Is buying an Airbnb in December to write off W-2 income actually legit?

The rules behind it are real. 100% bonus depreciation under IRC §168(k) is not prorated for a late start, and a rental with average stays of seven days or less escapes the per se passive rule under Treas. Reg. §1.469-1T(e)(3)(ii)(A). What makes it fragile is the calendar: material participation is tested over the taxable year, so a December start has to show more than 100 hours in a few weeks, and a loss resting mostly on pre-listing setup hours is the version I would expect an examiner to push on.

Does bonus depreciation get cut if the rental goes into service in December?

No. Bonus depreciation under IRC §168(k) is a first-year deduction with no monthly proration, so cost-segregated 5-, 7-, and 15-year property placed in service December 4 deducts the same 100% as property placed in service in March. Only the building shell is prorated: 39-year nonresidential real property placed in service in December gets 0.107% of its basis in year one under the mid-month convention in Publication 946.

Do my spouse's hours count toward the 100 hours?

Yes. IRC §469(h)(5) takes a spouse's participation into account, and Temp. Reg. §1.469-5T(f)(3) applies that whether or not the spouse owns any part of the property and whether or not you file jointly. Two spouses can combine hours to clear the 100-hour test of §1.469-5T(a)(3), though each should keep a log of what they actually did.

What if I close in December but don't list the cabin until January?

Then it is placed in service in 2027, and the depreciation, the seven-day average, and the hours all belong to 2027. Personal use before that falls in a year with no rental, so a family Christmas in 2026 does not touch the 2027 allocation under IRC §280A(e). Publication 527 starts rental expense deductions when the property is made available for rent, so the cost of carrying it empty in December is not part of a Schedule E loss.

Is there a cap on how much of a short-term rental loss can offset my salary?

Yes, above a high line. IRC §461(l) limits net business losses to $512,000 on a joint return and $256,000 otherwise for 2026, and wages do not count as business income in that test. The excess is not lost; it carries forward as a net operating loss. A single cabin rarely reaches it, but a cost-segregated luxury property can.

The Journal / Real Estate Tax

Buy a Short-Term Rental Before Year-End? Four Weeks Has to Do a Year's Work.

By Ewan Morkel, EAOctober 25, 20266 min read

A cabin closed in November, cost-segregated, and listed by December can put a six-figure loss against 2026 wages. It only works if the property, the guests, and your hours all fit inside the weeks that are left, and one family Christmas there can cost $17,289.

A-frame vacation rental illuminated at dusk in the mountains

A two-income couple with $420,000 of wages closes on a mountain cabin in mid-November. Somebody (a podcast, the agent, a cost segregation firm with a quarter to finish) has told them that if they buy a short-term rental before year-end, they can write most of it off against their salaries. That is not a myth. It is also not automatic. The deduction depends on four things being true by December 31, and the calendar works against every one of them.

The pitch

What the December pitch gets right.

The math in the pitch is mostly honest. Under §168(k), as amended by OBBBA, property acquired after January 19, 2025 gets 100% bonus depreciation, and bonus is a first-year deduction with no monthly proration. A cost segregation study that carves a quarter of the building into 5- and 15-year property deducts all of that quarter in year one, whether the cabin went into service in March or December. Only the building shell feels the late start. A 39-year building placed in service in December gets 0.107% of its basis under the mid-month convention, which on this cabin is $423.

What the pitch skips is everything that decides whether the loss is passive. Using a rental loss against wages takes the short-term rental loophole, and every test in that rule is measured over the taxable year. In a year that starts in December, each one has about four weeks to clear.

The four gates

To buy a short-term rental before year-end and deduct it, four things must be true.

01It is placed in service by December 31. Closing is not enough. Publication 527 treats a rental as placed in service when it is ready and available for rent, which means furnished, photographed, and listed with bookable dates. A cabin still waiting on beds on New Year's Eve depreciates in 2027.
02Real guests stay, and the average stay is seven days or less. Treas. Reg. §1.469-1T(e)(3)(ii)(A) works from the customer use that actually happened during the year. A listing that goes live December 29 with no bookings has nothing to average, and I would not file a nonpassive loss on a year with zero guests.
03You materially participate in the weeks that remain. Temp. Reg. §1.469-5T(a)(3) asks for more than 100 hours and no less than any other individual. Spouses count together under §469(h)(5), so two owners at 13 hours a week each clear it in four weeks. The 500-hour test is out of reach, and the substantially-all test in §1.469-5T(a)(2) only fits owners who do their own cleaning, with no percentage in the regulation to say how much is enough.
04Personal use stays out, or you count what it costs. This is the gate nobody mentions, and it gets its own section below.

December makes one test easier: the comparison against everyone else. Four turnovers is maybe 14 hours of cleaning, so the cleaner is rarely the problem in a short year. It makes another harder: the setup hours. How to prove short-term rental hours treats furnishing and building the listing as countable operational work, and I still think that is the better reading. But no Tax Court case squarely decides whether hours spent before the property was ready for guests count toward that year's test, and on a December return the setup is most of the log. I want the 100 hours to stand on what happened after the listing went live.

A $650,000 cabin, closed November 13 and listed December 4, 2026.

Building basis after a $130,000 land allocation: $520,000
Cost seg components (24% of the building), 100% bonus: $124,800
Furniture and fixtures bought for the listing, 100% bonus: $38,000
Remaining $395,200 of building, 39-year, December rate 0.107%: $423
December operating costs, including the study fee: $10,400
December rental income, 4 stays, 13 nights: $6,200
2026 loss with all four gates cleared: $167,423
Tax saved at 24% federal plus 4.45% Utah: $47,632
Family stays December 22 to 29: §280A(e) ratio, 13 of 20 days: 65%
2026 loss after the allocation: $106,655
Tax saved with the Christmas week: $30,343
Only 70 logged hours: loss suspended on Form 8582, 2026 tax saved: $0

Tax year 2026, married filing jointly, $420,000 of combined wages and the $32,200 standard deduction, so taxable income falls from $387,800 to $220,377 and every deducted dollar sits in the 24% bracket ($211,401 to $403,550 under Rev. Proc. 2025-32). Utah at 4.45% under S.B. 60, retroactive to January 1, 2026, ignoring the taxpayer credit phase-out. The building is nonresidential real property because stays average seven days or less. The Christmas case applies the Publication 527 ratio of days rented to total days used to every expense, interest and taxes included. No substantial services are provided, so no self-employment tax. The loss is well under the 2026 §461(l) excess business loss threshold of $512,000 for a joint return.

The Christmas trap

One family week at the cabin can take a third of the deduction.

Buying in November usually means the family wants Christmas there. §280A(e) says that in any year you use a dwelling unit for personal purposes on even one day, rental deductions are limited to the share of days rented at a fair rental over total days used. Unlike the 14-day test, it applies even when the personal use is too short to make the cabin a residence. In a normal year, seven family nights against 200 rented ones barely moves the ratio. In a December year with 13 rented nights, the same week makes it 65%, and the ratio reaches the bonus depreciation, not just the utilities. On the numbers above that is $60,768 of lost deduction and $17,289 of tax, and §280A(e) has no carryforward for the share it takes.

Days spent working on the place substantially full time are not personal use under the flush language of §280A(d)(2). A holiday with some shelf-hanging is still a holiday.

Where I land

The December 31 deadline is for the cabin, not for you.

If the keys arrive by early November and you will run the cabin yourselves, the December play is real, and the table shows what it is worth. List it by the first week of December, take real bookings, keep the family out until January, and log every hour from the day the listing goes live.

If the keys arrive after Thanksgiving, I would usually tell you to put the cabin in service in January instead. That sounds like giving up $47,632. It is closer to moving it a year, because the 100% bonus rate is permanent and no worse in 2027, and you get twelve months to meet the hours instead of four weeks. The expensive outcome is the middle one: list in late December, fall short of 100 hours, and the whole $167,423 becomes a suspended passive loss. Qualifying in 2027 does not rescue it. Under §469(f)(1), a suspended loss from an activity that later turns nonpassive offsets only that activity's own income, not your wages, so a failed first year can trap the biggest deduction the property will ever produce until the cabin earns it back or you sell.

None of this answers whether the cabin is a good investment. A loss that saves $47,632 on a property that loses money for five years is still a property that loses money.