BELLACOSA, J.

This tax certiorari case involves the real property assessment of the Saratoga Harness Track in the City of Saratoga Springs. The principal question is whether the comparable lease income method of valuation utilized by the taxpayer's expert is an impermissible method of valuation for owner-occupied property, as the Appellate Division found. In conjunction with the consideration of that issue, the courts below also reviewed whether the subject property constitutes a "specialty."

Saratoga Harness owns 161.3 acres of land in the City of Saratoga Springs, on which it operates a racetrack. Improvements include a one-half mile oval track, a grandstand, a clubhouse, barns, stables, veterinary facilities, and administrative and maintenance buildings. The City's assessments were predicated on a full property value of close to $19 million dollars. Its Assessor found the property to be a "specialty," and calculated its value using the replacement cost less depreciation method. This approach is based on "[t]he current cost of reproducing * * * the improvements, minus the loss in value from depreciation, plus site value" (Appraisal Inst., The Appraisal of Real Estate 81 [11th ed]).

Saratoga Harness filed written protests with the City's Board of Assessment Review prior to the finalization of the assessment roll for each year. The protests were rejected and separate proceedings were initiated to annul and correct the 1993 and 1994 assessments. Supreme Court conducted a bench trial and granted partial relief to the taxpayer, by reducing the assessments for the respective years to $519,300 and $516,600, from $1,085,000. The Appellate Division reversed and dismissed the proceedings. This Court granted leave to appeal to the taxpayer, and we now reverse and remit to the Appellate Division.

At trial, the taxpayer's expert testified to respective $4.3 and $3.8 million values for 1993 and 1994, based upon the comparable lease income method of valuation. The taxpayer also offered into evidence the expert's appraisal report supplementing and documenting the testimony. At the close of the evidence, the City rested without presenting any counter proof or theory of value. It relied on the presumption of validity that attaches to real property tax assessments, and its belief that Saratoga Harness failed to meet its burden to prove overvaluation.

Supreme Court held that the subject property is not a "specialty" because it was "established as fact, uncontroverted, that there is a market for racetrack property and that there are sales of racetrack properties for such use." The court then rejected the City's assertion of the replacement cost method to value the subject property. It found that the comparable lease income method employed by the taxpayer's expert sufficiently rebutted the presumption of correctness. The court then also disallowed the value tendered by the taxpayer's expert as too low. Instead, it set the value for both 1993 and 1994 at $9 million — the amount that Saratoga Harness paid for the property in a stock transfer transaction in 1987, an issue and rationale not reached by the Appellate Division because of the nature of its decision.

In reversing and dismissing the petitions outright, the Appellate Division agreed with the City's Assessor that the subject property is a "specialty" ( 241 A.D.2d 655, 657). The court disagreed with the taxpayer's use of the comparable lease formula and concluded that this methodology was improperly based on key facts not in evidence: "first, that there was a rental lease agreement between petitioner and a tenant-operator of the subject property and, second, that rental money was paid from the tenant to petitioner" ( id., at 658).

This appeal turns on whether the comparable lease income method of valuation is proper in these particular circumstances. As this Court recognized in Matter of Allied Corp v Town of Camillus ( 80 N.Y.2d 351), "while property must be assessed at market value, there is no fixed method for determining that value" ( id., at 356). "The ultimate purpose of valuation, whether in eminent domain or tax certiorari proceedings, is to arrive at a fair and realistic value of the property involved so that all property owners contribute equitably to the public fisc" ( id., at 356). The Court explained that "[a]ny fair and nondiscriminatory method that will achieve that result is acceptable" ( id., at 356).

With regard to the various methods for valuation, the Court stated that "[t]he best evidence of value, of course, is a recent sale of the subject property between a seller under no compulsion to sell and a buyer under no compulsion to buy" ( id., at 356). "Absent that evidence, however, the courts have traditionally valued property by one of three methods: comparable sales, capitalization of income or reproduction cost less depreciation" ( id., at 356). The Court stated that "[e]vidence of comparable sales is generally the preferred measure of a property's value for assessment, but where there is insufficient relevant data, value may be determined by other methods" ( id., at 356).

Importantly, the Court has emphasized that "even when alternative theories must be used, the courts have been cautious about applying the reproduction cost less depreciation method [the method utilized here by the City of Saratoga Springs and imposed by the reversal at the Appellate Division] because it is most likely to result in overvaluation" ( id., at 356-357). The rationale for that careful limitation on the reproduction formula is that it "ascribe[s] too little weight to such factors as rising construction costs and diminishing value by functional obsolescence" ( id., at 357). Generally, then, the Court has eschewed the use of that method except for properties qualifying as "specialties" ( id., at 357).

The comparable lease income method proffered at trial by Saratoga Harness, on the other hand, is generally used as a component of the income capitalization approach to real property valuation (i.e., "[t]he value of a property's earning power based on the capitalization of its income") (Appraisal Inst., The Appraisal of Real Estate, at 81 [11th ed]). The technique generally involves two-steps: (1) the appraiser determines the "market rent" of the property in question, and (2) capitalizes (or converts to present value) that rent to determine the proper value of the subject property ( see, e.g., Appraisal Inst., The Appraisal of Real Estate, supra, at 449).

This methodology is accepted as valid within the field of real property appraisal ( see, Appraisal Inst., The Appraisal of Real Estate, at 471-489 [11th ed]; see also, 1 Bonbright, The Valuation of Property, at 169, 216-232). In addition, this Court implicitly recognized and countenanced this formula in W.T. Grant Co. v Srogi ( 52 N.Y.2d 496, 508, 512 [upholding an appraisal of leased premises based on "the rent which a national chain store would agree to pay on a percentage lease"]).

Most relevantly here, when this method is utilized for owner-occupied space, it "is usually estimated at market rent levels" (Appraisal Inst., The Appraisal of Real Estate, at 489 11th ed]). The rubric is that "market rent is the rental income that a property would most probably command in the open market" (Appraisal Inst., The Appraisal of Real Estate, at 478-479 [11th ed] [emphasis omitted]). Estimated annual rent, or market rent, should be determined by analysis of the rent paid for other comparable properties and then imputed to the subject property:

"The rents of comparable properties can provide a basis for estimating market rent for a subject property once they have been reduced to the same unit basis applied to the subject property. Comparable rents may be adjusted just as the transaction prices of comparable properties are adjusted in the sales comparison approach" ( Appraisal Inst., The Appraisal of Real Estate, at 481 11th ed]).

In the instant case, the taxpayer's expert arrived at its proferred value by (1) gathering information on 21 comparable horse racetracks subject to lease agreements with tenant-operators, (2) analyzing the lease agreement for each comparable racetrack, (3) determining the rent paid by each tenant-operator as a percentage of income derived from the operations of each racetrack, (4) adjusting the rent paid by each tenant-operator to account for any differences between the business and regulatory conditions under which the racetracks operated, (5) computing an imputed market rent for the subject property, and (6) capitalizing the imputed market rent.

We are satisfied that the comparable lease income method is a valid and acceptable regimen to value this owner-occupied property. The Appellate Division mistakenly rejected this mode by effectively restricting it to properties actually leased. The use of "market rent" in the formula for this method, by definition, is not necessarily contingent on the existence of a lease of the subject property; rather, it more realistically assumes either the non-existence of a lease or the inaccuracy of an actual lease in reflecting true market value.

Since we disagree with the Appellate Division and are persuaded that the comparable lease income method is a proper method of valuation for owner-occupied property, we must also consider whether Supreme Court properly permitted its utilization here, while rejecting the City's use of the replacement cost less depreciation method. A key factor in resolving the inter-related puzzle is whether the subject property may be treated as a "specialty." We are convinced that Supreme Court correctly found that the property does not constitute a "specialty."

This Court has limited the use of the reproduction cost method to properties considered "specialties" ( Matter of Allied Corp v Town of Camillus, 80 N.Y.2d 351, 357, supra). Allied Corp. defined "specialty" properties as those "that are ` uniquely adapted to the business conducted upon [them] or use made of [them] and cannot be converted to other uses without the expenditure of substantial sums of money'" ( Matter of Allied Corp v Town of Camillus, supra, 80 N.Y.2d, at 357, quoting Matter of Great Atl. Pac. Tea Co. v Kiernan, 42 N.Y.2d 236, 240 [emphasis in original]). Four criteria have emerged for determining whether a property qualifies as a "specialty":

"(a) the improvement must be unique and must be specially built for the specific purpose for which it is designed; (b) there must be a special use for which the improvement is designed and the improvement must be so specially used; (c) there must be no market for the type of property and no sales of property for such use; and (d) the improvement must be an appropriate improvement at the time of the taking or assessment and its use must be economically feasible and reasonably expected to be replaced" ( supra, 80 N.Y.2d, at 357 [emphasis added]).

With regard to the market and sales component of the test, Supreme Court correctly concluded on the evidence submitted that both aspects were satisfied. The trial court weighed the "uncontroverted" evidence that there were 33 sales of horse racetracks in the United States and one in Canada between 1984 and 1992. Indeed, three of the sales were in New York State, namely Roosevelt Raceway, Vernon Downs and the subject Saratoga Raceway itself (the subject property was transferred to its present owner in 1987 for $9 million). We agree with the trial court's realistic assessment and application of Allied Corp. and, correspondingly, we reject the Appellate Division's conclusory "specialty" slotting of this property. Saratoga Harness established the existence of a "market" for racetrack properties by admissible evidence sufficient to take it out of the "specialty" niche.

As noted, the reproduction cost less depreciation formula (the method used by the City Tax Assessor here) is the one most likely to result in overvaluation and, thus, its use is generally limited to properties deemed "specialties" ( Matter of Allied Corp v Town of Camillus, supra, 80 N.Y.2d, at 357). Because this property is not a "specialty," as the trial court correctly found, the City and the Appellate Division relied on an inappropriate or unacceptable methodology premised on that classification. We are persuaded, instead, that the comparable lease income method, proffered by the taxpayer at trial, was proper.

That conclusion, however, does not tie up all the loose ends of this case because the Appellate Division, in substituting an erroneous methodology for a correct one, understandably never turned to the required review of the valuation adjudication made by the trial court. For that reason, we must consider the most appropriate remedial action, upon our law reversal of the Appellate Division rationale and order. We perceive no matter-of-law basis to reinstate the Supreme Court judgment outright; instead, we conclude that a remittal to the Appellate Division is prudent and required.

W.T. Grant Co. v Srogi ( 52 N.Y.2d 496, supra) discussed the power of this Court with regard to valuations in this way:

The determination of market value essentially is a question of fact. As such, the Appellate Division is empowered to make new findings of value where the trial court "`has failed to give conflicting evidence the relative weight which it should have.'" Furthermore, where the Appellate Division reverses the findings of value made by the trial court and makes new findings of fact, the Court of Appeals is empowered to choose between the findings of the trial court and the new findings of the Appellate Division by ascertaining which valuation is in accord with the weight of the evidence ( id., at 510-511 [citations omitted] [emphasis added]).

Because the Appellate Division neither reviewed the trial court's findings of value, nor made findings of its own, this case is most appropriately remitted to that court to complete its intermediate appellate court review. We should not render a final, matter-of-law resolution, as requested and urged by appellant Saratoga Harness.

Accordingly, the order of the Appellate Division should be reversed, with costs, and the matter should be remitted to the Appellate Division for consideration of issues raised, but not determined on the appeal to that court.

Order reversed, with costs, and matter remitted to the Appellate Division, Third Department, for consideration of issues raised but not determined on the appeal to that court. Opinion by Judge Bellacosa. Chief Judge Kaye and Judges Titone, Smith, Levine, Ciparick and Wesley concur.

Decided June 9, 1998