What a $16,364 Rental Write-Off Is Actually Worth

A rental owner posts that the property “lost” $9,000 last year and the replies treat it as a win. A cost segregation firm advertises a $99,000 first-year deduction. Someone on a podcast explains that depreciation will wipe out a W-2. In all three cases a tax number is doing the talking, and nobody has said what happened to anyone’s net worth.

Those are different questions, and the tax code answers only one of them. This guide separates them, prices a real deduction on one modeled property, and gets to the rule that decides whether a rental write-off is worth its face value, a fraction of it, or nothing at all this year.

Three ledgers, and why they disagree

Most arguments about real estate and tax strategy are really arguments between three measurement systems built for different purposes.

The third row is the problem. Tax accounting gets produced for you once a year by law. Cash accounting gets produced for you continuously by your bank. Nobody produces the economic ledger, so it tends to get replaced by whichever of the other two sounds better.

The three disagree constantly, and disagreement is normal rather than a sign that something is wrong. A rental can show positive cash flow, a tax loss, and a positive economic return in the same year. A stock portfolio can suffer a real economic loss years before any tax loss exists. An investor can take a large deduction in a year their net worth fell.

A deduction reduces a cost. It does not refund one.

Spend $10,000 on something fully deductible. Setting other interactions aside, the deduction is worth your marginal rate:

This is arithmetic a ten-year-old can do, and it is the source of an enormous amount of bad investment reasoning. “I can write it off” answers what an expense costs after tax. It never answers whether the expense was worth making.

Both answers can be good at once. A $10,000 roof that prevents $30,000 of water damage is worth doing at any marginal rate, and the deduction makes a sound decision cheaper. Spending $10,000 you did not need to spend because it generates a $2,400 deduction reduces your wealth by $7,600 with a tax-shaped explanation attached.

The Summitward version: never spend a dollar to save a quarter.

The clean case is a loss that already happened

Tax-loss harvesting is the easiest place to see the distinction, because there the tax loss and the economic loss are genuinely connected.

Put $100,000 in a fund. It falls to $80,000. You are $20,000 poorer right now, whether or not you sell. Realization sets the tax treatment; it does not cause the loss and it does not cure it. Selling and buying a similar but not substantially identical fund converts a loss you already suffered into a tax asset while keeping you invested.

The economic damage precedes the harvest by definition, which is why harvesting is worth doing when it is cheap. Ivković, Poterba, and Weisbenner, studying six years of individual brokerage accounts, found that of stock purchases of $10,000 or more, 21% produced a realized loss within a year while another 18% of all purchases sat on unrealized losses at the one-year mark. Their conclusion was that “nearly one-half of the stock purchases that could have been used to generate a short-term loss were not liquidated in time to generate this loss.”¹ The losses happened. Most were simply never claimed.

What harvesting is worth is a separate and much-abused question. The most-cited estimate is 1.08% a year, before transaction costs, falling to 0.82% once the wash-sale rule is enforced, in a simulation over the 500 largest CRSP stocks from 1926 to 2018 at assumed rates of 15% and 35%.² Harvesting also cuts your cost basis by the amount harvested, so much of the benefit is timing rather than forgiveness. We took that apart in what “tax alpha” actually means, priced a specific investor’s harvest in what $121,281 of harvested losses was worth, and made the realization point at length in yes, an unrealized loss is still a loss. The rest of this guide is about the harder case.

The same property, on three ledgers

Here is one modeled property. Every figure comes from running the rules, and the assumptions are deliberately generous to the rental: a 12.5 price-to-rent ratio, 40% down, and 3% annual appreciation.

• $600,000 residential rental, 25% of the price allocated to land
• $240,000 down, $360,000 borrowed at 6.5% over 30 years
• $46,968 of rent collected after a 5% vacancy allowance
• $16,365 of cash operating expenses: property tax, insurance, maintenance, and management
• Owned by a couple filing jointly with $200,000 of other income, a 22% marginal federal rate under the 2026 brackets³

Land is not depreciable, so the depreciable building basis is $450,000. Residential rental property is written off straight-line over 27.5 years under the general depreciation system, which is $16,364 a year.⁴

Treat that land split as an assumption that moves the answer. It is a facts-and-circumstances allocation, normally taken from the county assessor’s land-to-improvement ratio. FHFA’s value-weighted national land share was 39.9% in 2022, while its equal-weighted county medians run between 18% and 24% and the county spread reaches from under 8% to over 54%.⁵ At a 10% land share this property would depreciate $19,636 a year; at 39.9% it would depreciate $13,113. Use your own assessor’s numbers.

Here is the second full year on all three ledgers.

Three numbers, all correct, none contradicting the others. Cash flow is positive by $3,298. Schedule E shows an $8,887 loss. Net worth rose about $26,000, and nearly all of that came from a price assumption rather than from anything the tax code did.

Two lines explain the whole divergence. Mortgage principal is cash leaving your account that you cannot deduct, and it is not an economic cost either, because it moves wealth from your bank balance into your equity. Depreciation is a deduction that costs no cash at all. The cash ledger and the tax ledger are each missing exactly one of those, in opposite directions.

That is why “I lost $8,887 on the property” and “the property made me $26,000” are both defensible sentences about the same year, and why neither tells you whether the investment is good. Flip the price assumption to a 5% decline and the economic line becomes a $22,000 loss while the cash and tax lines do not move at all.

Tax depreciation and economic depreciation are different schedules

Real estate marketing likes to call depreciation a “phantom expense.” The real structure of the advantage is more specific than that, and it makes a stronger case for owning residential rental property.

Buildings really do wear out. The Bureau of Economic Analysis defines economic depreciation as “the decline in the value of the stock of fixed assets due to physical deterioration, normal obsolescence, and accidental damage”⁶ and measures it separately from anything on a tax return. BEA is explicit that the two are not the same thing: IRS data “are based on historical-cost valuation and on tax service lives. BEA must adjust these estimates to the NIPA definition of depreciation, consumption of fixed capital, which is based on current-cost valuation and economic service lives.”⁶ A roof is a real cost. The statutory schedule simply is not a measurement of it.

How far apart are they? The tax schedule writes off 1/27.5 of the building each year, or 3.64%. BEA’s geometric rate for new one- to four-unit residential structures is 1.14% a year.⁷ Treat that as a range rather than a point: BEA’s figure is a fallback declining-balance rate divided by an 80-year service life traceable to a 1963 study, and independent estimates from housing transaction data cluster between roughly 1% and 2%. What every one of them agrees on is that the number is well under 3.64%.

The cleanest version of the comparison comes from Treasury’s Office of Tax Analysis, which computes the present value of tax depreciation against the present value of economic depreciation for each asset class. A reading of 100 means the tax schedule matches economic reality. The figures already carry the inflation penalty that historical-cost deductions suffer.

Residential rental at 27.5 years is the outlier. The 39-year categories sit within a few points of neutral, which means Congress set one of these recovery periods close to the economic evidence and the other well inside it. That asymmetry, rather than depreciation in general, is the real tax advantage in residential rental property, and it is a legitimate reason to own one.

The counterargument deserves saying out loud. Accelerated cost recovery was not designed as a giveaway. As the Congressional Research Service put it, the 1986 rules “allowed for accelerated deductions to offset the lack of indexing for inflation, so that the discounted present value of tax depreciation deductions was roughly equal to the discounted present value of economic depreciation. Since that time, inflation has declined and caused the value of tax depreciation to be more beneficial than economic depreciation.”⁸ The schedule did not get more generous. Inflation got lower.

Congress understood the gap when it wrote these rules. Explaining why the passive activity loss rules were created in 1986, the Joint Committee on Taxation wrote that “the statutory allowance for depreciation ... reflects broad industry averages, as opposed to providing precise item-by-item measurements. Accordingly, taxpayers with assets that depreciate less rapidly than the average, or that appreciate over time (as may be the case with certain real estate), could engage in tax sheltering.”⁹ The limits in the next section exist because the schedule in this one is deliberately approximate.

Whether you can use the loss at all

Generating an $8,887 rental loss does not mean deducting $8,887 against your salary.

Rental real estate is passive by default under section 469, and passive losses cannot offset nonpassive income such as wages. There is a special allowance of up to $25,000 for taxpayers who actively participate, but IRS Publication 925 states the limit plainly: the allowance “is reduced by 50% of the amount of your modified adjusted gross income that is more than $100,000,” and at $150,000 or more “you generally can’t use the special allowance.”¹⁰

Neither threshold is indexed. The $25,000 figure and the $100,000 phaseout start have been the same numbers in section 469(i) since the Tax Reform Act of 1986, with no cost-of-living provision, and they do not appear among the items the IRS adjusts each year.¹¹ A household comfortably inside the allowance in 1986 is well past it today on the same real income.

For the modeled couple at $200,000 of modified AGI, the allowance is zero. The $8,887 loss is suspended and carried forward. Over ten years the property throws off $156,136 of depreciation and $40,499 of cumulative Schedule E losses, and the federal tax saved is $0. The losses are not destroyed. They wait, and are generally released when you dispose of your entire interest in a fully taxable sale to an unrelated party.¹⁰ A benefit ten years out is worth less than the same benefit today.

The phaseout only bites once the remaining allowance drops below the loss itself, which is why a modest loss survives well into the phaseout range while a large one gets cut early. Two owners with identical properties and different incomes get completely different answers, and a deduction quoted without an income attached is not information.

Price your own property

Put your own numbers in. The middle column separates the Schedule E loss from the part of it you can actually deduct this year.

The two routes around the passive rules

Both routes are real, both are in the regulations, and both get sold far more casually than they should be.

Real estate professional status. Publication 925 sets two tests that must both be met: more than half of the personal services you performed in all trades or businesses during the year were in real property trades or businesses in which you materially participated, and you performed more than 750 hours of such services.¹⁰ On a joint return you cannot add your spouse’s hours to clear those two tests. One spouse has to meet both alone, and spousal participation counts only for the separate question of material participation. A full-time W-2 job makes the more-than-half test close to unreachable.

The short-term rental route. Under Temporary Regulation § 1.469-1T(e)(3)(ii)(A), an activity is not a rental activity at all if “the average period of customer use for such property is seven days or less.”¹² That is a real regulation, correctly described.

What usually gets left out is what happens next. Falling outside the definition of a rental activity does not make the loss nonpassive. It removes the automatic presumption. The activity is then tested as an ordinary trade or business, which still requires material participation under one of the seven tests in the regulations. Fail that, and the loss is still passive, and it is now also ineligible for the $25,000 allowance, which section 469(i) grants only to rental real estate activities.¹¹ Running the modeled property as a seven-day-average short-term rental without materially participating is strictly worse than leaving it an ordinary rental.

Rows three and four are the same building with the same seven-day average stay. The $8,910 gap between them is material participation, paid for in hours of genuine work and contemporaneous records you may have to produce years later. That is a real price, and it is the part that never makes it into the pitch.

One more limit sits behind all of this. Clearing section 469 is not the last gate. Publication 925 notes that losses allowed after the basis, at-risk, and passive limits may still be capped by the excess business loss limitation, which the 2025 law made permanent and which is $256,000, or $512,000 on a joint return, for 2026.¹³

Cost segregation and 100% bonus depreciation

A cost segregation study identifies components of a building that qualify as shorter-lived property rather than part of the 27.5-year or 39-year structure. The IRS describes the mechanism in its own Cost Segregation Audit Technique Guide: tangible personal property “has a shorter recovery period (e.g., 5 or 7 years) and can be also eligible for accelerated depreciation,” so “a faster depreciation write-off (and tax benefit) can be obtained by allocating costs to § 1245 property.”¹⁴

The 2025 tax law reinstated a 100% first-year depreciation allowance and made it permanent, for qualified property acquired after January 19, 2025.¹⁵ One detail matters for anyone who bought around that date: the effective date turns on acquisition, and property under a written binding contract entered into before January 20, 2025 is not treated as acquired after that date, so it stays under the old phase-down.¹⁶ Verified as of September 2026.

Bonus depreciation never applies to the residential building itself. Section 168(k) is limited to property with a recovery period of 20 years or less, and residential rental property is 27.5-year property. It reaches only the shorter-lived components a study carves out, which is exactly why the two get marketed together.

Assume a study reclassifies 22% of the $450,000 building basis, or $99,000, into five-, seven-, and fifteen-year property, all bonused in year one. That share is an illustration; the IRS warns specifically against “rule of thumb” allocations based on a preparer’s experience with a property type, naming residential rental property as an example, and says an examiner “should view this approach with caution.”¹⁴ Here is the honest price list.

Every row is a different number and only the first is a deduction. The gap between $99,000 and $0 is the same phaseout from two sections ago, applied to a bigger loss.

Two more things before paying for a study. The IRS guide states that it “has not established any requirements or standards for the preparation of cost segregation studies,” that there are “no prescribed qualifications for cost segregation preparers,” and that contingency fee arrangements “create the incentive to maximize the amount of costs attributed to § 1245 property.”¹⁴ Published fee data is thin and old: the last figures from a source that does not sell studies are a $10,000 to $25,000 range from the Journal of Accountancy in 2004 and a $5,000 to $30,000 range from a Texas A&M real estate center piece in 2010.¹⁷ Nearly everything published on cost segregation pricing since then comes from firms that sell the studies, which is worth knowing when you compare a $500 product and a $15,000 product that carry the same name. And section 179, which gets mentioned in the same breath, does not reach residential rental property at all: the qualified real property it covers is improvements to nonresidential real property.¹⁸

Depreciation lowers basis, and basis comes back at sale

Every dollar of depreciation reduces the property’s adjusted basis, which increases the gain when you sell. The rule is unforgiving about owners who skip it. Publication 544 says that if you took no depreciation deduction at all, “your adjustments to basis for depreciation allowable are figured by using the straight-line method.”¹⁹ Not claiming it does not preserve your basis.

On sale, gain attributable to depreciation on section 1250 real property is unrecaptured section 1250 gain, taxed at a maximum federal rate of 25%.¹⁹ That 25% is a ceiling rather than a flat rate, so a taxpayer below it pays their ordinary rate. The shorter-lived components a cost segregation study creates are section 1245 property, and section 1245 gain is recaptured as ordinary income with no such ceiling. Cost segregation therefore raises the rate at sale only for taxpayers above 25%. On $99,000 of reclassified basis the extra tax at sale is $0 at a 24% rate, $6,930 at 32%, and $11,880 at 37%.

Selling the modeled property in year ten at 3% annual appreciation and 7% selling costs produces $156,136 of accumulated depreciation, a $306,042 gain, $56,836 of federal tax, and the release of $37,833 of suspended losses worth $8,323. Depreciation moved income through time. It did not erase it.

There are real exits that change the ending. A 1031 exchange defers gain into a replacement property, and a step-up in basis at death can eliminate it. Those are planning questions with their own conditions, and they are why depreciation usually comes back rather than always.

Running the claims backwards

What we recommend

• Underwrite the property first, then add the tax treatment. If a rental works on rent, expenses, and a defensible price assumption, depreciation makes a sound investment better, and Treasury’s own figures say the residential schedule is genuinely favorable. If it does not work, no deduction rescues it. We reached the same conclusion from the other direction in rental property vs. global index funds.
• Find your modified AGI before you find a property. Above $150,000 without material participation, a rental loss produces a carryforward and no current benefit. That is fine if you wanted the property. It is a bad surprise if the deduction was the reason.
• Price a cost segregation study as deduction times rate, discounted, minus the fee. Pay for one when the basis is large, the loss is usable this year, and the fee is small against the timing benefit. Skip it when the loss would be suspended anyway, and read the allocation method before you sign.
• Treat the short-term rental route as a job. It works only with material participation, which means hours and records. Without them it is worse than an ordinary rental.
• Keep an economic ledger. Nobody will produce it for you. Cash flow, plus principal paid down, plus price change, against the equity you have tied up and what that equity could have earned elsewhere.

How Summitward helps

The question underneath all of this is your marginal rate and your modified AGI, because those two numbers decide what any deduction is worth to you. Tax Projection computes projected AGI and marginal federal rate from your own income, and Cash Flow keeps the cash ledger separate from the tax one.

Frequently asked questions

Is rental depreciation a fake expense?

No. Buildings deteriorate, roofs and heating systems fail, kitchens become obsolete, and replacing them costs real cash. What is true is that the statutory schedule and the rate at which a building loses value are set by different processes. For residential rental the statutory schedule is the faster of the two, which is a genuine advantage and a different claim from calling the expense imaginary.

If my rental loss is suspended, have I lost it?

No. Suspended passive losses carry forward and are generally allowed in full when you dispose of your entire interest in the activity, in a transaction where all gain or loss is recognized, to someone not related to you.¹⁰ All three conditions have to hold. The cost is timing: a deduction years from now is worth less than the same deduction today.

Does the $25,000 allowance ever get adjusted for inflation?

It has not been. The $25,000 amount and the $100,000 phaseout threshold are stated as flat dollar figures in section 469(i) with no cost-of-living provision, and they do not appear among the items the IRS adjusts annually.¹¹ They are the numbers Congress wrote in 1986.

Is a cost segregation study ever worth it?

Yes, when three things line up: the building basis is large enough that reclassification moves real money, the resulting loss is usable this year rather than suspended, and the fee is small relative to the present value of moving deductions forward. The study also has to be defensible, and the IRS is explicit that it has set no standards for how one is prepared.¹⁴

Why does my rental show positive cash flow and a tax loss?

Because mortgage principal is cash you pay that you cannot deduct, and depreciation is a deduction that costs no cash. In most leveraged rentals the depreciation figure exceeds the principal figure in the early years, so the tax result is more negative than the cash result. Both numbers are correct.

How do I split the purchase price between land and building?

It is a facts-and-circumstances allocation, and most taxpayers use the county assessor’s land-to-improvement ratio or an appraisal. The spread across US counties runs from under 8% land to over 54%, so a national average is not a substitute for your own property’s numbers.⁵ A higher land share means a smaller depreciable basis and a smaller deduction.

Should I buy a rental for the tax benefits?

No. Buy one because the property clears your hurdle on rent, expenses, risk, effort, and the return you could get elsewhere. Tax treatment then improves the after-tax outcome of a decision that was already sound.

Key takeaways

• Tax income, cash flow, and economic wealth are three different measures. They disagree routinely, and only the first two get produced for you automatically.
• A deduction is worth your marginal rate. At 24%, a $10,000 deductible expense still leaves you $7,600 poorer.
• Residential rental depreciation is genuinely favorable, and the tax code says so. Treasury puts the present value of tax depreciation on tenant-occupied residential buildings at about 175% of economic depreciation, against roughly 100% for offices, warehouses, and lodging.
• A rental loss is passive by default, and the $25,000 allowance is gone at $150,000 of modified AGI. Neither threshold has been indexed since 1986.
• For one modeled $600,000 rental owned by a couple earning $200,000, ten years of depreciation totaled $156,136 and saved $0 of federal tax. Every dollar was suspended to a carryforward. A different income makes it a different answer.
• The short-term rental route works only with material participation. Without it the activity is still passive and has also forfeited the $25,000 allowance.
• Depreciation lowers basis and mostly comes back at sale, at a 25% ceiling for section 1250 gain and at ordinary rates for the section 1245 components a cost segregation study creates.

Related guides

• Rental Property vs. Global Index Funds underwrites the property itself, before any of the tax treatment here applies.
• What “Tax Alpha” Actually Means prices a harvested loss under every exit and explains where the headline percentages come from.
• What $121,281 of Harvested Losses Was Worth runs the same arithmetic on a specific public claim.
• Yes, an Unrealized Loss Is Still a Loss on why realization sets the tax treatment and not the economics.
• There Are Only Four Things You Can Do With a Capital Gain on what deferral is worth in present value.
• The True Cost of Owning a Home applies the same three-ledger split to a house you live in.

Sources

1. Zoran Ivković, James Poterba, and Scott Weisbenner, “Tax-Motivated Trading by Individual Investors,” American Economic Review 95, no. 5 (2005): 1605–1630. NBER Working Paper 10275. nber.org
2. Shomesh E. Chaudhuri, Terence C. Burnham, and Andrew W. Lo, “An Empirical Evaluation of Tax-Loss-Harvesting Alpha,” Financial Analysts Journal 76, no. 3 (2020): 99–108. The 1.08% figure is before transaction costs; 0.82% enforces the wash-sale rule. MIT DSpace
3. IRS Revenue Procedure 2025-32, inflation adjustments for tax year 2026, including the federal rate brackets and the standard deduction. irs.gov
4. IRS Publication 527, Residential Rental Property: 27.5-year GDS straight-line recovery, the mid-month convention, and the rule that land is not depreciable. irs.gov
5. Federal Housing Finance Agency, “Land Price Appreciation During the COVID-19 Pandemic,” August 2024 (national aggregate land share of 39.9% in 2022), and FHFA Working Paper 19-01, Davis, Larson, Oliner, and Shui, “The Price of Residential Land for Counties, ZIP Codes, and Census Tracts in the United States” (county distribution). fhfa.gov
6. Bureau of Economic Analysis, NIPA Handbook: Concepts and Methods of the U.S. National Income and Product Accounts, glossary (definition of consumption of fixed capital) and chapter 4 (adjustment of IRS tax-return depreciation to an economic measure). bea.gov
7. Bureau of Economic Analysis, “BEA Depreciation Estimates” (rates by asset type; 1-to-4-unit residential structures at 0.0114 a year), following Barbara M. Fraumeni, “The Measurement of Depreciation in the U.S. National Income and Product Accounts,” Survey of Current Business, July 1997. The residential rate is a Hulten-Wykoff default declining-balance rate divided by a service life drawn from a 1963 study, so treat it as an approximation. bea.gov
8. Tracy Foertsch, U.S. Cost of Capital Model Methodology, U.S. Department of the Treasury, Office of Tax Analysis Technical Paper 10, May 2022, Table 3. Figures are present values per dollar of investment, computed on a corporate basis. treasury.gov
9. Jane G. Gravelle, Bonus Depreciation: Economic and Budgetary Issues, Congressional Research Service R43432, updated October 2014. congress.gov
10. Staff of the Joint Committee on Taxation, General Explanation of the Tax Reform Act of 1986 (JCS-10-87), May 1987, the “Reasons for Change” discussion accompanying new section 469. jct.gov
11. IRS Publication 925, Passive Activity and At-Risk Rules: the $25,000 special allowance and its phaseout, the real estate professional tests, the seven-day rental exception, the excess business loss ordering, and the release of suspended losses on a qualifying disposition. irs.gov
12. 26 U.S.C. § 469(i), which states the $25,000 allowance and the $100,000 phaseout threshold as flat dollar amounts with no cost-of-living provision, and limits the allowance to rental real estate activities. law.cornell.edu
13. Temporary Treasury Regulation § 1.469-1T(e)(3)(ii)(A): an activity is not a rental activity if the average period of customer use is seven days or less. ecfr.gov
14. 26 U.S.C. § 461(l), excess business loss limitation, made permanent by P.L. 119-21 § 70601; the 2026 thresholds of $256,000 and $512,000 appear in Rev. Proc. 2025-32 § 4.31. law.cornell.edu
15. IRS Publication 5653 (February 2025), Cost Segregation Audit Technique Guide. This edition predates P.L. 119-21, so its bonus depreciation and section 179 figures are superseded; the section 1245 and 1250 classification framework is not. irs.gov
16. IRS Instructions for Form 4562 (2025), on the reinstated 100% special depreciation allowance for qualified property acquired and placed in service after January 19, 2025. irs.gov
17. IRS Notice 2026-11, describing the permanent 100% additional first year depreciation deduction and the written binding contract rule in P.L. 119-21 § 70301(c)(4). irs.gov
18. Jay A. Soled and Charles E. Falk, “Cost Segregation Applied,” Journal of Accountancy, August 2004 ($10,000 to $25,000), and Jerrold Stern, “Cost Segregation Yields Cash Flow,” Texas Real Estate Research Center, October 2010 ($5,000 to $30,000). Both predate current law and neither breaks fees out by property value. journalofaccountancy.com
19. IRS Publication 946, How to Depreciate Property: MACRS recovery periods, the 20-year ceiling on bonus-eligible property, and the limitation of qualified section 179 real property to improvements to nonresidential real property. irs.gov
20. IRS Publication 544, Sales and Other Dispositions of Assets: depreciation allowed or allowable, unrecaptured section 1250 gain at a maximum 25% rate, and section 1245 ordinary income recapture. irs.gov

Editor’s note

Educational content, not tax advice. Federal rules only; state treatment differs, and several states do not conform to federal bonus depreciation. Tax rules described here were verified against the IRS, Treasury, and BEA sources listed above in September 2026 and change over time. The worked example is one modeled property under stated assumptions; different inputs produce materially different answers. Consult a CPA before acting on any of it.