a) floor is 20.9% and the 504 floor 15.2%, measured on loans approved in fiscal 2017 to 2024. Charge-off rates are withheld for both programs, because neither program reaches 30 disbursed loans in the FY2017 to FY2021 window, the minimum we require before printing a rate. That gap should not be read as a clean loss record. For a lender, the absence of a measured loss rate means the credit case has to stand on the project's own contracts and collateral.
Contracts, power and a special building
Demand in cold storage is contractual. A feasibility study has to show who will fill the rooms: food producers, distributors, grocers or processors in the area, and whether they will sign multi-year agreements for pallet positions or dedicated space, at which temperatures. It should map competing public refrigerated warehouses and the gap they leave. Cost has to include insulated panels, the refrigeration plant, backup power, dock doors and the under-slab heating that freezer floors need, along with a realistic utility budget. Debt service coverage depends on power cost and handling fees as much as on storage rent, so the study should stress energy prices. The appraisal must address a specialized building with few local buyers, and what part of its cost would survive a sale to a dry warehouse user.
On counts this small, one project can swing a year: 7(a) lending reached $14.4 million on 5 loans in FY2018, against $1.4 million on 4 loans in FY2017. The national record therefore says more about how these projects are financed than about the health of the trade. Loans are classed by the borrower's own industry, refrigerated warehousing, so a food processor that builds a freezer for its own product is counted in its own industry and not here.
SBA loan record for cold storage, United States, FY2017 to FY2026
| Program | Loans | Dollars | Median loan | Cancelled (floor) | Charged off (floor) |
|---|---|---|---|---|---|
| SBA 7(a) | 49 | $66.1 million | $615,000 | 20.9% | not reported: fewer than 30 loans in the cohort |
| SBA 504 | 39 | $102.3 million | $2.1 million | 15.2% | not reported: fewer than 30 loans in the cohort |
Download this table (CSV). Cite as: Loan Analytics, SBA loan record for cold storage, United States, FY2017 to FY2026, https://www.analytics.loan/cold-storage-feasibility-study#record-C13.
Cancellation counts FY2017 to FY2024 approvals; charge-off counts FY2017 to FY2021 disbursed loans to date. Both are floors, and a rate needs at least 30 loans in its cohort.
SBA 7(a) cold storage loans approved nationally, by fiscal year
| Fiscal year | Loans | Dollars |
|---|---|---|
| FY2017 | 4 | $1.4 million |
| FY2018 | 5 | $14.4 million |
| FY2019 | 3 | $1.9 million |
| FY2020 | 10 | $7.6 million |
| FY2021 | 5 | $8.8 million |
| FY2022 | 4 | $5.8 million |
| FY2023 | 6 | $7.8 million |
| FY2024 | 6 | $5.2 million |
| FY2025 | 3 | $8.1 million |
| FY2026 | 3 | $5.1 million |
Download this table (CSV). Cite as: Loan Analytics, SBA loan record for cold storage, United States, FY2017 to FY2026, https://www.analytics.loan/cold-storage-feasibility-study#record-C13.
SBA 504 cold storage loans approved nationally, by fiscal year
| Fiscal year | Loans | Dollars |
|---|---|---|
| FY2017 | 7 | $19.8 million |
| FY2018 | 4 | $3.4 million |
| FY2019 | 6 | $13.4 million |
| FY2020 | 2 | $3.2 million |
| FY2021 | 7 | $22.5 million |
| FY2022 | 5 | $12.3 million |
| FY2023 | 2 | $5.9 million |
| FY2024 | 0 | $0 |
| FY2025 | 5 | $18.4 million |
| FY2026 | 1 | $3.4 million |
Download this table (CSV). Cite as: Loan Analytics, SBA loan record for cold storage, United States, FY2017 to FY2026, https://www.analytics.loan/cold-storage-feasibility-study#record-C13.
Source: U.S. Small Business Administration, 7(a) and 504 loan-level FOIA files (data.sba.gov), as of 2026-09-30.
Related feasibility study pages
SBA lending to cold storage projects, FY1991 to FY2026
Every SBA 7(a) and 504 loan approved for a cold storage project since FY1991, by fiscal year, nationally.
Every year has a 7(a) loan, none has many
Since FY1991 the SBA has approved 171 cold storage loans through 7(a), worth $161.7 million, and 118 through 504, worth $199.1 million. The 7(a) count does not fall to zero in any fiscal year of the table: its low is 2, in FY1995, FY1997, FY1999 and FY2008, and its high is 12, in FY2011. Refrigerated warehousing is a trade the program touches every year and never in volume. That shapes the lender's position. A cold storage request is rarely one of a batch in its year, so the credit officer has no current set of comparable approvals to read it beside, and the study has to supply the comparison itself, from the long record on this page and from the market around the site.
504 carries the dollars
The two programs divide the class in opposite ways. 7(a) made more cold storage loans in the full record, 171 against 118, and in FY2017 to FY2026, 49 against 39. 504 carried more dollars in both spans: $199.1 million against $161.7 million since FY1991, and $102.3 million against $66.1 million from FY2017 on. The years in which the long table prints both programs' dollars show the same split within a single year. In FY2012, 11 loans under 7(a) came to $15.2 million and 6 under 504 to $15.3 million. In FY2021, 5 loans under 7(a) came to $8.8 million and 7 under 504 to $22.5 million, the largest annual dollar figure printed anywhere on this page. A study for a 504 cold storage project is written for the program that finances the larger refrigerated buildings, and its construction budget and fallback value carry more of the credit decision than they would on a typical 7(a) file in the class.
Runs, peaks and a quiet stretch
The counts come in runs. The busiest 7(a) stretch is FY2010 to FY2012, with 7, 12 and 11 loans, after 2 in FY2008 and 3 in FY2009; by FY2015 and FY2016 the count is back to 3 a year. The next 7(a) high is FY2020 at 10, the most in any year after FY2012. 504 peaks at 7 twice, in FY2017 and FY2021, and the two programs seldom rise together: FY2011, the 7(a) high, has a single 504 loan, and FY2020, the recent 7(a) high, has 2. The latest years sit low on both sides, with 3 under 7(a) in each of FY2025 and FY2026, and 5 and then 1 under 504. FY2024 is the one year in the table with no 504 cold storage approval, while FY1991 is the one year the table cannot measure for 504. A borrower applying now applies in a quiet period for the class, and a lender will want the study to show demand from customer commitments, signed or in negotiation, rather than from any trend in the record.
What the dollar columns leave out
Most annual dollar cells in the long table are not printed. Where a year has fewer than 5 loans the cell says so, and in a number of years with 5 or more loans the dollars are withheld. Printed annual dollars remain only for FY2010, FY2012 and FY2021 in both programs, and for FY2024 under 504, where the table shows 0 loans and $0. The span totals do print, and they are the figures a lender can use: $161.7 million on 171 7(a) loans and $199.1 million on 118 504 loans since FY1991, and $66.1 million and $102.3 million from FY2017 to FY2026. A study should not try to rebuild the missing years, and it should never treat a withheld cell as a small one. What the counts do support is a statement about frequency: in a typical year either kind of cold storage loan is approved only a few times across the whole country, and a credit committee will weigh the proposed loan as a case with few precedents.
Approvals, not closings, and counts that may be floors
Two notes under the table limit how a study may quote it. The counts are gross approvals, so they include approvals that were later cancelled; with cancellation floors of 20.9% under 7(a) and 15.2% under 504, the funded record is thinner than the approval counts make it look. And a class count is itself a floor in any year in which too many of SBA's loan records lack a usable industry code, so the table may understate some years. Neither note changes the shape of the record, a thin and steady class with 504 carrying the larger loans, but both limit what it can prove. A study that cites the table should call the counts approvals, mark them as floors where the source note applies, and build its demand case so that it does not depend on them.
Whose loans these are
The class counts borrowers whose own industry code is refrigerated warehousing. A processor, packer or grower that builds a freezer or cooler for its own product is classed in its own industry, so the 171 7(a) loans and 118 504 loans here describe businesses whose trade is cold storage itself. That decides the first question a study must answer. A cold storage operator sells space, handling and time at temperature, and its feasibility rests on customers who are not the borrower; an owner-user's rests on its own production. For an operator, the lender will expect the study to identify the outside customers, by name where agreements exist and by documented demand where they do not, because the loans on this page belong to businesses whose revenue depends entirely on those customers staying.
| Fiscal year | SBA 7(a) loans | SBA 7(a) dollars | SBA 504 loans | SBA 504 dollars |
|---|---|---|---|---|
| FY1991 | 3 | fewer than 5 loans | not measured | not measured |
| FY1992 | 6 | withheld | 4 | fewer than 5 loans |
| FY1993 | 3 | fewer than 5 loans | 2 | fewer than 5 loans |
| FY1994 | 3 | fewer than 5 loans | 1 | fewer than 5 loans |
| FY1995 | 2 | fewer than 5 loans | 5 | withheld |
| FY1996 | 4 | fewer than 5 loans | 4 | fewer than 5 loans |
| FY1997 | 2 | fewer than 5 loans | 1 | fewer than 5 loans |
| FY1998 | 4 | fewer than 5 loans | 4 | fewer than 5 loans |
| FY1999 | 2 | fewer than 5 loans | 1 | fewer than 5 loans |
| FY2000 | 3 | fewer than 5 loans | 2 | fewer than 5 loans |
| FY2001 | 8 | withheld | 4 | fewer than 5 loans |
| FY2002 | 7 | withheld | 3 | fewer than 5 loans |
| FY2003 | 6 | withheld | 4 | fewer than 5 loans |
| FY2004 | 5 | withheld | 4 | fewer than 5 loans |
| FY2005 | 3 | fewer than 5 loans | 2 | fewer than 5 loans |
| FY2006 | 5 | withheld | 4 | fewer than 5 loans |
| FY2007 | 5 | withheld | 1 | fewer than 5 loans |
| FY2008 | 2 | fewer than 5 loans | 3 | fewer than 5 loans |
| FY2009 | 3 | fewer than 5 loans | 6 | withheld |
| FY2010 | 7 | $7.2 million | 6 | $5.7 million |
| FY2011 | 12 | withheld | 1 | fewer than 5 loans |
| FY2012 | 11 | $15.2 million | 6 | $15.3 million |
| FY2013 | 5 | withheld | 3 | fewer than 5 loans |
| FY2014 | 5 | withheld | 4 | fewer than 5 loans |
| FY2015 | 3 | fewer than 5 loans | 3 | fewer than 5 loans |
| FY2016 | 3 | fewer than 5 loans | 1 | fewer than 5 loans |
| FY2017 | 4 | fewer than 5 loans | 7 | withheld |
| FY2018 | 5 | withheld | 4 | fewer than 5 loans |
| FY2019 | 3 | fewer than 5 loans | 6 | withheld |
| FY2020 | 10 | withheld | 2 | fewer than 5 loans |
| FY2021 | 5 | $8.8 million | 7 | $22.5 million |
| FY2022 | 4 | fewer than 5 loans | 5 | withheld |
| FY2023 | 6 | withheld | 2 | fewer than 5 loans |
| FY2024 | 6 | withheld | 0 | $0 |
| FY2025 | 3 | fewer than 5 loans | 5 | withheld |
| FY2026 | 3 | fewer than 5 loans | 1 | fewer than 5 loans |
| FY1991 to FY2026 | 171 | $161.7 million | 118 | $199.1 million |
| FY2017 to FY2026 | 49 | $66.1 million | 39 | $102.3 million |
Source: SBA 7(a) and 504 loan-level FOIA files as of 2026-09-30 (data.sba.gov; gross approvals, cancelled approvals included), aggregated by Loan Analytics; each figure is kept with its source file and retrieval date. A class figure counts the loans whose NAICS code places them in the class; in years when SBA's files leave the code missing or unplaceable for more than one loan in a hundred, the figure is a floor, and where too few loans are coded the year is not measured. Download this table (CSV). Cite as: Loan Analytics, "SBA 7(a) and 504 loans to cold storage projects by fiscal year", https://www.analytics.loan/cold-storage-feasibility-study#sba-record-sba-by-year.
Every state ranked for cold storage lending
SBA loans to cold storage projects in each state from FY2017 to FY2026, ranked, with loan size, program split and the cohort floors where at least 30 loans support them.
California by count, Texas by dollars
California leads the state table with 18 cold storage loans and $26.5 million from FY2017 to FY2026. Ranked by dollars, Texas is first, with $28.6 million on 11 loans, more than California's total on fewer loans. Wisconsin is third on both measures, 10 loans and $23.7 million, and Massachusetts fourth on both, 7 loans and $20.5 million, close to Wisconsin's dollars and not far below California's on far fewer loans. The gap between the count ranking and the dollar ranking says what kind of projects each state has financed: the Texas, Wisconsin and Massachusetts rows carry larger loans on average than California's. A building-scale refrigerated warehouse in Texas or Massachusetts joins a state record weighted toward large loans, while a California study should expect its lender to read a state row in which the loans run smaller on average.
One leader in each region
California, Texas, Wisconsin and Massachusetts each sit in a different Census region, so the class has no single home region. Within each region, though, the leader holds at least twice the count of the next state. In the West, California's 18 loans stand beside 2 each for Washington and Utah and 1 each for Arizona, Idaho and Nevada; in the Midwest, Wisconsin's 10 stand beside Ohio's 5, 2 each for Indiana, Michigan and Missouri, and 1 for Iowa; in the Northeast, Massachusetts' 7 stand beside 2 for New Jersey and 1 each for New York, Pennsylvania and Connecticut. In those three regions only the leader's row and Ohio's print a dollar total; every other row sits under 5 loans and prints none. The South is more spread: Texas leads with 11, Florida adds 5, Virginia 3 and Oklahoma 2, and Arkansas, North Carolina and Tennessee 1 each. For a project in a region's leading state, the state row is the comparison. For a project elsewhere in that region, the leader's row is the nearest record with any depth, and a study can cite it as regional context while saying plainly that it describes another state's market. No region in the table reaches the 30 loans this page requires before it prints a rate.
Same count, different scale
Counts alone mislead in this class. Florida and Ohio both show 5 loans, but Florida's come to $6.2 million and Ohio's to $726,100, which is less, all together, than the $1.2 million at which the 504 lower quartile sits. Ohio's record is a record of small credits. Below those two rows the dollar column stops printing: Virginia's 3 loans and the 2 apiece in Indiana, Michigan, Missouri, New Jersey, Oklahoma, Utah and Washington show 'fewer than five loans' where a total would stand, so once a state's count drops under 5 the table no longer separates a record of small credits from a record of building loans. A study for a new refrigerated building in Ohio cannot point to a state loan of comparable size in this table and should say so. In a state with 2 or 3 loans the table supports even less: it shows that the state has financed the class a few times, not at what scale, and the study should neither guess the size of those loans nor offer them to the lender as a benchmark. The question of scale then moves to the national rows, where the program medians and quartiles print and most size bands carry a dollar total.
One loan, or no row
Arkansas, Arizona, Connecticut, Iowa, Idaho, North Carolina, Nevada, New York, Pennsylvania and Tennessee each show a single cold storage loan from FY2017 to FY2026, and states such as Georgia, Illinois, Minnesota, Colorado and Oregon have no row in the table. In those states the lender's question is not how the project compares with the state's record but whether there is a state record at all, and the answer is that there is almost none. The study then carries the comparison at the national level, with the program medians, quartiles, size bands and cancellation floors, and then assembles the local case through fieldwork: the food businesses that would contract for space, the public refrigerated warehouses within trucking distance and their occupancy, the rates they quote, and the electrical service the site can obtain. A single loan is a data point, not a benchmark, and a study should not present it as one.
What no state row prints
Every state row stops at counts and dollars. The median, program split and floor columns carry no values; in their place every row prints the note 'too few loans for these measures', California's with 18 loans included, so the national rows are the only place on this page where a cold storage median, quartile or cancellation floor is printed. A state-specific study has to be explicit about that limit. It can say that a Texas project joins a state record of 11 loans and $28.6 million, or a Wisconsin project one of 10 loans and $23.7 million, but it cannot say how cold storage loans in either state have performed, because the table does not measure it, and it should not borrow a rate from another class or another state to fill the gap. In every state the credit case rests on the project's own evidence, with the state row as context and the national floors as the only performance measure the lender can cite.
| State | Loans | Dollars | Median loan | SBA 7(a) loans | SBA 504 loans | Cancelled (floor) | Charged off (floor) |
|---|---|---|---|---|---|---|---|
| California | 18 | $26.5 million | too few loans for these measures | ||||
| Texas | 11 | $28.6 million | too few loans for these measures | ||||
| Wisconsin | 10 | $23.7 million | too few loans for these measures | ||||
| Massachusetts | 7 | $20.5 million | too few loans for these measures | ||||
| Florida | 5 | $6.2 million | too few loans for these measures | ||||
| Ohio | 5 | $726,100 | too few loans for these measures | ||||
| Virginia | 3 | fewer than five loans | too few loans for these measures | ||||
| Indiana | 2 | fewer than five loans | too few loans for these measures | ||||
| Michigan | 2 | fewer than five loans | too few loans for these measures | ||||
| Missouri | 2 | fewer than five loans | too few loans for these measures | ||||
| New Jersey | 2 | fewer than five loans | too few loans for these measures | ||||
| Oklahoma | 2 | fewer than five loans | too few loans for these measures | ||||
| Utah | 2 | fewer than five loans | too few loans for these measures | ||||
| Washington | 2 | fewer than five loans | too few loans for these measures | ||||
| Arkansas | 1 | fewer than five loans | too few loans for these measures | ||||
| Arizona | 1 | fewer than five loans | too few loans for these measures | ||||
| Connecticut | 1 | fewer than five loans | too few loans for these measures | ||||
| Iowa | 1 | fewer than five loans | too few loans for these measures | ||||
| Idaho | 1 | fewer than five loans | too few loans for these measures | ||||
| North Carolina | 1 | fewer than five loans | too few loans for these measures | ||||
| Nevada | 1 | fewer than five loans | too few loans for these measures | ||||
| New York | 1 | fewer than five loans | too few loans for these measures | ||||
| Pennsylvania | 1 | fewer than five loans | too few loans for these measures | ||||
| Tennessee | 1 | fewer than five loans | too few loans for these measures |
Source: Loan Analytics' SBA loan record from SBA's 7(a) and 504 loan-level files, approvals FY2017 to FY2026 as of 2026-09-30; rates are floors and need at least 30 loans in their cohort; dollars are not shown where fewer than five loans would reveal a single loan. Download this table (CSV). Cite as: Loan Analytics, "SBA cold storage loans by state, FY2017 to FY2026", https://www.analytics.loan/cold-storage-feasibility-study#states-states.
Loan sizes, program split and cohort floors for cold storage loans
How large cold storage loans run under each SBA program, how they fall into size bands, and how often they were cancelled or charged off.
Two programs, two price points
The median 504 cold storage loan, $2.1 million, is far above the 7(a) median of $615,000, and the 504 lower quartile, $1.2 million, is itself above the 7(a) median. The difference runs through both distributions: the 504 middle spans $1.2 million to $4.9 million, the 7(a) middle $175,000 to $2.3 million. At the top the programs meet, because the 7(a) upper quartile of $2.3 million is above the 504 median, so the largest 7(a) loans reach into the middle of the 504 range. This is the first sizing question a study answers for a borrower. A refrigerated building financed through 504 at a typical size is a loan within that program's $1.2 million to $4.9 million middle; a 7(a) request of the same size sits in the upper part of 7(a)'s own record, where the lender has fewer comparable loans to read it against and will look harder at the construction and collateral case.
7(a) splits in two
The 7(a) size bands show why its median describes few of its loans. The smallest band holds 12 loans for $1.1 million in all, and the band just below the top holds 13 loans for $46.5 million, while the bands between them hold 9, 8 and 7. The median, $615,000, falls in a middle band of 8 loans worth $5.5 million. 7(a) cold storage lending is two kinds of credit filed under one class: small loans in a band where 504 has no loan at all, and building-scale loans that carry $46.5 million of the program's $66.1 million. No 7(a) cold storage loan sits in the top band. A study for a 7(a) request should say which of the two groups the loan belongs to and benchmark it against that group, not against the program median; for a small credit the work centres on the operating business, its customers and its equipment, and for a building loan on the construction budget, the power supply and the fallback value.
504 climbs into its upper bands
504 cold storage lending gathers in its larger bands. No 504 loan falls in the smallest band, and the band above it holds 2. From there the count rises with size: 5 loans for $3.8 million, then 10 for $13.8 million, then 19 for $67.8 million in the band just below the top, the busiest band in either program. Above it, 3 loans pass the top boundary, the largest cold storage loans in the record; their dollars do not print, since the band holds fewer than 5 loans, and under 7(a) the same band shows 0 loans and $0. At the largest scale in this record, a cold storage project has been a 504 project. A study for one should expect the lender to read it beside a very small group of comparable approvals, and to rely on the study's cost and value work more than on any precedent.
Where the class keeps its dollars
Each program's busiest band is the one that runs over $2 million to $5 million, with 19 approvals on the 504 side and 13 on the 7(a) side, and it holds most of each program's dollars: $67.8 million of the 504 total of $102.3 million, and $46.5 million of the 7(a) total of $66.1 million. SBA's cold storage exposure therefore sits in building loans, and that is the size at which a study's construction budget, its utility budget and its estimate of the building's value without refrigeration decide the credit. A lender reviewing a loan in that band will want each cost line for the insulated panels, the refrigeration plant, the heated or insulated slab and the standby generator supported by a dated bid, because those items separate a refrigerated building's cost from what a dry warehouse buyer would pay for the same walls.
Cancellation floors and the cohort behind them
The cancellation floors, 20.9% for 7(a) and 15.2% for 504, are measured on approvals from FY2017 to FY2024, and they are the only performance measures this page prints for cold storage. Both are floors, so the share of approvals that never went forward may be higher. Cancellation runs higher under 7(a), the program with the smaller loans, than under 504. For a borrower, the floor is the figure here that matters most in practice: an approval in this class is not yet a closing. The table does not record why cold storage approvals fall away, but a study can remove the causes a lender can see in advance: customer commitments never signed, a construction budget without dated bids for the plant and envelope, an electrical service assumed rather than confirmed by the utility, and a valuation that falls short of cost because the building's refrigerated value was never supported.
Why no charge-off rate prints
The charge-off columns, including the measure within 5 years, print no rate for either program, each noting that its cohort falls short of 30 loans, and the yearly counts show why. The charge-off cohort is loans disbursed from approvals of FY2017 to FY2021. In those years 7(a) approved 4, 5, 3, 10 and 5 cold storage loans, and 504 approved 7, 4, 6, 2 and 7; neither program's counts for those years come to 30 even before cancelled approvals are taken out. A missing rate is not a low rate. The lender has no measured loss history for cold storage to price against, so the study's coverage test, its evidence of contracted demand and its collateral fallback carry the weight that a printed loss rate carries in a deeper class. A study that leaves the lender to infer safety from the blank column has not done its job.
| Program | Loans | Dollars | Lower quartile | Median loan | Upper quartile | Cancelled (floor) | Charged off (floor) | Charged off within 5 years |
|---|---|---|---|---|---|---|---|---|
| SBA 7(a) | 49 | $66.1 million | $175,000 | $615,000 | $2.3 million | 20.9% | not reported: fewer than 30 loans in the cohort | not reported: fewer than 30 loans in the cohort |
| SBA 504 | 39 | $102.3 million | $1.2 million | $2.1 million | $4.9 million | 15.2% | not reported: fewer than 30 loans in the cohort | not reported: fewer than 30 loans in the cohort |
Source: Loan Analytics' SBA loan record from SBA's 7(a) and 504 loan-level files, approvals FY2017 to FY2026 as of 2026-09-30; rates are floors and need at least 30 loans in their cohort; dollars are not shown where fewer than five loans would reveal a single loan. Download this table (CSV). Cite as: Loan Analytics, "SBA cold storage loans by program, FY2017 to FY2026", https://www.analytics.loan/cold-storage-feasibility-study#loan-sizes-programs.
| Loan size | SBA 7(a) loans | SBA 7(a) dollars | SBA 504 loans | SBA 504 dollars |
|---|---|---|---|---|
| $150,000 or less | 12 | $1.1 million | 0 | $0 |
| Over $150,000 to $350,000 | 9 | $2.5 million | 2 | fewer than 5 loans |
| Over $350,000 to $1 million | 8 | $5.5 million | 5 | $3.8 million |
| Over $1 million to $2 million | 7 | $10.4 million | 10 | $13.8 million |
| Over $2 million to $5 million | 13 | $46.5 million | 19 | $67.8 million |
| Over $5 million | 0 | $0 | 3 | fewer than 5 loans |
Source: SBA 7(a) and 504 loan-level FOIA files as of 2026-09-30 (data.sba.gov; gross approvals, cancelled approvals included), aggregated by Loan Analytics; each figure is kept with its source file and retrieval date. Download this table (CSV). Cite as: Loan Analytics, "SBA cold storage loans by size band, FY2017 to FY2026 approvals", https://www.analytics.loan/cold-storage-feasibility-study#loan-sizes-size-bands.
Questions buyers ask
Who prepares a cold storage feasibility study that a lender will accept?
A lender wants the study from an author with no stake in the project, written as a third-party report it can keep in the credit file. Loan Analytics writes refrigerated warehouse studies for each route a cold storage loan can take (SBA 7(a), SBA 504, a USDA guarantee or plain bank credit), and Daniel Smith, MAI signs each refrigerated warehouse report. Our work begins with the national record on this page, 49 7(a) loans and 39 504 loans from FY2017 to FY2026, because the bank may have no refrigerated warehouse loan of its own to compare with. It then builds the local case: the customers who will contract for space, the cold storage already serving them, the cost of the building and plant, the power budget, coverage under stress, and the value the walls keep for an occupant with no use for refrigeration.
Does a lender require a feasibility study for a cold storage loan?
Not by one fixed rule, but a refrigerated warehouse fits the circumstances in which a study is asked for. This site's 2026 fee survey cites the SBA's authority to ask for one, and among the circumstances it lists are highly specialized project property and a unique market concept; it adds that lenders order studies under their own credit policies for special-purpose property and start-ups. A cold storage building, with its insulated envelope, refrigeration plant and freezer floors, is specialized by construction. The record supplies a second reason: with cancellation floors of 20.9% under 7(a) and 15.2% under 504, and no charge-off rate printed for either program, a lender has little history to rely on. And 19 of the 39 504 loans sit in the band just below the top, where the lender is weighing a large building with a narrow resale market, the very question a study is ordered to answer.
What will a refrigerated warehouse study cost us?
Our source is the 2026 fee survey, a post on this site that compiles the ranges study providers publish, and refrigerated warehouses go unnamed in it. Read against its bands, a single refrigerated building financed through SBA and built to a conventional plan sits with the special-purpose work priced from $5,000 to $12,500; a phased build, a campus of several buildings or an unusual concept moves up to the complex band, $10,000 to $25,000; and USDA work, written to answer 7 CFR Part 5001, is quoted from $7,500 up to $18,750 or beyond. Within a band the fact pattern sets the price, by the survey's own logic: an operating warehouse with trailing statements needs less work than a ground-up build, and several temperature zones delivered in stages read as complex. The survey also records that SBA neither fixes nor limits what a study may cost. Our fee for a given building comes back with its scope.
Can the cost of the study go into the SBA loan?
The 2026 fee survey on this site says it can in both programs: under 504 the debenture finances eligible soft costs as part of what the whole project costs, while the proceeds of a 7(a) loan may pay for the third-party reports the file needs. On a cold storage project the fee normally sits in sources and uses beside the appraisal and the environmental report. The survey's basis-point table also shows the fee shrinking against the loan as the loan grows; above roughly $2 million, it says, a study costs less than the appraisal on a going-concern property. The 504 cold storage median on this page is $2.1 million, and the busiest 504 band, over $2 million to $5 million, counts 19 loans, with 13 more from 7(a) in the same range; much of the class's building lending therefore sits where the fee is small beside the loan.
How long does a cold storage feasibility study take?
No fixed schedule applies to every study; we state the timeline in the reply that carries the scope. Among the factors this site's 2026 fee survey says add time, several fit refrigerated buildings: a ground-up project needs its construction budget checked and its ramp counted from opening day; each extra site needs a market section of its own; a USDA overlay brings the enumerated factors and a signed opinion; and a sparse rural market means checking comparable properties in person. Cold storage adds its own time where customer agreements are still in negotiation or the utility has not confirmed the electrical service, because a study should report those facts rather than assume them. Sending the draft agreements and the utility correspondence with the scope request lets the timeline account for them from the start.
How large are SBA loans for cold storage facilities?
From FY2017 to FY2026 the 7(a) median is $615,000, with a lower quartile of $175,000 and an upper quartile of $2.3 million; the 504 median is $2.1 million, with quartiles of $1.2 million and $4.9 million. Under 7(a) the loans split between small credits, 12 of them in the smallest band for $1.1 million in all, and building-scale loans, 13 in the band just below the top for $46.5 million. Under 504, 19 loans sit in that band for $67.8 million, and 3 go above it. No state has enough loans to print a median of its own, so a cold storage study benchmarks the proposed loan against these national figures and explains its position by building size, temperature zones and what the loan pays for.
Is 7(a) or 504 the better fit for a refrigerated warehouse loan?
Both programs lend to the class, and they split it by size. 7(a) made more cold storage loans, 49 against 39 from FY2017 to FY2026, but 504 carried more dollars, $102.3 million against $66.1 million, and every cold storage loan in the top size band was a 504 loan. The 504 middle, $1.2 million to $4.9 million, is building scale; 7(a) reaches that scale in its upper band and also makes small credits that 504 does not, since 504 has no loan in the smallest band. Cancellation runs higher under 7(a), 20.9% against 15.2%, both floors. The choice is the borrower's and the lender's to make; the study is written for the program chosen and, where the request may move between programs, sized so that either reader can rely on it.
Which states have the most SBA cold storage loans?
California leads by count with 18 loans and $26.5 million from FY2017 to FY2026, and Texas leads by dollars with $28.6 million on 11 loans. Wisconsin follows with 10 loans and $23.7 million, then Massachusetts with 7 and $20.5 million; Florida and Ohio show 5 each, and Virginia 3. The ranking by count hides differences in scale: Ohio's 5 loans total $726,100 against $6.2 million for Florida's 5, and from Virginia down every row prints its count with no dollar total, because each rests on fewer than five loans. None of the state rows carries a median, a split between programs or a floor, because none has enough loans. For a site in one of these states, the study uses the state's row as context, reads its total against its count to see what size of project the state has financed, and carries its performance comparison at the national level.
My state has one cold storage loan in the table, or no row. What does the study compare my project with?
With the national program rows. Where the state table shows a single loan, as it does for New York, Pennsylvania and North Carolina, or no row at all, as for Georgia and Illinois, there is no state benchmark, and one loan cannot be made into one. The study sets the project against the medians, quartiles, size bands and cancellation floors of its program and, where a neighbouring state leads its Census region, cites that row as regional context. The local evidence is then built directly: the processors, distributors and retailers who would use the space, the public refrigerated warehouses within trucking distance and how full they are, the rates they charge, and the electrical service the site can get. A lender will notice the thin state record, so the study addresses it openly.
Is SBA lending to cold storage growing or slowing?
By count it is slowing. 7(a) approved 3 cold storage loans in each of FY2025 and FY2026, against 10 in FY2020 and 6 in each of FY2023 and FY2024. 504 approved 5 in FY2025 and 1 in FY2026, after none in FY2024. The class has never been large: the 7(a) high is 12, in FY2011, and the 504 high is 7, reached in FY2017 and again in FY2021. Most annual dollar figures in the long table are withheld or rest on too few loans to print, so the table cannot show whether loan sizes are rising. A borrower applying now joins a quiet stretch for the class, and a study that documents committed customer demand gives the lender a reason to approve that the recent record does not.
How often are SBA cold storage loans cancelled or charged off?
Cancellation is the measured risk. Measured on FY2017 to FY2024 approvals, at least 20.9% of 7(a) cold storage approvals and 15.2% of 504 approvals were cancelled, and both figures are floors. Charge-offs are not reported for either program, on the full measure or within 5 years, because the disbursed cohort from FY2017 to FY2021 approvals holds fewer than 30 loans in each program. That blank is not evidence of low losses. For a cold storage borrower the practical reading is that the study has two jobs: making the approval hold, with signed or documented customer demand, a bid-backed budget and a confirmed power supply, and giving the lender a coverage and collateral case that stands without any loss history behind it.
We process food and want to build our own freezer. Is this the record our lender will use?
Usually not as your own industry's record. The class on this page counts borrowers whose industry is refrigerated warehousing, and a processor that builds cold space for its own product is classed in its own industry, so its loan would not be among these 171 7(a) and 118 504 loans. A lender may still read this page for the building, since construction cost, power and fallback value raise the same questions, but the demand case differs: an owner-user's feasibility rests on its own production and sales, and an operator's on outside customers. The study is scoped to the business the borrower actually runs, and where the building will also rent space to others, it tests both sources of revenue.
How is a cold storage feasibility study different from the appraisal?
An appraiser gives an opinion of value. The study asks a different question: will paying customers fill the rooms, and will the margin they leave cover the debt? For cold storage that means naming the customers and the terms on which they will take space, by temperature zone; surveying competing refrigerated space with its occupancy and rates; checking the construction budget for the envelope, plant, docks, slab and standby power; building a utility budget and running coverage at higher power prices and lower occupancy; and estimating the building's value to an occupant that would never switch the refrigeration on. The appraisal can draw on that market evidence where comparable sales are scarce. The record on this page, with cancellation floors of 20.9% and 15.2% and no printed charge-off rate, is why a lender wants both documents in the file.
Will my lender have financed a cold storage facility before?
It may not have. Across the whole country from FY2017 to FY2026 the record holds 49 cold storage approvals through 7(a) and another 39 through 504, scattered over many states, and this page does not rank lenders for the class. A credit officer may therefore meet a refrigerated warehouse for the first time in your file and read it against dry warehouse lending that the bank knows better. The study is written for that reader. It explains, with local evidence, what separates cold storage from dry space: the refrigeration plant, insulated envelope, freezer floors and power draw on the cost side, and handling charges and customer contracts on the revenue side. It also shows what part of the cost would hold if the building had to be sold to a dry warehouse user.
We are buying an operating cold storage business. Is a quality of earnings report the same as a feasibility study?
They are different documents. As this site's 2026 fee survey reports it, SOP 50 10 8.1 (effective October 1, 2026) makes a quality of earnings report a requirement where a business is bought for $3 million or more, lender summaries price it at $10,000 to $30,000, and the survey keeps it distinct from the feasibility study. A cold storage purchase at the scale of the 504 middle on this page, $1.2 million to $4.9 million, can cross that threshold. The quality of earnings report tests the seller's historical figures. The feasibility study tests the market going forward: the customer agreements and their renewal dates, the condition and remaining life of the compressors, evaporators and panels the buyer inherits, the utility cost, and whether coverage holds after the purchase.
What other third-party reports will a lender order for a cold storage loan?
Expect a stack of reports around the study. This site's 2026 fee survey gives their published ranges: the appraisal, $2,500 to $7,500 for ordinary assignments, with going-concern work on special-purpose property priced between $5,000 and $15,000; the environmental screen, a questionnaire and records search costing $300 to $1,500; a Phase I when the screen calls for one, and the package of title, survey and recording, each quoted between $2,000 and $5,000; and, if the business changes hands, a valuation costing $1,800 to $7,000. The survey notes that SBA's rules call for an appraiser with industry experience and a going-concern allocation on special-purpose property, the treatment a lender may give a refrigerated building with few local buyers. The feasibility study supplies the market and cost evidence those reports draw on, which is a reason to order it early.
Can a rural cold storage project use a USDA B&I guarantee?
We write cold storage studies for B&I lenders to the standard set out in 7 CFR Part 5001. On the USDA side, the survey of 2026 study fees that this site publishes makes a study compulsory when a new business, a refrigerated warehouse start-up among them, borrows more than $1 million under B&I, and it prices that work from $7,500 at the low end to $18,750 or more at the high end. Whether a given project qualifies is for the lender and the Agency to decide. No USDA loan is counted in any table here, so for a rural refrigerated warehouse the study's evidence comes from the trade area itself: the growers, processors and shippers who would use the space, the distance to the nearest public cold storage, and the power the site can draw.
What should I send to get a cold storage study scoped?
The address or parcel; whether the project is a new building, an expansion or a purchase; and the building as planned: floor area, pallet positions, temperature zones, dock doors and any blast freezing or handling services. Add the customer position, from signed agreements to letters of intent, the utility's service confirmation if there is one, the construction budget or the seller's statements, which program or lender you are working with, if any, and the loan amount. We read it against this page first: the state row, the program median and quartiles, the size band the loan falls in, and the cancellation floor that applies, and then reply with scope and timeline.
Loan Analytics prepares the feasibility study a lender reads before it commits: for a cold storage project anywhere in the United States, financed under SBA 7(a), SBA 504, USDA Business and Industry or a conventional bank loan. Tell us the project and the lender's deadline, and we reply with scope, timeline and fee.
Daniel Smith, MAI
Request Scope & Timeline
Start with the data.
Contact Us