When cost segregation doesn't make sense

The cost segregation industry has a marketing problem, which is that the product is genuinely useful and therefore gets sold to people for whom it is useless. The study accelerates depreciation; it does not create it. The benefit is the time value of deferred tax, and time value can be small, zero, or negative once you account for the study fee, ordinary-income recapture on sale, suspended passive losses, and the owner's actual bracket. A CPA's job is to run that math before the engineering firm's proposal does it for them.

This piece is the checklist of situations where the honest answer is "no," "not yet," or "smaller than the proposal says." It is the counterweight to what a cost segregation study is and the bonus depreciation interaction, and it is where we think the analysis should start, not end.

Short holding periods: deferral needs runway

The core benefit is an interest-free loan from the Treasury: deduct now, repay (via forgone future depreciation and recapture) later. A loan's value depends on how long you keep it. Sell in year two and the loan is called almost immediately — and repaid on worse terms.

The terms are worse because of character. Depreciation claimed on reclassified 5-, 7-, and 15-year Section 1245 property is recaptured as ordinary income on sale, to the extent of gain — up to 37 percent federally. Straight-line depreciation on the building itself, by contrast, generates unrecaptured Section 1250 gain taxed at a maximum of 25 percent. Cost segregation therefore trades a 25-percent-rate future liability for a 37-percent-rate one, in exchange for the use of the money in between. Over fifteen years, that trade wins comfortably. Over three, the rate spread can eat most of the time-value gain, and the study fee eats the rest. The full arithmetic is in depreciation recapture explained.

A reasonable screen: if the realistic holding period is under about five years, the study needs unusually favorable facts (very high reclassification percentage, very high current bracket, low expected gain) to pencil.

The passive loss trap: a deduction you can't touch

This is the most common failure mode, and the one promoters mention least. Under Section 469, rental real estate losses are per se passive unless the owner is a real estate professional — more than 750 hours and more than half of personal-services time in real property trades or businesses, plus material participation in the rental (often requiring a grouping election for multiple properties). Passive losses offset passive income only; the excess is suspended and carried forward.

Now run the standard pitch through that filter. A surgeon buys a $2 million rental; a study plus 100 percent bonus generates a $500,000 year-one loss; the proposal multiplies by 37 percent and advertises $185,000 of tax savings. But the surgeon is not a real estate professional, has no passive income, and the $500,000 loss goes into a suspension account — where it sits, at zero present value, until the property is sold in a fully taxable disposition or passive income materializes. (The $25,000 active-participation allowance of Section 469(i) phases out entirely at $150,000 of AGI, so it rarely helps the clients being pitched.)

The short-term-rental fact pattern is the well-known workaround — average stays of seven days or less take the activity out of the "rental" definition, so material participation alone suffices — but it has real requirements and real audit attention, covered neutrally in our STR brief. Even non-passive losses then face the Section 461(l) excess business loss cap and, for pass-through owners, basis and at-risk limits.

Low brackets and NOL positions: the wrong year to deduct

A deduction is worth its marginal rate. Accelerating $400,000 of depreciation into a year where the owner's marginal federal rate is 12 percent — a retirement year, a start-up loss year — while forgoing those deductions in future 32-percent years is negative arbitrage, before considering that recapture on sale will still come back at ordinary rates.

The same logic applies to taxpayers already in an NOL position. Adding depreciation to a loss year just enlarges a carryforward that (post-2017) offsets only 80 percent of future taxable income and carries no interest. And C corporations should note the flat 21 percent rate: the corporate deferral benefit exists but is smaller per dollar than the individual-rate examples in most proposals, while a corporation contemplating asset-sale exit faces the same 21 percent on recapture — muting both sides.

The right response to a temporarily low bracket is usually timing, not abstinence: defer the study, or take it but elect out of bonus by class under Section 168(k)(7) to spread deductions into higher-rate years. See depreciation basics for the election mechanics.

Planned 1031 exchanges

Owners who intend to exchange under Section 1031 within a few years get a compressed version of the short-hold problem plus complications of their own. Post-2017, Section 1031 applies only to real property; the regulations under §1.1031(a)-3 define real property broadly for exchange purposes (many items a study classifies as Section 1245 property for depreciation still count as real property for 1031), so a study does not usually blow the exchange — but it does not defer the Section 1245 recapture story forever either. The carryover-basis mechanics mean the accelerated deductions reduce exchanged basis, the replacement property's depreciable base shrinks, and the ordinary-income recapture attribute rides along until a taxable sale finally reckons it, potentially at higher future rates.

For a serial exchanger with indefinite deferral intentions, the study can still work — the loan may never be called during life, and death delivers a basis step-up that extinguishes it. That is a legitimate estate-horizon strategy. But it requires the owner to actually mean "hold till death," and most people who say that in year one sell in year six.

Small buildings: the fee is the hurdle

Study fees are largely fixed against basis. A quality engineering-based study on a modest property might cost $5,000 to $8,000; the property might have $400,000 of depreciable basis with 20 percent reclassifiable. That is $80,000 of accelerated deductions — a present-value benefit on the order of $10,000 to $15,000 at typical rates and discount assumptions. Positive, but a thin margin for the audit-profile and recordkeeping burden, and it goes negative if any of the factors above applies. Some providers sell cheaper "modeled" or fully remote studies for small properties; those are exactly the residual-method products the IRS Audit Techniques Guide ranks lowest, so the fee savings buys examination risk.

There is no statutory floor, but as a practical screen, below roughly $500,000–$750,000 of depreciable basis the burden of proof shifts to the proposal.

Other situations that deserve a pause

• Imminent gift or estate transfers. Property expected to pass through an estate gets a basis step-up; depreciation harvested now is recaptured never, which favors studies for hold-till-death owners — but property intended for lifetime gifts carries its basis and its recapture attributes to the donee.
• Tax-exempt and low-income owners, including entities with unrelated-business considerations, may have no appetite for deductions at all.
• States that decouple. In nonconforming states, the study creates a permanent federal/state depreciation divergence to track for decades. The federal math may still work; the compliance cost is real and belongs in the analysis.
• Electing real property trades or businesses under Section 163(j)(7) must depreciate real property under ADS, which eliminates bonus on 15-year land improvements and QIP — shrinking the study's yield for exactly the leveraged owners most likely to have made the election.

The honest sequencing question

For most owners the question is rarely "study or no study" but "study when." Because look-back studies via Form 3115 capture every missed dollar as a one-year Section 481(a) adjustment with no amended returns, the option to study later is nearly free — the cost of waiting is time value on the interim years, not the deductions themselves. That option value cuts against commissioning a study in a year when the deduction would suspend, be wasted in a low bracket, or precede a near-term sale.

The strategy is sound. The engineering is real. The law is settled. None of that makes it right for this client, this building, this year — and a proposal that never mentions Section 469, recapture, or the seller's timeline is telling you who it was written for.