Rate snapshot
| Fed funds target | 3.75% to 4.00% | Set September 16, 2026, a 25 bp increase |
| 5-year Treasury | 5.02% | October 9, 2026 close, up 3 bp from October 8 |
| 10-year Treasury | 5.24% | October 9, 2026 close, down 7 bp from October 5 |
| Top federal ordinary rate, 2026 | 37% | Taxable income above $768,700, married filing jointly |
| Unrecaptured section 1250 gain | 25% max | Plus the 3.8% net investment income tax |
| Industrial cap rates | Compressed | CBRE survey of first-half 2026 closings |
The bond market is closed today for Columbus Day, so Friday’s curve is the newest print anyone has: the 5-year at 5.02% and the 10-year at 5.24%. At that cost of capital, a lot of middle-market sponsors are getting to their return target through the tax line rather than the rent line, and cost segregation is the first lever they reach for. The year-one number is genuinely large. On the industrial deal modeled below, a study moves $2,346,850 of basis out of the 39-year schedule and produces a $957,515 first-year tax deferral.
What almost nobody models is the other side of that entry. The deduction is not forgiven, it is borrowed, and at sale it comes back at a rate you do not choose. On a five-year hold, $451,586 of that $957,515 gets handed back, in present-value terms, leaving $325,908 after the study fee. That is 34% of the headline.
The non-obvious part is where the expensive recapture comes from. It is not the equipment. On this deal, 74.5% of the ordinary-income recapture is generated by the 15-year site work, because bonus depreciation on land improvements is “additional depreciation” under section 1250(b)(1) and comes back as ordinary income, not as unrecaptured section 1250 gain at 25%. Sponsors who assume the site-work bucket exits at 25% are mispricing it by 12 full points.
Two code sections, two very different exit rates
Cost segregation does not create deductions out of nothing. It moves basis from a 39-year straight-line schedule into shorter-lived classes that qualify for the 100% special depreciation allowance, which the One Big Beautiful Bill Act reinstated for property acquired and placed in service after January 19, 2025. Where that basis lands determines how it is taxed at exit, and there are three distinct answers.
Section 1245 property, which is personal property such as racking, process equipment and specialty electrical, recaptures as ordinary income. Section 1245(a)(1) treats as ordinary income the excess of the lower of recomputed basis or amount realized over adjusted basis. In plain terms, every dollar of depreciation you took on that equipment comes back at your ordinary rate, up to the gain allocable to it.
Section 1250 property is the category sponsors misread. Under section 1250(c) it means any depreciable real property that is not section 1245 property, which includes land improvements: paving, curbs, yard, site lighting, fencing, landscaping. Section 1250(a)(1) recaptures “additional depreciation” as ordinary income, and section 1250(b)(1) defines additional depreciation for property held more than a year as the adjustments that exceed what straight line would have produced.
That definition is the whole problem. A 100% bonus deduction on a 15-year asset is wildly in excess of straight line, so most of it is additional depreciation. Only the straight-line-equivalent slice escapes into unrecaptured section 1250 gain, where the 25% maximum rate applies.
| Where basis lands | Code section | Exit character | Top rate with NIIT |
|---|---|---|---|
| 5-year personal property | 1245 | Ordinary income, all depreciation | 40.8% |
| 15-year land improvements, bonus portion | 1250(a)(1) | Ordinary income, additional depreciation | 40.8% |
| 15-year land improvements, straight-line portion | 1250 | Unrecaptured section 1250 gain | 28.8% |
| 39-year building, straight line only | 1250 | Unrecaptured section 1250 gain | 28.8% |
| Appreciation above original basis | 1231 / 1221 | Long-term capital gain | 23.8% |
The deal
A sponsor buys a 118,000 square foot small-bay industrial building for $12,400,000 with $150,000 of capitalized closing costs. Land is 15% of total basis. The property is placed in service in July, held 60 months, and sold for $14,200,000 with 1.5% selling costs. The sponsor is a passive investor in the top 2026 bracket, so ordinary income carries 37% plus the 3.8% net investment income tax.
| Basis item | Amount | Share |
|---|---|---|
| Purchase price | $12,400,000 | |
| Capitalized closing costs | $150,000 | |
| Total basis | $12,550,000 | 100.0% |
| Land, not depreciable | $1,882,500 | 15.0% |
| Depreciable basis | $10,667,500 | 85.0% |
| Study: 5-year personal property | $853,400 | 8.0% |
| Study: 15-year land improvements | $1,493,450 | 14.0% |
| Study: 39-year building | $8,320,650 | 78.0% |
| Total reallocated out of 39-year | $2,346,850 | 22.0% |
The 8% and 14% allocations are the point of the example rather than a market average. Industrial is paving-heavy and equipment-light, which is exactly why the site-work problem shows up more sharply here than on an apartment deal. A real study on your building will produce its own percentages.
What the study does to the depreciation schedule
Year one is the whole story, and years two through six are the giveback. Once $2,346,850 has been expensed, the remaining 39-year basis is smaller, so annual depreciation during the hold is lower than it would have been without the study. The cumulative extra deduction over the hold is $2,045,972, not the full $2,346,850.
| Year | No study | With study | Difference |
|---|---|---|---|
| 1 | $125,366 | $2,444,635 | $2,319,269 |
| 2 | $273,526 | $213,350 | ($60,176) |
| 3 | $273,526 | $213,350 | ($60,176) |
| 4 | $273,526 | $213,350 | ($60,176) |
| 5 | $273,526 | $213,350 | ($60,176) |
| 6, through sale | $148,160 | $115,565 | ($32,595) |
| Total | $1,367,628 | $3,413,600 | $2,045,972 |
The exit without a study
Net proceeds are $13,987,000. Allocate $2,200,000 to land and the residual $11,787,000 to the building, against a building basis of $9,299,872, for a building gain of $2,487,128. All depreciation was 39-year straight line, so there is no additional depreciation and no ordinary recapture at all. The accumulated depreciation becomes unrecaptured section 1250 gain, capped by gain, and the rest is long-term capital gain.
| Gain component | Amount | Rate | Tax |
|---|---|---|---|
| Ordinary recapture | $0 | 40.8% | $0 |
| Unrecaptured section 1250 gain | $1,367,628 | 28.8% | $393,877 |
| Long-term capital gain | $1,437,000 | 23.8% | $342,006 |
| Total gain and tax | $2,804,628 | 26.2% | $735,883 |
The exit with a study
Now the proceeds have to be allocated among asset classes, because recapture is computed asset by asset. Assume the 5-year equipment is worth 40% of original cost after five years and the site work holds 90%, with the building taking the residual. Both short-life buckets have zero adjusted basis, since bonus expensed them entirely in year one.
| Exit allocation | Proceeds | Adjusted basis | Gain |
|---|---|---|---|
| Land | $2,200,000 | $1,882,500 | $317,500 |
| 5-year personal property | $341,360 | $0 | $341,360 |
| 15-year land improvements | $1,344,105 | $0 | $1,344,105 |
| 39-year building | $10,101,535 | $7,253,900 | $2,847,635 |
| Total | $13,987,000 | $9,136,400 | $4,850,600 |
The site-work line is where the arithmetic turns. Straight line on $1,493,450 over 60 of 180 months would have been $497,817. Bonus took the full $1,493,450, so additional depreciation is $995,633, which is 66.7% of the bucket.
Section 1250(a)(1) then recaptures all $995,633 as ordinary income, because it is less than the $1,344,105 gain on that asset. Only $348,472 of the remainder falls into unrecaptured section 1250 gain, and nothing on that asset reaches the capital gain rate.
| Gain component | Amount | Rate | Tax |
|---|---|---|---|
| Section 1245 ordinary, equipment | $341,360 | 40.8% | $139,275 |
| Section 1250 ordinary, site work | $995,633 | 40.8% | $406,218 |
| Unrecaptured section 1250 gain | $1,415,222 | 28.8% | $407,584 |
| Long-term capital gain | $2,098,385 | 23.8% | $499,416 |
| Total gain and tax | $4,850,600 | 29.9% | $1,452,493 |
Total gain rose by $2,045,972, exactly the extra depreciation taken, which is expected. The blended rate on gain rose from 26.2% to 29.9%, and the exit tax bill rose by $716,610. Ordinary income went from zero to 27.6% of total gain.
Where the ordinary income actually comes from
Of the $1,336,993 recaptured at ordinary rates, the 5-year equipment accounts for $341,360 and the 15-year site work accounts for $995,633. The bucket most sponsors treat as the safe, boring, real-property part of the study generates three quarters of the expensive recapture.
| Source of ordinary recapture | Amount | Share |
|---|---|---|
| 5-year personal property, section 1245 | $341,360 | 25.5% |
| 15-year land improvements, section 1250 | $995,633 | 74.5% |
| Total ordinary recapture | $1,336,993 | 100.0% |
Price the mispricing directly. If all $1,336,993 recaptured as unrecaptured section 1250 gain at 28.8%, the tax would be $385,054. At 40.8% it is $545,493.
The 12-point differential costs $160,439, which is 1.29% of the purchase price and 16.8% of the headline year-one deferral. On the site-work bucket alone the penalty is $119,476.
Simplifications: top 2026 federal rates only, no state income tax, no section 1231 loss netting, no suspended passive losses released at sale, and no entity-level allocation among partners. The 3.8% net investment income tax is applied to all gain components on the assumption of a passive investor above the MAGI threshold. Allocation percentages, exit price, selling costs, residual values and the $15,000 study fee are assumptions, not market data; depreciation uses a mid-month convention for 39-year property placed in service in July and 60 months of ownership. Dollar figures are rounded, so a column may differ from its displayed components by $1.
The net, after discounting
Run both sides through a discount rate and the decision becomes legible. The extra deductions are worth $792,493 in present value at 8%, almost all of it in year one. The extra exit tax of $716,610 arrives in year six, so its present value is $451,586.
| Component | Nominal | PV at 8% |
|---|---|---|
| Extra deductions, years 1 through 6 | $2,045,972 | |
| Tax saved on those deductions | $834,756 | $792,493 |
| Extra exit tax, year 6 | ($716,610) | ($451,586) |
| Cost segregation study fee | ($15,000) | ($15,000) |
| Net value of the study | $103,146 | $325,908 |
The study is still clearly worth doing: $325,908 of present value on a $15,000 fee. But the honest framing is that a sponsor keeps 34.0% of the headline number, not 100% of it. The remaining 47.2% is a loan from the Treasury at the price of a rate upgrade from 28.8% to 40.8% on most of the reallocated basis.
The cushion is wide, which is the reassuring part. Holding the deduction side fixed, the extra exit tax would have to reach $1,233,784 before the study broke even. Actual extra exit tax is $716,610, leaving $517,174 of room.
Sensitivity: hold period and discount rate
Here the exit price grows at 2.75% a year from the purchase price rather than being fixed, so each hold period gets its own gain. Two results run against intuition.
| Hold | 6% discount | 8% | 10% | 12% |
|---|---|---|---|---|
| 2 years | $169,256 | $190,968 | $210,565 | $228,251 |
| 3 years | $210,697 | $241,084 | $267,969 | $291,745 |
| 5 years | $282,885 | $325,908 | $362,473 | $393,500 |
| 7 years | $342,590 | $393,366 | $434,839 | $468,602 |
| 10 years | $412,770 | $468,897 | $512,179 | $545,328 |
First, the study gets more valuable as the hold lengthens, not less. A two-year hold at 8% nets $190,968 against $325,908 at five years and $468,897 at ten. The reason is that additional depreciation on the 15-year bucket decays as straight line catches up, so the ordinary slice shrinks, and the recapture bill is pushed further out.
| Hold | Extra exit tax | PV of deductions at 8% |
|---|---|---|
| 2 years | $804,444 | $844,562 |
| 3 years | $775,166 | $825,854 |
| 5 years | $716,610 | $792,493 |
| 7 years | $658,053 | $763,892 |
| 10 years | $570,219 | $728,454 |
Second, a higher discount rate helps the study rather than hurting it. The benefit is front-loaded and the cost is back-loaded, so impatient capital likes this trade. A sponsor underwriting to a 12% cost of equity gets more out of a study than one underwriting to 6%, which is the opposite of how most other tax timing decisions behave.
Does a 1031 exchange solve it?
Partly, and with a catch that cost segregation makes worse. Section 1031(a)(1) now reads “no gain or loss shall be recognized on the exchange of real property,” the 2017 amendment having removed personal property from its scope. A fully deferred exchange of real property defers the section 1250 recapture on the building and the site work along with everything else.
The section 1245 bucket is the exposure. Equipment reallocated out of the building is not real property, so it is not like-kind, and to the extent proceeds are allocable to it the sponsor can recognize ordinary income even inside an otherwise clean exchange. On this deal that is $341,360 of ordinary income, or $139,275 of tax, in a transaction a sponsor believed was fully deferred.
Whether a given item is real property for section 1031 purposes is a separate question from whether it is section 1245 property for depreciation purposes, and the two do not always line up. Get that mapping done by your CPA before the exchange closes, not after.
Nine checks before you commission a study
- Ask for the 15-year land improvement number separately. It is usually the largest reallocated bucket on industrial and retail, and it is the one that recaptures as ordinary income.
- Model the exit tax in the same spreadsheet as the year-one deduction. A study proposal that shows only the deduction is showing you half the trade.
- Confirm your actual marginal rate. The 12-point ordinary-versus-25% gap is the entire cost of the strategy, and it narrows for sponsors below the top bracket.
- Add your state. Several states do not conform to bonus depreciation, which can delay the benefit while doing nothing to soften the recapture.
- Check whether the deduction is usable. Passive activity rules can suspend the loss, and a deduction you cannot use this year is worth far less than the proposal assumes.
- Decide on the exit path first. A planned 1031 changes the analysis, and a planned taxable sale inside three years changes it more.
- Price the partial-asset disposition benefit too. Having a cost segregation study in place lets you write off a replaced roof or parking lot instead of depreciating two.
- Ask how proceeds will be allocated at sale, and get the purchase and sale agreement to say it. An allocation drafted by the buyer’s advisers will not favor you.
- Keep the study. Recapture is computed from your depreciation history, and a study you cannot produce in an exam is a study you did not buy.
Red flags in a study proposal
Three patterns should slow you down. A proposal quoting a reallocation percentage before anyone has walked the building is selling a template, not an engineering study. A proposal that compares the year-one deduction to the fee and calls the ratio a return is ignoring the entire liability side. And a proposal that describes recapture as “25% on the real property” without distinguishing additional depreciation from straight-line depreciation has the law wrong in the direction that flatters the sale.
None of this argues against cost segregation. On the deal above it creates real value at every hold period and every discount rate tested. It argues for underwriting it as what it is: a financing decision with a known cost, not a discount.
How QuadBlock can help
We finance lower middle-market industrial, multifamily, retail, self-storage and specialty assets from $250K to $30M, and we underwrite the exit alongside the entry because that is where leverage and tax decisions meet. If you are weighing a hold-versus-sell question, a refinance, or a bridge-to-stabilization plan where the after-tax outcome drives the structure, send the deal to us at quadblockcapital.com. We will tell you quickly what we can do and where the structure needs work.
This article is for general educational purposes and is not tax, legal, or investment advice. QuadBlock Capital is a commercial real estate lender, not a tax advisor, CPA, or law firm. Depreciation, recapture, and exchange outcomes depend on facts specific to each taxpayer, entity structure, state of residence, and property, and the figures above rest on stated assumptions that will not match your deal. Consult a qualified CPA or tax attorney before acting on anything here.
Sources
- 26 U.S.C. section 1245, Gain from dispositions of certain depreciable property
- 26 U.S.C. section 1250, Gain from dispositions of certain depreciable realty
- 26 U.S.C. section 1031, Exchange of real property held for productive use or investment
- IRS Topic No. 409, Capital Gains and Losses
- IRS Publication 946, How To Depreciate Property
- IRS, Net Investment Income Tax
- KPMG TaxNewsFlash, Rev. Proc. 2025-32 inflation adjustments for 2026
- Federal Reserve, FOMC statement, September 16, 2026
- U.S. Treasury, Daily Treasury Par Yield Curve Rates, October 2026
- Federal Reserve, H.15 Selected Interest Rates
- CBRE, U.S. Cap Rate Survey H1 2026
