When Cost Segregation Isn't Worth It: A Decision Framework for Rental Investors

Engineering firms market cost-seg aggressively. Four conditions where it actually loses money — small basis, short holds, §469 PAL trap, §163(j) ADS election. The honest decision framework.

The cost segregation industry has a marketing problem that quietly transfers into a math problem on the owner's tax return. Engineering firms publish glossy one-pagers full of six-figure first-year deductions and "20-30% reclassification" claims, then charge $5,000 to $8,000 for a study built around those numbers. For the right property, in the right hands, the result is a meaningful pull-forward of deductions. For the wrong property — and a surprisingly large share of properties marketed to are the wrong ones — the study costs more than it produces, the deduction is suspended under §469, or the math reverses on disposition.

This article is the contrarian take. It walks through the four conditions under which cost segregation is uneconomic, the citations that put the realistic residential reclassification range much lower than the brochures claim, and the questions a CPA asks before signing off on a study. Nothing here is tax advice; it is a framework an investor uses to ask better questions before paying an engineer.

The four conditions that make cost-seg uneconomic

A cost segregation study has a fixed cost (the engineering fee) and a variable benefit (the time value of money on the deductions pulled forward). The benefit must exceed the fee, after tax, after recapture, and after passive-loss limitations. Four conditions, in combination or alone, can flip the sign of that calculation.

Condition 1 — Property tax basis under approximately $200,000

The engineering fee for a residential cost segregation study generally runs $4,000 to $8,000 depending on scope, plus the time cost of dealing with the engineer. For a single-family rental with a depreciable basis under roughly $200,000, the math gets thin fast. The IRS Cost Segregation Audit Techniques Guide, Chapter 7.2 (Industry-Specific Guidance — Residential Rental), characterizes 5-year personal property in single-family and small multifamily rentals as roughly 5-15% of basis, not the 20-30% that engineering-firm landing pages quote. On a $180,000 depreciable basis at the midpoint of that range — call it 10% to 5-year, 7% to 15-year — the reclassified bucket is around $30,000. With 100% bonus depreciation under IRC §168(k) as restored by OBBBA §70301, that produces a year-one deduction near $30,000. At a 32% marginal bracket the federal cash effect is roughly $9,600 — before subtracting the engineer's fee, before any state add-back, and before the §1245 recapture liability that follows the property to disposition.

For a small rental, the feasibility calculator at /cost-seg/?purchase\_price=200000 is a screening tool — it estimates the year-one swing before an investor commits to the engineering fee. If the screen shows a marginal benefit, that is itself the answer.

Condition 2 — Hold period under three years

Cost segregation is a timing tool, not a tax cut. The reclassified property is §1245 property under IRC §1245(a)(1) and §1245(a)(3), and on disposition the lesser of accumulated depreciation or realized gain comes back as ordinary income. The longer the hold, the more time value of money the owner captures on the pulled-forward deduction; the shorter the hold, the smaller the spread between "accelerated now" and "would have happened anyway." The §1245 recapture trap article walks through the disposition mechanics in full.

A useful rule of thumb: under three years, the §1245 recapture differential plus the engineering fee often consumes most of the year-one benefit. The /cost-seg/?planned\_disposition=lt-2yr screen makes that visible. A 1031 exchange can defer the recapture (see Treas. Reg. §1.1031(a)-1 and §1.1245-4), but only if the replacement property is itself eligible and the exchange is properly structured — that is a planning conversation with a CPA, not a marketing bullet.

Condition 3 — Passive losses with no REPS qualification or STR carve-out

This is the single most common reason a study produces a deduction the owner cannot use in year one. IRC §469(c)(2) classifies all rental activity — by definition, regardless of participation — as passive, and §469(a) suspends passive losses against active or portfolio income. The deduction does not vanish; it sits on the return as a suspended passive activity loss until either the owner has passive income to absorb it, the property is fully disposed of in a taxable transaction, or the owner qualifies for an exception.

Two exceptions matter for cost-seg owners:

• Real estate professional status (REPS) under IRC §469(c)(7) — the owner spends more than 750 hours and more than half of all personal services in real property trades or businesses, materially participates in the rental activity, and (typically) makes the §469(c)(7)(A) aggregation election. The full REPS analysis covers the documentation bar and the audit risk.
• Short-term rental carve-out under Treas. Reg. §1.469-1T(e)(3)(ii)(A) — average customer use of seven days or less takes the activity out of the §469(c)(2) per-se passive bucket, and material participation under §469(h) then determines whether losses are active.

Without one of those, the year-one deduction is suspended. The PAL screen at /cost-seg/?is\_re\_professional=false flags this trap before the engineering fee gets paid. An owner with a W-2 day job, a single long-term rental, and no STR carve-out is the textbook case where a study produces a deduction the return cannot absorb.

Condition 4 — §163(j) electing real property trade — the ADS lockout

A trap that almost no engineering-firm landing page mentions. IRC §163(j) caps business interest deductions at 30% of adjusted taxable income, but §163(j)(7)(B) lets a "real property trade or business" elect out of the cap. The price of that election is a permanent shift to the Alternative Depreciation System under IRC §168(g) for residential (30-year ADS life), nonresidential real property (40-year ADS life), and qualified improvement property — and ADS property is not eligible for bonus depreciation under §168(k)(2)(D).

Concretely: an owner who elected real-property-trade-or-business status to preserve mortgage interest deductibility on a leveraged portfolio gave up bonus depreciation on the building, the QIP, and the ADS-affected components going forward. A cost segregation study on a property already inside that election still reclassifies 5-year and 15-year personal property to shorter recovery periods, but the §168(k)(2)(D) lockout strips the 100% bonus benefit that drives the headline year-one number. The pull-forward shrinks from "100% of the reclassified bucket in year one" to MACRS straight-line over 5, 7, or 15 years — a fundamentally different economic story.

The §163(j) election is irrevocable. An owner who made it without modeling the cost-seg interaction has a structural answer: a study still adds modest acceleration, but the bonus-depreciation hammer is gone.

Why "20-30% reclassification" is selection-biased

Engineering-firm marketing pages routinely quote 20-30% of basis as the typical reclassification range. The IRS Cost Segregation Audit Techniques Guide Chapter 7.2, addressing residential rental specifically, puts 5-year personal property at roughly 5-15% of depreciable basis for single-family and small multifamily — about half the marketing claim. The two numbers are not measuring the same thing.

The brochure number combines 5-year and 15-year property, includes hardscape and landscaping that is typically modest on a tract-built single-family rental, and is drawn from the engineer's own portfolio of completed studies. That portfolio is selection-biased: properties where the screening math did not pencil out never become studies, never appear in the case-study deck, and never appear in the average. The realistic 5-15% range from the IRS ATG is what shows up across an unfiltered population of residential rentals — including the small, plain, recently-built tract houses where most of the upgrade-driven 5-year basis simply isn't present.

The constants table the feasibility calculator uses is calibrated to the IRS ATG range, not to the engineering-firm marketing range, precisely so the screening output reflects what an unselected residential property is likely to produce — not what a successful case study did produce.

The hidden cost: state non-conformity (CA, NY, NJ)

Federal bonus depreciation is the engine of cost-seg's year-one number. Several high-tax states do not conform.

• California — Rev. & Tax. Code §17250 decouples California from federal §168(k) bonus depreciation entirely. The state requires straight-line MACRS adjustment, which means a federal $30,000 year-one bonus deduction produces a $30,000 federal benefit and roughly a fifth or sixth of that on the California return — with annual schedule adjustments to track the state-vs-federal book/tax difference for the life of the asset.
• New York — Tax Law §208 decouples New York from federal bonus depreciation for general corporate and personal income tax purposes (with limited exceptions for resurgence-zone and certain manufacturing property). The same federal-vs-state schedule overhead applies.
• New Jersey — Title 54A decouples New Jersey gross income tax from federal bonus depreciation; partnerships and LLCs taxed as partnerships pass through the state-conforming (non-bonus) numbers.

For an owner in one of those states, the headline federal benefit is real, but the after-tax-after-state benefit is meaningfully smaller, and the recordkeeping cost (a state depreciation schedule for every reclassified asset for the full recovery period) eats further into the net. None of that shows up in a federal-only year-one estimate. A high-bracket California owner of a small rental can run the federal screen, see a positive number, pay $6,000 for a study, and end up roughly break-even after the California offset and the recordkeeping burden.

AMT exposure and §168(g) ADS election fallout

Two more landmines that rarely make the marketing.

Alternative Minimum Tax (IRC §55-§59) and the §56(a)(1) adjustment. Property depreciated under MACRS general system has historically required an AMT adjustment to reflect the longer ADS life. The 2017 TCJA softened this for most real property, and bonus-depreciated property is largely treated consistently between regular and AMT, but specific class-life property and certain pre-2018 acquisitions still trigger a §56(a)(1) adjustment. An owner with a high AMT exposure profile — large state tax deductions, large miscellaneous deductions, ISO exercise, certain private-activity-bond interest — should model the AMT effect, not just the regular-tax effect, before paying for a study.

The §168(g)(7) ADS election. An owner can elect to depreciate property under ADS instead of MACRS for any class of property in a given tax year. Once made, the election is irrevocable for the property class in that year. Owners who elected ADS in a prior year — sometimes for state-conformity reasons, sometimes inadvertently via §163(j)(7)(B) — locked themselves out of the bonus-depreciation engine for the affected property. A cost segregation study on top of an ADS election still reclassifies, but the year-one bonus is gone.

The combined message: cost-seg's economics are not federal-marginal-rate-times-reclassified-basis. They are federal-marginal-rate-times-reclassified-basis, minus state non-conformity drag, minus AMT adjustment if applicable, minus engineering fee, minus §1245 recapture differential at disposition — divided by the time-value discount of the years held. That formula has a wider range of negative outcomes than the brochures admit.

How the engineering-firm marketing skews the answer

The structural conflict in the cost-seg industry is straightforward. Engineering firms are paid to produce studies, not to advise against them. A firm that screens 100 leads, signs 30, and politely declines 70 has a worse income statement than a firm that signs 60. The incentive runs toward "more studies" — which means the marketing reaches for the property profiles where the math is most marginal, because that is where the unsold demand sits.

Three patterns recur in firm-published content:

• The 20-30% reclassification claim detailed above.
• Combined-savings figures that sum federal year-one benefit, state benefit, and time-value-of-money over a hold period, then present the total as if it landed in year one.
• No mention of the §469 passive activity loss limitation, the §1245 recapture cost on disposition, the §163(j) ADS lockout, or state non-conformity.

The honest version: cost segregation is the right answer for a specific subset of properties — depreciable basis north of $300,000, hold period of five years or more, owner with REPS or STR carve-out (or sufficient passive income to absorb the loss), no §163(j) ADS election in the chain, federal-conforming state. Outside that subset, the screening conversation is the work; the engineering study isn't.

The five questions a CPA asks before signing off on a study

A cost segregation study under the IRS Cost Segregation Audit Techniques Guide Chapter 4 has 13 elements that distinguish a quality study (engineer with construction-cost expertise, detailed component-level analysis, contemporaneous documentation, statistical sampling discipline, and so on). Before getting to those 13 elements, a CPA typically asks five qualifying questions:

1. 1. What is the depreciable basis? Below roughly $200,000-$250,000 on a residential rental, the engineering fee eats most of the year-one benefit.
2. 2. What is the planned hold period? Under three years, the §1245 recapture differential reverses much of the pull-forward.
3. 3. Does the taxpayer qualify under §469(c)(7) REPS, the Treas. Reg. §1.469-1T(e)(3)(ii)(A) STR carve-out, or have offsetting passive income? Without one, the deduction is suspended.
4. 4. Has any §163(j)(7)(B) electing-real-property-trade-or-business election been made — by the taxpayer, the partnership, or any tiered entity? ADS lockout under §168(g) eliminates bonus depreciation.
5. 5. What is the state of residence and do any conforming states attach? California, New York, New Jersey, and several others decouple from §168(k) and reduce the after-state benefit.

A "no" or "uncertain" on any of those generally pushes the conversation back to feasibility screening before any engineering fee is committed.

When DIY screening is enough vs. when an engineer is needed

A feasibility screening tool — like the /cost-seg/ calculator — answers the first-order question: given a property's basis, type, location, planned hold, and tax profile, does the federal year-one number look big enough to justify the engineering fee, after state drag and recapture? It is a screening tool, not a substitute for an engineering analysis. The IRS ATG Chapter 4 is explicit about what a defensible study requires: an engineer with appropriate construction-cost expertise, on-site or detailed-photographic component identification, contemporaneous source documentation, and statistical sampling protocols where applicable. A spreadsheet, however well-calibrated, does not produce that record.

The honest split: screening calculators answer "is it worth getting an engineer involved?" Engineering studies answer "what are the defensible numbers to put on the return?" The first is a 10-minute exercise; the second is a $4,000-$8,000, multi-week engagement. Confusing the two — using a screening estimate as the basis for a tax return — is what generates the audit exposure the IRS ATG was written to address. The cost-seg basics FAQ walks through where the line sits.

Run the numbers honestly

The contrarian frame is not "cost-seg is bad." It is "cost-seg is a leveraged tool, and leverage cuts both ways." The leverage is good when basis is large, hold is long, the §469 trap doesn't apply, the §163(j) election doesn't apply, and the state conforms. The leverage is bad when any of those reverse — and the engineering firm's incentive is to wave the bad cases through, because every screened-out property is a study not sold.

The honest answer comes from running the numbers with the back-end costs included, not just the front-end benefit. The /cost-seg/ calculator is a screening tool that surfaces the four conditions in this article — small basis, short hold, §469 trap, §163(j) ADS lockout — alongside the federal year-one estimate, so the screening conversation happens before the engineering fee is committed. If the screen returns a marginal number, the screen is the answer. If it returns a strong number, the screen is the prompt to call a CPA and an engineer.

---

Sources

• IRC §168(k) (bonus depreciation), as amended by the One Big Beautiful Bill Act (OBBBA) §70301, Pub. L. 119-21 (July 4, 2025).
• IRC §168(g) (Alternative Depreciation System).
• IRC §163(j) and §163(j)(7)(B) (electing real property trade or business; ADS consequence).
• IRC §469(a), §469(c)(2), §469(c)(7) (passive activity loss rules; real estate professional status).
• IRC §1245(a)(1), §1245(a)(3) (recapture as ordinary income on disposition of §1245 property).
• IRC §55-§59 and §56(a)(1) (Alternative Minimum Tax and depreciation adjustment).
• Treas. Reg. §1.469-1T(e)(3)(ii)(A) (seven-day average customer use carve-out from per-se passive treatment).
• Treas. Reg. §1.1245-4 (recapture and §1031 exchange interaction).
• IRS Cost Segregation Audit Techniques Guide, Chapter 4 (the 13 elements of a quality study).
• IRS Cost Segregation Audit Techniques Guide, Chapter 7.2 (Industry-Specific Guidance — Residential Rental; 5-15% range for 5-year personal property in single-family and small multifamily).
• California Revenue & Taxation Code §17250 (state non-conformity to federal §168(k) bonus).
• New York Tax Law §208 (state non-conformity to federal bonus depreciation).
• New Jersey Title 54A (gross income tax non-conformity to federal bonus).
• Hospital Corp. of America v. Commissioner, 109 T.C. 21 (1997) (cost segregation as defensible methodology).
• Whiteco Industries v. Commissioner, 65 T.C. 664 (1975) (six-factor permanence test for §1245 vs. §1250 line).
• IRS Notice 2026-11 (post-OBBBA bonus depreciation transition guidance).

---

Disclaimer. This article describes general federal tax concepts. TaxProtestTx (Nought Labs LLC) is a feasibility-screening tool, not tax advice or a cost segregation study. Calculator output cannot be relied on under Treasury Circular 230. Consult a qualified CPA, EA, or attorney before filing. Results are not guaranteed.