*587In these consolidated tax certiorari special proceedings, for tax years 1996-1997 through 2003-2004, the petitioner, Mutual of America Life Insurance Company (hereinafter referred to as Mutual), owner of a building at 320 Park Avenue, Manhattan, challenges the assessed valuation of the property by the Tax Commission and the Commissioner of Finance of the City of New York (hereinafter referred to as the City). On appeal, Mutual asserts, inter alia, that the trial court overvalued the property by disallowing an annual capital expenditure deduction for leasing costs on owner-occupied space. Because there is no legal authority directly on point, and neither party cites to any section of the tax code to support its arguments, this is a case of first impression on the issue of whether such leasing costs can be taken as a below-the-line deduction1 in the valuation of an investment property for tax purposes. We agree with the trial court that they may not be so deducted except where owner-occupied space becomes a de facto vacancy and the deductions are subject to proof.
The undisputed facts of this case are as follows: Mutual acquired the 34-story commercial office building at Park Avenue and 51st Street in 1992. Mutual refitted the building for use as its company headquarters in 1994-1995. The building offered approximately 675,000 square feet of rentable office space, as well as retail space on the ground floor and storage space in the below-grade levels. Of the rentable office space, Mutual occupied about 40%, that is 263,652 square feet (hereinafter referred to as owner-occupied space). The balance of rentable space was available to outside tenants. In the first tax year at issue, 1996-1997, about half of this space (30%) remained vacant. By the second tax year at issue, 1997-1998, approximately only 5% of the office space was still vacant. For the next three years the office space available to outside tenants was fully rented.
In 1996, Mutual sought an administrative correction of the building’s assessed value on which the property tax was based. Mutual failed to get a correction, paid the property tax and *588timely commenced a special proceeding against the City to challenge the assessment pursuant to RPTL 702, 706 (2) and New York City Charter § 163 (f); § 166. Mutual thereafter brought similar proceedings for the next seven tax years. Unable to reach a settlement, the eight cases were jointly tried in September 2006. At trial, each party submitted an expert’s written report containing eight appraisals. Based on its expert’s conclusions, Mutual withdrew its challenge to the last two tax years and thus the trial was limited to the first six tax years 1996-1997 through 2001-2002.
Both Mutual’s appraiser (Jerome Haimes) and the City’s appraiser (Terrence Tener) used the income capitalization approach to valuation, and both testified as experts at trial. Both experts valued the building as of the taxable status date of each year (January 5) pursuant to New York City Charter § 1507. Each applied a 45% equalization rate, appropriate for tax class four buildings in New York City, to produce a “fractional assessment” as required by state law pursuant to RPTL 305, 720 (3); 1802 (1).
The record reflects that the accepted income capitalization approach used by both experts results in an assessment of the building as an investment, with the building’s income stream representing the return on investment. In this case, both parties used the same method for determining income stream (hereinafter referred to as net operating income) as follows: by adding the actual annual gross rental income from all leased space (based on the rates and terms as reflected in lease contracts) to a hypothetical amount of gross rental income from any unleased space (based on the market rent at the time and estimated lease terms). The resultant annual gross potential income for each year was then reduced by deducting a vacancy and collection loss (calculated as an estimate based on market conditions and expressed as a percentage of the annual gross potential income). Also deducted were the annual operating expenses such as cost of utilities, maintenance, insurance, security and ongoing leasing costs. The latter costs were understood by both experts to be those annual costs budgeted for in anticipation of a future, steady turnover of tenants. Mutual characterized these — and the City did not disagree — as expenses which are “anticipated and amortized anticipating the large outlay to be made when a current tenant’s lease ends and a replacement tenant must be found.”
Further, both expert appraisers applied the same complex formula in calculating ongoing leasing costs to the total square footage of rentable space in the building, that is, to both leased *589and unleased space. The formula involves factoring in a lease renewal probability of 70% to reflect that not every tenant, would vacate at the end of a lease; the length of an average lease (Mutual’s expert used 13 years; the City’s appraiser used 10 years); tenant concessions expressed as. a dollar per square foot amount as well as leasing commissions expressed as a percentage of the annual gross income.
Tenant concessions sometimes characterized as “work letter items” were agreed to be improvements and .renovations made to prepare the space for a particular tenant as a result of a lease or tenant/landlord agreement. The concessions also included lost rent between leases and free rent offered as inducement. Leasing commissions were simply broker and cobroker commissions associated with renewals and new leases.
As a result, both expert appraisers arrived within the same range of expenses associated with ongoing leasing costs, and thus at similar values for the net operating income of the building.2 Both appraisers then utilized an estimated annual rate of return in order to arrive at the capitalized value of the building. For 1996-1997, Mutual’s appraiser used a rate of 14.7% while the City’s expert used a rate of 12.64% to reach capitalized values of $156,268,033 and $178,293,118 respectively. Finally, the assessed value could be estimated pursuant to RPTL 305, 720 (3) by applying an agreed-upon equalization rate, in this case 45%, to the capitalized value which is also generally considered to be the market value of the property.
Both expert appraisers agreed, however, that occasionally an intermediate step is required before establishing the market value of a property when a one-time, nonrecurring expense needs to be deducted as a below-the-line capital expenditure from the capitalized value. Testimony at trial established that this can be the cost of construction in rendering vacant raw space habitable by erecting drywall, installing ductwork and ceilings and floors. In situations where the space is already habitable, it is the.cost of tenant improvements and leasing commissions associated with an initial, occupancy (hereinafter referred to as lease-up costs) of habitable, retrofitted space. Both appraisers in this case agreed that annual recurring ongoing leasing costs do not account for the same expense as initial one-time leasing costs; that because initial costs are applied to space which does not involve a renewal but an entirely new tenant, a full and comprehensive leasing package is usually *590required; and, that a full leasing package on an initial occupancy would cost two to three times the expected annual rent for the space. Ongoing leasing costs were viewed as being lower than costs associated with preparing space for an initial occupancy because renewal tenants require significantly lower, or even no, work letter items such as painting or carpet replacement.
In this case, the City conceded that lease-up costs should be deducted from the capitalized value of the property for 1996-1997 for the 30% of rentable space that was vacant in the building in that tax year. The City’s appraiser allowed for almost $23 million as a lease-up cost, and so arrived at a market value of $155,309,341 for the building for 1996-1997.
Mutual’s appraiser however, applied lease-up costs to owner-occupied space as well as to the vacant space, noting that attribution of hypothetical rent should be accompanied by a deduction of appurtenant corresponding costs which would not be covered by the ongoing operational expenses. Thus, at this point the valuations of the expert appraisers diverged dramatically since Mutual further argued that the lease-up costs for the owner-occupied space should be deducted for each of the disputed years. For tax year 1996-1997, Mutual’s appraiser deducted $59,698,490 as a capital expenditure item to allow for leasing costs associated with both the 30% vacancy and the 40% of owner-occupied space and so arrived at a market value of $96,569,543.
The trial court engaged in a thorough comparison and analysis of both expert appraisals for each of the tax years in dispute. For clearly stated reasons, the court selected either Mutual’s or the City’s value for each step of the appraisal — except that, sua sponte, it used market, not actual, rent to calculate the gross income from leased space as well as unleased space. Thus, the court reached its own results for the capitalized value of the building in successive years.
The court further determined that the capitalized value was also the market value of the subject building for all the tax years in dispute except for the first tax year of 1996-1997. This determination took into account the concession of the City’s appraiser that consideration must be given for lease-up costs on the 30% vacant portion of the building. The court adopted the City’s estimate and deducted a rounded-up amount of $23 million from the capitalized value during the first tax year at issue.
Finally, the court noted that the “considerable” disparities between the market values found by the court and those opined by Mutual’s expert are due to the incorrect opinion of Mutual that the “costs of preparing owner-occupied space for a market *591tenant in each disputed year should be deducted from the capitalized value of the net operating income.” The court rejected that opinion, and thus reduced the assessed valuation only for the 1996-1997 tax year. For the reasons set forth below, we agree with the trial court, and modify only to the extent of finding that it erred in applying market rent to leased space when actual rents were available for the calculation of gross income.
It is a cardinal principle, enshrined in the State Constitution, that in property valuations for tax purposes “[assessments shall in no case exceed full value.” (NY Const art XVI, § 2; see Matter of Commerce Holding Corp. v Board of Assessors of Town of Babylon, 88 NY2d 724, 729 [1996].) Further, the Court of Appeals has held that the “concept of full value is typically equated with market value” or “what a seller under no compulsion to sell and a buyer under no compulsion to buy” would agree is the most probable price a property would bring on a specific date (88 NY2d at 729). Thus, the Court of Appeals observed that “the assessment of property value for tax purposes must take into account any factor affecting a property’s marketability” {id., citing RPTL 302 [1] [“(t)he taxable status of real property . . . shall be determined annually according to its condition”]). Moreover, each annual assessment is separate and distinct from each other (Matter of Northville Indus. Corp. v Board of Assessors of Town of Riverhead, 143 AD2d 135, 138 [1988]).
On this basis, Mutual argues that valuing real property according to its condition on each taxable status date means that an appraiser must hypothesize a sale of the property on that date each year. Indeed, Mutual argues that “[w]here six annual valuations are in dispute, as here, appraisers must imagine six separate sales . . . [even though] each sale is a fiction.”
Further, relying on Commerce Holding Corp., Mutual argues that for any factor that depresses the value of the property, the cost to cure must be deducted. Moreover, if the improvement has not been made, it must also be deducted the following year. Mutual asserts this does not mean that the amount is spent “again and again” — just that it must be accounted for hypothetically each year.
In this case, Mutual contends the owner-occupied space comprising 40% of the building’s rentable space is a negative factor affecting the property’s marketability. Upon closing, it argues, a “willing” buyer would be facing a 40% vacancy, and hence would be confronted with initial lease-up costs for 40% of the building which, like those associated with the 30% vacancy in 1996-1997, must be deducted as a capital expenditure.
*592Mutual further relies on Matter of CCB Assoc. v Penale (266 AD2d 805 [1999], lv dismissed in part and denied in part 95 NY2d 788 [2000]) to argue that since market rent is attributed to the 263,652 square feet of owner-occupied space not earning rent, the cost necessary to obtain that rent must be accounted for because such potential costs reduce the market value of a property. In this case, Mutual argues that the trial court failed to recognize the hypothetical nature of the costs in the context of a hypothetical sale for valuation.
Mutual’s argument is based on a flawed analysis of the sparse applicable case law. First, its reliance on Commerce Holding Corp. is misplaced. In that case, subsurface contamination indisputably affected marketability of the realty, and the Court held that the full cost to complete the cleanup had to be deducted, from each valuation. In other words, the “total remaining cost” was to be deducted not just the amount expended in a particular year. However, in Commerce Holding Corp., the building suffered from de facto contamination, and the costs associated with the cleanup were subject to proof. As the cleanup progressed, the total of the remaining costs was a real, not a hypothetical, amount just as the 30% vacancy of the subject building in 1996 was a de facto vacancy not a hypothetical.
Even if we were to accept, as Mutual posits that, “the construct of hypothesized sales ordinarily presents no unusual dilemma” in valuations, the 40% vacancy facing a new hypothetical buyer in each successive year is an unacceptable construct. That proposition would entail assuming that each new buyer becomes an owner-occupier of the same 40% of office space since, according to Mutual a hypothetical buyer in year two is also confronted with a 40% vacancy rate as is hypothetical buyer in year three and so on for each of the years at issue. It cannot even be assumed that a hypothetical buyer would be confronted with a 40% vacancy upon closing. It could be equally well hypothesized that an owner-occupier while selling the property would still remain as a tenant of some portion if not all of the current space. Similarly, the hypothetical new buyer could occupy some of the space and would be looking to lease only a portion of the current 40% of owner-occupied space. Indeed, it is obvious that a vacancy to which lease-up costs are properly attributed as a below-the-line deduction cannot be hypothetical. The proposition that a new owner would be faced with a vacancy of 40% of the space for each disputed year simply cannot be assumed where such a hypothetical does not require actual accrual of costs. It merely results in a tax windfall for the petitioner.
*593Indeed, if there is one holding to be extrapolated from the sparse case law cited by Mutual, it is that lease-up costs qualify as a capital expenditure only when the vacancy actually exists (Matter of CCB Assoc., 266 AD2d at 807). In CCB Associates, the petitioners presented evidence that the building’s major tenant had vacated the premises the year prior to their petition, leaving a 58% vacancy rate. More significantly, the case suggests that the vacancy must be of sufficient size to destabilize occupancy.
CCB Associates supports what we assume is the City’s view, that unless the area of vacant space is sufficiently sizeable the leasing costs associated with retenanting it must fall within the normal, ongoing operating expenses. At trial, the City conceded that the costs associated with tenanting 30% of the rentable office space that remained vacant in 1996 should be viewed as below-the-line capital expenditure. The City, relying on the Appraisal Institute’s Appraisal of Real Estate (12th ed 2001), agreed that “[a]n investment grade building is not completed till it has stabilized at market occupancy.” In this regard, because in 1996, the building was only 70.5% tenanted, the City agreed it was not at a stabilized occupancy, which it did not achieve until the following year.
However, there is no authority, in tax code, statute or case law, nor are we inclined to set a precedent, for classifying space which is, in effect, occupied (albeit by the owner, not a paying tenant) as vacant based on the fiction that it will be leased to a paying tenant at the start of each new tax year. As the City asserts, relying on Matter of Ernst v Board of Assessors of City of Lockport (58 Misc 2d 504 [1968], affd 33 AD2d 655 [1969]), while owner-occupied space is calculated as if leased at market rent to produce a hypothetical revenue stream, the fact that it does not do so is entirely by Mutual’s own choice, and Mutual “cannot expect [its] fellow taxpayers to compensate [it] for the difference.”
The trial court also correctly determined that lease-up costs could not be taken as a capital expenditure for the next two tax years in dispute even though some of the 30% of vacant space remained untenanted. It is true that the Court of Appeals has held that the total remaining costs of curing any negative factor must be applied each year irrespective of whether they are actually spent (see Matter of Commerce Holding, 88 NY2d at 731). It is also undisputed that, in this case, less than 5% of the building’s rentable space was untenanted as of the beginning of the second tax year in dispute. However, in this case, both appraisers factored in a 5% vacancy and collection loss as part of *594the ongoing operating expense. Hence, both parties essentially agreed that occupancy was stabilized at 95%, and thus Mutual cannot claim the remaining vacancy as a below-the-line capital expenditure.
Finally, we agree with Mutual that the trial court erred in using market rent rather than actual rents for leased space in its calculations. “As a rule, actual income is the best indicator of value.” (Matter of Conifer Baldwinsville Assoc. v Town of Van Buren, 115 AD2d 325, 325 [1985], affd 68 NY2d 783 [1986]; see also Matter of City of New York [First Elephant Estates — La Hermosa Church], 17 AD2d 317, 320 [1962] [“(generally, with respect to income-earning property . . . the net income is . . . the surest index of value”].) In essence, market rent is the rent at which a space, under current and ordinary conditions, would command on an open market. As a result, actual rent for tenant occupied space will always be a more accurate barometer of the subject property value than market rent. In this case, both appraisers, in applying the income capitalization approach, used actual rents, not market rents, for space that was actually leased (tenant occupied space), and applied market rent only to owner-occupied space and vacant space. Consequently, we find that the trial court should have used actual rents, where they are available.
Likewise, as Mutual asserts, and the City concedes, the trial court erred in adding escalation to the market rents. Escalations are increases in a tenant’s rent, typically stipulated to in a commercial lease, that compensate the owner for general inflation or specific expense increases. It follows that escalation should only be applied to actual rent and not market rent. Since the trial court only applied market rent in its calculations of operating income, we conclude that it erred in its application of escalation and thus erroneously arrived at net operating income totals for each of the tax years greater than the totals submitted by either of the parties’ experts. Concur — Mazzarelli, J.P., Andrias, Nardelli and Catterson, JJ.
. Generally, one-time, nonrecurring expenses that are not ongoing annual expenses may be taken as a capital expenditure, and thus as a below-the-line deduction that reduces the market value of a property for tax purposes.
. For example, for the 1996-1997 tax year, Mutual’s appraiser estimated the net operating income at $22,971,411. while the City’s appraiser estimated it at $22,536,259.