Pushing Your DIY Cost Seg Envelope

By Jason Watson, CPA
Posted Sunday, July 12, 2026

It is not surprising that many rental property owners who use a do-it-yourself cost segregation tool such as CostSegEZ.com want to push down land values. Lower land value means higher building value. Higher building value means more depreciable basis. More depreciable basis means more property that might be accelerated. See the temptation?

Careful.

As mentioned previously, the easiest starting point is usually the county assessor’s land-to-building ratio. Let’s say you purchase a rental property for $500,000, and according to the assessor’s data, the parcel is assessed at $45,000 land and $180,000 building, for a total assessed value of $225,000. In this example, land is 20%. Using that ratio, you might allocate $100,000 of your $500,000 purchase price to land, leaving $400,000 as depreciable building value.

The Internal Revenue Manual, Part 4, Examining Process, Chapter 48, Engineering Program, Section 6, Real Property Valuation Guidelines, specifically 4.48.6.3, says “assessment and tax data” should be considered when determining value. Ok, that is quite the mouthful.

How about a tax court case? In Nielsen v. Commissioner, Tax Court Summary Opinion 2017-31, the court found the county assessor’s allocation between land and improvements more reliable than the taxpayers’ proposed values. The court also noted that the taxpayers did not identify an authority suggesting they were qualified to allocate value between land and improvements on their own. Shocking.

Sidebar: Worth noting, the case involved Los Angeles County properties, and the court leaned on the assessor numbers partly because the taxpayers’ alternative methods were weak.

Furthermore, in Nielsen, an independent appraisal on one of the properties actually lined up with the assessor-based result, which is a nice reminder that assessor data and an appraisal can back each other up rather than compete. Boom, but with a caution label: this court matter was a summary opinion, so treat it as persuasive, not a magic shield.

If you disagree with the county assessor ratio method, then spend the money on an appraisal. The opinion of a disinterested third party will usually look a lot better than “I just really wanted more depreciation.” Funny how that works.

Another sanity check is average construction cost per square foot. If you purchase a lovely short-term rental property in Malibu for $2.5 million and want to claim that $1.8 million is allocable to the building on an 1,800 square foot property, you are implying $1,000 per square foot of building value. Is that impossible? No. Location, location, location, right? High-end finishes, complex construction, coastal requirements and expensive markets can get silly fast. But you need support. The more aggressive the allocation, the more the facts need to carry the weight.

Yet another consideration is the discussion and application of replacement cost found in the Cost Segregation Audit Techniques Guide. Here is a blurb-

When construction cost information for the acquired property is not available, it must be estimated using the construction cost data, methods, and techniques normally employed for a property appraisal. These estimated property costs (replacement cost new (RCN)), which include indirect costs, are then adjusted for the asset’s age and condition at the acquisition date, including physical depreciation, functional obsolescence, and economic obsolescence. These adjustments are generally expected to be different among the various items of acquired property (e.g., a building and its structural components, land improvements, and personal property), given that each of these items of property have varying expected useful lives, levels of use, and may have been constructed and/or installed on different dates. The total replacement cost new less depreciation (RCNLD) of the acquired items of property, along with the fair market value (FMV) of the land, should reasonably approximate the total purchase price of the acquired property.

Ok, neat. Later on, in the ATG, the IRS issues this guidance for examination-

Ensure that Replacement Cost Values are Properly Adjusted
for the actual condition and remaining economic useful life of the assets.

The value of used components must be reduced from the replacement cost new value in proportion to the observed economic obsolescence or physical depreciation as compared to similar new assets. This principle is discussed in regard to the “Helipot Building” in Lesser v. Commissioner, 42 T.C. 688 (1964), aff’d, 352 F.2d 789 (9th Cir. 1965), acq., 1966-2 C.B. 5, cert. Denied, 384 U.S. 927 (1966).

And they pile on with this-

Real Estate Allocations
The fair market value of land should be based on the highest and best use of the land as though vacant, even if the land has improvements. The land value may equal the value of the total real estate even if the real estate has substantial improvements when such improvements do not contribute value to the property. Whereas land has value, improvements contribute value. The value of the total real estate, less the value of the land, results in the contributory value of the improvements. Accordingly, it is inappropriate to estimate the value of the land by subtracting the estimated value of the improvements from the lump real estate price. Basis assigned to land in this residual fashion may result in understating the appropriate basis in the land and overstating the appropriate basis in the depreciable improvements. Examiners should also be wary if a cost segregation study relies solely on local assessed values rather than appropriately determining fair market values.

Yuck.

So, what is the bottom line? The county assessor ratio method is often a reasonable starting point, and Nielsen gives taxpayers a helpful example where assessor data carried the day. But do not treat the assessor ratio as a license to jam land value as low as possible. Use the assessor data, consider an appraisal when the dollars matter, and sanity-check the result against building cost, age, condition, location and replacement cost. If the land allocation feels like it was reverse-engineered to make the depreciation number delicious, it probably needs more support.