Summary
A cost segregation study is judged on two things: whether each component is placed in the right class, and whether the components were found at all. Misclassification creates audit exposure. Omission leaves deductions unclaimed, and because the component was never listed, nobody notices. This paper explains the legal tests the study applies, how SegFlow reads a property from the preparer's uploads, how costs are assigned and reconciled, and where the limits of a photo-based study lie. An illustrative example shows that coverage usually decides the result, ahead of classification.
Two kinds of error
Every item in a building is either part of the building or it is not. IRC (Internal Revenue Code) §1250 property is the building and its structural components, depreciated over 27.5 years for residential rental property or 39 years for nonresidential real property. IRC §1245 property is tangible personal property, which falls into the 5- or 7-year classes of MACRS (Modified Accelerated Cost Recovery System), the depreciation schedule of the Internal Revenue Service (IRS). Land improvements are a third group, recovered over 15 years under asset class 00.3 of Rev. Proc. 87-56.
Placing a structural component in the 5-year class overstates the first-year deduction, and it is the adjustment an examiner looks for first. Leaving an eligible component inside the building total understates the deduction, and the gap is invisible after the fact. A study can be conservative in every classification and still be wrong by omission.
The tests the study applies
Treas. Reg. §1.48-1(e)(2) lists the structural components: walls, partitions, floors, ceilings, permanent coverings such as paneling or tiling, windows and doors, central heating and air conditioning, plumbing and plumbing fixtures, electric wiring and lighting fixtures, chimneys, stairs, elevators, sprinkler systems, and other components relating to the operation or maintenance of the building. An item on that list stays in §1250 unless a recognized exception applies.
For items the list does not settle, the study applies the six factors from Whiteco Industries v. Commissioner, 65 T.C. 664 (1975): whether the item can be moved and has been; whether it was designed to remain in place; the expected length of affixation; how substantial removal would be; how much damage removal would cause; and the manner of attachment. Hospital Corporation of America v. Commissioner, 109 T.C. 21 (1997) confirmed that these investment-credit-era tests govern depreciation, which is why branch wiring dedicated to equipment, plumbing serving specific equipment, carpet and certain removable coverings can be §1245 property. Recovery periods then follow Rev. Proc. 87-56: class 57.0 (distributive trades and services) at 5 years for most personal property in rental and service uses, 7 years for personal property with no assigned class life, and class 00.3 at 15 years for land improvements.
Each ledger line carries the class and the authority relied on, so the reviewer sees the reasoning rather than a bare label.
How the study reads a property
The preparer uploads documents and images. The closing statement or purchase contract supplies the basis and the closing date, which drives the bonus depreciation test and the placed-in-service year. The county assessor record supplies the land and improvement split, the year built and the square footage. Photos of every room, the four exterior sides, the mechanical spaces and the site supply the components. An appraisal, renovation invoices, a site plan, a prior depreciation schedule or a lease help where available.
Before anything runs, the intake step shows what was found in the uploads, the assumptions the study will make unless changed, and at most two or three questions that change the result: the source of the land value, the use type or average customer stay, whether the property was furnished at purchase, and the year for a look-back study. The rooms seen in photos are compared with the bedroom and bathroom counts and the square footage. An area that was not photographed becomes a stated assumption in the study, never a silent gap, and anything that cannot be photographed can be described in text.
The study then catalogs the components in every photo and document in one pass: name, quantity and unit, location, the photos that evidence it, the class with its authority, and the cost source. Items considered and not reclassified are listed as exclusions with the reason, so the reviewer sees what was rejected as well as what was accepted.
How costs are assigned and reconciled
The vast majority of components are priced from benchmark and market data, using benchmark pricing adjusted for location and purchase date. Where a component has no close benchmark, the study estimates it and writes the basis for the estimate into the ledger line. All component costs are then reconciled to the depreciable basis: the reclassified components plus the residual building equal purchase price less land. Reconciliation of allocated costs to actual costs is one of the 13 principal elements of a quality study in the IRS Cost Segregation Audit Techniques Guide (ATG, the examiner's handbook for these studies), and it is the check that stops a study from pricing components as if the building were worth more than the client paid.
Where evidence is thin, the study defaults to the longer life. A hard-wired fixture that may or may not be dedicated to equipment stays in the building unless the photos or documents show otherwise, and the review notes tell the preparer what evidence would support the shorter life.
An illustrative example
Round numbers for illustration; this is not a client result. An eight-unit apartment building is bought for $2,000,000. The assessor record supports $400,000 of land, leaving $1,600,000 of depreciable basis. From the uploads, the study lists $300,000 of 5-year property (unit appliances, carpet and pad, window coverings, decorative fixtures in the common areas, dedicated circuits and venting for laundry equipment) and $120,000 of 15-year land improvements (front parking, sidewalks, fencing, landscaping, site lighting). Reclassified basis is $420,000, about 26 percent.
The intake step had noted that photos covered six of the eight units and that the rear of the lot did not appear in any photo. The study carried both as assumptions: the two unseen units were assumed to match the six seen, and the rear lot was excluded. The preparer forwards the photo checklist to the owner, who replies with pictures of the remaining units and of the rear parking area with its fenced dumpster enclosure. The included second run adds $30,000 of 15-year site work and $10,000 of 5-year equipment wiring and exhaust in the second laundry room. Reclassified basis rises to $460,000, about 29 percent.
Under 100% bonus depreciation the added $40,000 is deductible in the first year, worth about $14,800 at a 37 percent marginal rate. No classification changed between the two versions. The difference was coverage, and the study made the gap visible before the return was filed.
Where photos are not enough
A photo-based study infers what it cannot see. In-wall wiring, below-grade plumbing and utility runs are inferred from visible panels, fixtures and equipment, and each inference is stated as an assumption with its consequence. There is no site visit, no interview with the contractor and no engineer's signature. For large or specialized commercial property, or where an examination is likely, an engineered study with a site inspection may be the right instrument, and we say so when asked.
What this means for your firm
The study arrives as a ledger in which every line shows its evidence, class, authority and cost source, and every omission is written down as an assumption. Your review time goes to the lines the review notes single out and to the questions only you can answer, such as whether the property was furnished at purchase. Because coverage is visible at intake, the fastest way to improve a result is to send the owner the checklist before the study runs, not to argue with a classification afterward. The study is a workpaper prepared with software assistance for your review and signature; the judgment and the signature remain yours.