The IRS Cost Segregation Audit Technique Guide: What Examiners Are Actually Told to Look For
The Rare Case Where the IRS Shows Its Work
Cost segregation occupies an unusual position in tax practice. There is no statute that authorizes it by name, no regulation that prescribes a methodology, and no form to file. What exists instead is a body of case law running back to Hospital Corporation of America, and an Audit Technique Guide the IRS publishes for its own examiners.
The ATG is publicly available. It tells you, in the government's own words, how a study will be evaluated if it is examined. Reading it is the closest thing available to seeing the grading rubric before the test.
It is also the reason quality varies so much across providers. Firms that build to the ATG produce reports that look a certain way. Firms that do not produce reports that are shorter, cheaper, and considerably harder to defend.
The Thirteen Principal Elements
The guide identifies thirteen elements it associates with a quality study. Paraphrased, they are: preparation by an individual with expertise and experience; a detailed methodology description; use of appropriate documentation; interviews with people who have knowledge of the property; a common nomenclature; a standard numbering system; an explanation of the legal analysis; a determination of unit costs and engineering takeoffs; an organized and detailed asset listing; reconciliation of total costs; an explanation of the treatment of indirect costs; identification and listing of Section 1245 property; and consideration of related aspects such as Section 263A and change in accounting method requirements.
None of these are optional in practice. A study missing several is not a study the IRS considers reliable, and the examiner is instructed to say so.
The element most commonly missing from low-cost studies is the engineering takeoff. That is the actual measurement and unit-cost work. Without it, the reclassification is an estimate dressed as an analysis.
The Methodologies the Guide Recognizes
The ATG describes six approaches and ranks them implicitly by reliability. The detailed engineering approach from actual cost records is the gold standard, available where original construction invoices exist. The detailed engineering cost estimate approach is the standard for acquired property where records do not exist, and it is what most quality studies on purchased buildings use.
Below those sit the survey or letter approach, the residual estimation approach, the sampling or modeling approach, and the rule of thumb approach. The guide is direct about the last one: rule of thumb studies have little or no supporting documentation and are not considered reliable.
If a provider quotes you a price that seems impossibly low and turns the study around in three days without a site visit or a document request, you are almost certainly buying a rule of thumb study. It will produce a number. It will not produce support.
What Examiners Are Told to Challenge
The guide directs examiners toward specific pressure points. Classification of electrical and plumbing systems is one: the distinction between building systems serving the structure generally, which are structural, and components serving specific equipment or process needs, which may be 1245 property, is fact-intensive and frequently overreached.
Site improvements are another. The line between land, which is non-depreciable, land improvements at 15 years, and building components is often drawn aggressively.
Cost allocation and reconciliation get sustained attention. The examiner is looking for whether the sum of the reclassified components plus the remaining structural basis equals the total depreciable basis. Reports that do not reconcile invite the conclusion that the numbers were assembled rather than derived.
The Change in Accounting Method Overlay
The guide devotes real attention to the mechanics of applying a study to a property placed in service in a prior year. That is a change in method of accounting requiring Form 3115, filed under the automatic consent procedures, with a Section 481(a) adjustment.
Examiners check whether the Form 3115 was filed, whether the 481(a) computation is supported, and whether the taxpayer improperly amended prior returns instead of using the method change. Amending is generally not permitted for depreciation method changes after the second year, and doing it anyway is a straightforward adjustment for an examiner to make.
This is one of the more common procedural failures in self-prepared or low-cost studies. The engineering may be fine and the filing mechanics wrong, which produces the same outcome as bad engineering. AE Tax Advisors handles the procedural side in their Form 3115 cost segregation guidance.
How to Read Your Own Report Against the Guide
Open your report and look for four things. Is there a stated methodology naming which of the recognized approaches was used? Is there an asset-level listing with quantities, unit costs, and extended totals rather than category percentages? Is there a reconciliation tying components back to total basis? Is there a legal analysis section citing the authorities relied on?
If all four are present, your report is built the way the guide expects. If your report is twelve pages of summary tables with a percentage allocation and no takeoff detail, it will not hold under examination, and the price you paid was not a bargain.
Ask who performed the work. The guide expects expertise in both construction and tax. A study signed by someone with neither is a document, not an opinion.
What This Means for Choosing a Provider
The practical takeaway is that the standard is knowable and testable. You do not have to take a provider's word about quality. You can ask which of the six methodologies they use, whether they perform a site visit, whether they produce engineering takeoffs, and whether they provide audit support if the study is examined.
Providers building to the ATG answer those questions easily because the answers are their selling points. Providers who do not tend to redirect toward price and turnaround time.
Every Stratum study is built to the thirteen elements, includes the reconciliation and legal analysis sections, and comes with audit support at no additional cost. That is not a differentiator so much as a baseline, but it is a baseline a surprising share of the market does not meet.