In these consolidated matters plaintiff, Ford Motor Company (Ford), seeks a reduction in its local property tax assessments for the tax years of 1983, 1984 and 1985. Defendant, Edison *156Township, feeling aggrieved by the assessed valuation for each of the tax years seeks an increase in assessment.
The property involved in this proceeding is known and designated as Block 198L, Lot 37A. 1 on the tax map of Edison Township and is commonly identified as 939 U.S. Route 1. It was assessed for the tax years in question as follows:
1983
Original Assessment Middlesex County Board Judgment
Land $ 3,827,000 Land $ 3,827,000
Improvements 17,130,600 Improvements 24,595,000
Total $20,957,600 Total $28,422,000'
1984 and 1985
Original Assessments
Land $ 3,827,000
Improvements 24,595,000
Total $28,422,000
The 1983 assessment, which was a product of a complete revaluation in Edison, was increased by the Middlesex County Board of Taxation at the behest of Edison. Ford sought review of that determination in this court while Edison filed a counterclaim seeking an even greater increase. In 1984 and 1985 Ford sought direct review of its assessments in this court pursuant to N.J.S.A. 54:3-21 while Edison filed counterclaims seeking an increase in those assessments.
Ford contends that the fair market value of the subject property as of the assessment date of October 1, 1982 for the tax year of 1983 was $14,500,000, $15,000,000 as of the assessment date of October 1, 1983 for the tax year of 1984 and $16,000,000 as of the assessment date of October 1,1984 for the tax year 1985. It was further Ford’s position that, with the application of the appropriate ratios, the fair assessable value of its property was $14,500,000 for 1983, $14,580,000 for 1984 and $15,128,000 for 1985.
*157Edison contends that the fair market value, as of the respective assessment dates, was $30,000,000 for 1983, $33,000,000 for 1984 and $36,000,000 for 1985. Edison also claims that these values should be utilized as the assessments for the respective tax years.
At the outset of the proceeding, the parties stipulated that in 1983 Edison implemented a district-wide revaluation, and thus, the applicable assessment ratio was 100%. For the tax years of 1984 and 1985 the parties agreed that the appropriate assessment ratios would be in accordance with the ratios set forth by the Director of the Division of Taxation pursuant to chapter 123, N.J.S.A. 54:l-35a, :51A-6. These were 97.20% for 1984 and 94.55% for 1985. The parties also agreed that the total amount of floor space for all improvements for purposes of the present litigation was 1,350,162 square feet.
After the initial trial of this matter was completed, our Legislature enacted L. 1986, c. 117 which amended N.J.S.A. 54:4-1 and N.J.S.A. 54:llA-2 by establishing new criteria for determining whether property is taxable as real property. As a result, Edison moved to reopen proofs and for additional discovery. According to Edison, the Ford complex contains a number of items and systems which its appraisers did not consider as real property at the original hearing but as a consequence of c. 117 constitute assessable real property which must be considered in making any local property tax assessment determination. See Chevron U.S.A., Inc. v. Perth Amboy, 9 N.J.Tax 205 (Tax Ct.1987).
Edison’s motion to reopen this case was granted and additional discovery is still proceeding. In the interim, counsel for plaintiff, Ford Motor Company, has requested that this court render its valuation determination with regard to the proofs that were adduced at the original trial, on the basis that such a decision would be helpful to the parties, and perhaps, even be conducive to settlement negotiations, and thus obviate what is anticipated to be a protracted trial on the issue of whether the *158items to which Edison now makes reference pursuant to c. 117 should be deemed, and assessed as, realty.
Since this court would have to come to a value conclusion with regard to the proofs offered at the initial hearing, in any event, before any consideration is given to the additional items asserted by Edison to be real property, it is felt that Ford’s request is well founded. Therefore, I make the following findings and conclusions relative to the initial trial of the present matter.
Description.
The subject of this proceeding consists of an automobile assembly complex with manufacturing, warehousing and office areas commonly known as the “Edison Assembly Plant” and is used by a part of the “Body and Assembly Division of Ford Motor Company.” This facility produces, or attempts to produce, approximately 580 cars a day. The improvements are roughly centered on an irregularly shaped parcel of land comprising 76.54 acres. The site has approximately 2900 feet of frontage on the northwest side of U.S. Route 1, slightly south of the U.S. Route 1 interchange at Interstate 287, in Edison.
The property, which is in easy reach of New York and Philadelphia, enjoys an excellent road network provided by Interstate Highways 95, 287, 78, U.S. Routes 1, 9, 130, 22, 202 and the Garden State Parkway. There are also rail facilities at the site. The area north of the subject is generally industrial, while the uses south of the property are generally commercial. To the rear of the subject there is generally a single-family residential development and some small commercial uses.
The improvements, originally constructed in 1947 and 1948, with additions in 1954, 1957-1962, 1968-1972, 1977 and 1980, consist of three major buildings, i.e., a main assembly building (which has a mezzanine containing a cafeteria, hospital or medical clinic, conference rooms and offices), a warehouse and a two-story administration building, and six outbuildings (basically metal single-story structures) which, in total, as stipulated *159by the parties, comprise 1,350,162 square feet of floor space. The manufacturing and warehouse improvements have clear ceiling heights ranging from 16 feet to 30 feet and bay sizes ranging from 40 feet by 50 feet to 40 feet by 70 feet. There are 16 loading docks or platforms for the receipt of materials.1 The weighted average physical age of the improvements was considered to be approximately 25 years. The site improvements consist mainly of a paved parking lot which provides parking for approximately 1,600 automobiles, interior roadways, four railroad spurs comprising approximately 3,000 lineal feet of rail (all four rail lines enter the plant), cyclone fencing, water and fuel tanks, a water tower and exterior lighting.
The subject is zoned L-I (light-industrial) pursuant to the zoning ordinance of Edison and conforms in all respects with the exception of one 80-foot high stack on the boiler house which is nonconforming due to its height. Since the maximum percentage of lot coverage by building permitted by the ordinance is 50% and the land to building ratio of the subject is 2.5 to 1 there is room available for some expansion, if necessary, at the site.
The property is serviced by all public utilities including water, sewer, electric and gas. In this regard the Ford facility has features that are somewhat unique to automobile assembly as compared to general purpose manufacturing structures. Specifically, the Ford plant enjoys electric service of 26.4 KV power which, according to the report of Carl F. Walter, III, Ford’s architectural witness:
... is divided into two main distribution circuits by two transformers each with 6000/8000 KVA capacity. Power is transformed from 26.4KV to 13.2 KV30 *160and then distributed to eight substations each containing a 2000 KVA, 1500 KVA or 1000 KVA transformer.
As noted by Dr. Kevin McDermott, defendant’s industrial engineering expert (who is also an electrical engineer) there are “literally miles of [electrical] cable going through the overheads____” Additionally, according to Walter’s report, steam is provided by a boiler house located on the premises which:
... contains three ... boilers each with a rated capacity of 60,000 pounds/hour steam pressure.
Steam is distributed throughout all areas of the manufacturing and office buildings via an 8 inch line from each boiler which feeds a 12 inch diameter main header.
There are two waste treatment systems, one designed to facilitate sanitary waste removal and a second to provide for removal of industrial waste. There are also air compressors that provide compressed air through “miles” of air lines throughout the facility to power nonelectric power tools and equipment. Additionally, there are a “couple thousand feet” of water lines, separate and apart from the normal plumbing system, which carry cooling water to cool such items as welding transformers and heat exchangers. The report of Walter indicates that the capacity of each service is sufficient to provide for the automobile assembly use but “for the most part [is] in excess of that required for a multitude of other single or multi-tenant operations.” There is also a paint-spraying service which involves a system of piping designed to transport paint from a utility building into the main assembly building.
The condition of the property in general has been denoted as fair by plaintiff’s appraisal expert and very good by defendant’s appraisal expert. The condition of a large portion of the roof (631,118 square feet) was characterized as poor by plaintiff’s architectural witness, Walter, based on his visual inspection, while defendant’s appraisal expert was of the opinion that the roof was well maintained and in good condition. It is anticipated that the existing use, automobile assembly, will continue for the foreseeable future. In spite of this, however, the parties were not in agreement with regard to the highest and best use of the subject property.
*161 Highest and Best Use.
The concept of highest and best use is not only fundamental to valuation but is a crucial determinant of market value. American Institute of Real Estate Appraisers, The Appraisal of Real Estate (9 ed. 1987) at 41-42, 68, 269-270, 292. This is why it is the first and most important step in the valuation process. Id. at 41-42, 68. The highest and best use for the subject property is that use which at the time of the appraisal (here, the appropriate assessment dates) is the most profitable likely use or produces the highest property value. Inmar Assocs., Inc. v. Edison Tp., 2 N.J.Tax 59, 64-65 (Tax Ct.1980). It has been defined as the “reasonably probable and legal use of ... property ... that results in the highest value.” The Appraisal of Real Estate, supra at 269.
The Appraisal of Real Estate indicates there are two reasons which necessitate a highest and best use analysis. Id. at 68. One reason is for the purpose of determining the most profitable use of the property while a second is to assist in “identifying comparable properties” in a direct-sales comparison approach to value. Id. at 68, 273-274.
Four criteria are implicated in any highest and best use analysis. In order for a particular use to be the highest and best it must be: 1) legally permitted, 2) physically possible, 3) economically feasible, and 4) the most profitable. Id. at 274. Plaintiff insists that, considering these criteria, the subject’s highest and best use is for general manufacturing purposes with associated warehouse and office for a single user. Defendant, however, maintains that the highest and best use for the subject would be a continuation of its current use, i.e., an automobile assembly plant or like use such as small airplane assembly, truck or farm tractor assembly.2
*162There is no dispute in this case with regard to the first two criteria of highest and best use. Either of the uses advanced by the parties is both legally permitted and physically possible. Contention arises, however, with regard to whether the existing use is economically feasible and the most profitable. Plaintiffs appraisal expert, Jeffrey R. Parkhouse, was of the opinion that the present use lacked economic feasibility because economic demand for major industrial facilities, in general, has been particularly weak in the Northeast due to increased competition from sunbelt states and foreign manufacturers.
After a review and analysis of large manufacturing industries, business climate studies, and labor and work-force statistics he concluded that there was a limited market demand for large industrial facilities, and although the subject would have to compete with a number of other large industrial facilities currently available in the market, it was his opinion that the highest and best use for the subject was as a general purpose manufacturing facility. It is interesting to note that Parkhouse did not consider a sale of the subject for use as an automobile assembly plant nor did he explain why he did not.3 (Presumably it was because he felt that if, in fact, the plant were offered for sale there would be no buyers with requirements similar to that of the seller, i.e., an automobile or like manufacturer, but the record does not persuasively support this assumption.)
Plaintiff argues that the opinion of its expert is correct because there would be no demand in the market for an automobile manufacturing plant as such. Plaintiff also asserts *163that Edison did not establish any demand for an automobile manufacturing facility and therefore its current use cannot be its highest and best.
Ford also argues that in spite of the fact that its property has some features which render it suitable to its present operation these features do not impair the property’s adaptability for other uses. In this regard Ford maintains that those features such as the extensive electrical capacity, plumbing, steam capacity, air compressor system, waste treatment systems, air handling system, railroad spurs and trackage, loading docks and platforms might be excessive for most general manufacturing users, nevertheless, it would make the improvements more attractive to a wider range of users looking for a large general purpose manufacturing facility. Essentially, Ford asserts, the fact that the improvements or its specialized features would be “underutilized” is irrelevant in any highest and best use analysis.
In contrast, Edison contends that the fact that the Ford plant remains in its present use, and since there is no evidence that there are any plans by Ford to discontinue the present use, demonstrates that there is a demand for the subject as an automobile assembly facility. Additionally, although denying that the sales are comparable for other reasons, Edison asserts that two of Parkhouse’s comparable sales (Parkhouse’s comparable sales one [Volkswagen of America, Inc. to Chrysler Corporation] and four [Ford Motor Company to Northrop Corp.]) illustrate that large assembly plants “are sold for continued use as heavy industrial assembly plants,” and thus, demonstrate a demand for facilities which can accommodate this use. The Volkswagen sale shows demand for continued use of automobile assembly plants and the Northrop sale shows demand for an allied or like use of aircraft assembly.
Edison further maintains that any use characterization that ignores the many special features of the subject is “short-sighted.” It concedes that there would be potential purchasers who would be interested in the property for general industrial *164purposes but these potential purchasers would, and could, not take full advantage of the amenities offered by the property, and thus, would not be interested in paying for these items, but an automobile manufacturer or kindred user would.
Ford rests its highest and best use argument in large measure upon this court’s decision in WCI-Westinghouse v. Edison Tp., 7 N.J.Tax 610, 616-617 (Tax Ct.1985), aff’d o.b. per curiam 9 N.J.Tax 86 (App.Div.1986). In WCI-Westinghouse plaintiff maintained that a 943,725-square foot manufacturing-warehouse-office complex that had produced a wide variety of products at various times had a highest and best use as a general purpose manufacturing facility. Defendant argued that the improvements were special purpose in nature and as such plaintiff’s specific use at the time of the assessment was the highest and best use. This court observed:
In order to constitute a particular use as the property’s highest and best use, the selected use in a value in exchange context must be a probable use for which there must be a demand in the relevant market. [7 N.J.Tax at 616]
It was determined that there was nothing in the record which would demonstrate a demand for plaintiff’s precise use in the market generally. Moreover, because of the wide diversity of manufacturing uses for which the subject had been utilized and was readily adaptable, this court concluded that that facility’s highest and best use was as a general purpose manufacturing facility.
Edison argues that WCI-Westinghouse is inapposite. I agree. The record in this case reveals that there is a relevant market for automobile assembly plants either in continued use or in a kindred use. See Sunshine Biscuits, Inc. v. Sayreville, 4 N.J.Tax 486, 496 (Tax Ct.1982) (the market for a particular parcel of industrial real estate consists of the potential purchasers, unlimited by a specific geographical location). As previously mentioned the Volkswagen to Chrysler Corporation and Ford Motor Company to Northrop sales are illustrative of demand in the relevant market.
Additionally, as Judge Lario noted in Owens-Illinois Glass Co. v. Bridgeton, 8 N.J.Tax 495 (Tax Ct.1986), it is still *165the general rule in local property taxation that property must be valued “in the actual condition in which the owner holds it.” Id. at 501-502 (quoting from Colwell v. Abbott, 42 N.J.L. 111, 115 (Sup.Ct.1880)); see also Trustees of Stevens Inst. of Technology v. State Bd., 105 N.J.L. 99, 101, 143 A. 356 (Sup.Ct.1928), aff'd 105 N.J.L. 655, 146 A. 919 (E. & A.1929) (property should be valued in the actual condition in which the owner holds it).
In any highest and best use analysis all the capabilities of the property and all the uses to which it may be applied, or for which it is adapted, are to be considered and examined and that use which yields the highest value should be selected. Inmar Assocs, Inc. v. Edison Tp., supra, 2 N.J.Tax at 64. Any examination of potential uses cannot overlook the reality of the use for which the plant was designed, as an automobile assembly plant, and the actual condition in which Ford has held this property for the past 35 years, and will hold it indefinitely in the future, again, as an automobile assembly plant in daily operation. See Owens-Illinois Glass Co. v. Bridgeton, supra at 502-503; see also Hackensack Water Co. v. Bor. of Old Tappan, 77 N.J. 208, 214, 390 A.2d 122 (1978) (since property valuation must “have some relationship to reality,” it would be proper to consider actual use as the highest and best use) and McGinley Mills, Inc. v. Phillipsburg, 9 N.J.Tax 508, 514-515 (Tax Ct.1988) (highest and best use of an industrial plant designed, built and used as a textile manufacturing plant was its current use as a textile manufacturing plant).
In this context, it is interesting to note that it has long been accepted in eminent domain cases that if the property were being employed, at the time of the taking, in the use which is asserted to be the highest and best, that proof of actual use satisfies the requirement of showing the existence of a demand for that use.4 See Housing Auth. of the City of *166Bridgeport v. Lustig, 139 Conn. 73, 90 A.2d 169 (Conn.Sup.Ct. 1952) (“In particular, it is proper to take into consideration the existence of a going business on the land in question as indicative of the highest economic use to which the land may be put.” Id. at 171).5 Compare CPC Int’l v. Bor. of Englewood Cliffs, 193 N.J.Super. 261, 268, 473 A.2d 548 (App.Div.1984) (the sale contemplated as the criterion of market value is a “hypothetical sale” in which the hypothetical buyer is one whose requirements are reasonably accommodated by the property in question).
Defendant asserts that if this court does not consider the undisputed fact that the subject plant “will continue to be devoted to automotive assembly use” the court will “ignore the value of all those features that are presently being utilized by Ford” as if those “improvements and features contributed nothing to the value of the facility.” There is nothing in the record of this case that suggests the Ford plant is unmarketable for its current use. To consider it for lesser uses might make it more marketable (in the sense of available for more users) but would indeed fail to recognize those special features which make it exceptionably suitable for its present use. This would permit valuable features presently being used to entirely escape their just share of the burden of taxation. See Westing*167house Elec. Corp. v. Bloomfield Tp., 9 N.J.Tax 92, 97-100 (Tax Ct.1985), aff'd o.b. per curiam 9 N.J.Tax 108 (App.Div.1986).
Moreover, if plaintiff seeks to demonstrate that a property’s highest and best use is other than its current use, it is incumbent upon plaintiff, and not defendant, to establish that proposition (alternate use) by a fair preponderance of the evidence. Plaintiff has failed to do this.
Thus, I conclude under the factual circumstances demonstrated in the record that the highest and best use of the subject is its current use, i.e., an automobile assembly plant.6
With this determination in mind, I turn now to the various valuation methodologys advanced by the parties.
Market Data Approach.
The market data or direct sales comparison approach was considered by the appraisal experts for both parties. Edison’s expert, Barry J. Polayes, however, concluded that because of a *168lack of any sale properties with direct comparability it was not possible to employ a market data approach to value that would produce any meaningful result. Plaintiff’s expert, Parkhouse, expressed a strikingly different opinion. He felt that the market data approach should be given preponderant influence because that method reflects the actions of buyers and sellers in the marketplace. In his direct sales comparison approach he relied upon the sales of six (five for tax years 1983 and 1984 and a sixth for 1985) large industrial facilities, one as distant as Pico Rivera, California and another within a few miles of the Ford plant. These sales had an unadjusted sale-price range of $7.94 a square foot to $32.17 a square foot for building area with the land merged. The alleged comparable properties ranged in improvement size from 819,148 square feet to 2,710,-000 square feet.
Parkhouse adjusted the alleged comparable properties for the perceived dissimilarities between the sale properties and the Ford facility for (1) location and land-to-building ratio, (2) physical characteristics of the improvements, (3) size, (4) time and (5) condition. While his location and land-to-building ratio adjustment was based on a mathematical formula all of the other adjustments found support solely in the observation, experience and judgment of the appraiser.
After the application of adjustments, Parkhouse’s adjusted sale price range was from $9.93 a square foot to $12.05 a square foot including the land. He concluded that based on the sales the subject had a value of $10.75 a square foot for 1983, $11.00 a square foot for 1984 and $11.85 a square foot for 1985 which produced rounded value estimates of $14,500,000 for 1983, $15,000,000 for 1984 and $16,000,000 for 1985. As further support for his concluded values Parkhouse noted that he also considered sales data on 12 sales of industrial facilities, most of which were located in northern New Jersey, which had an unadjusted price range of $6.23 a square foot to $21.40 a square foot. These sales, while considered, were not deemed to be as suitable or relevant for a number of reasons, primarily because of smaller size, as the six sales upon which he relied, *169yet, were helpful and supportive of his value estimates for the Ford facility.
Edison contends, as does its appraiser, Polayes, that plaintiffs market data approach does not provide a reliable basis for deriving the market value of the Ford facility. Polayes attempted to utilize the market data or direct sales comparison approach but came to the conclusion that there was insufficient market data to derive a value conclusion from sales. He reviewed nine sales of large industrial facilities comparable in size to the Ford plant but concluded that realistic and meaningful comparisons could not be made to the subject property.
Essentially, Edison claims that plaintiffs market data approach is flawed for four reasons. First, with the exception of two of Parkhouse’s sales, the alleged comparable properties did not involve the same highest and best use as the Ford facility and therefore are not comparable. Second, the stated consideration in each of the sales did not reflect the value of the real estate. Third, the sale properties were not comparable to the Ford plant and could not be made comparable by adjustments of the dissimilarities. Fourth, many of the adjustments made by Parkhouse were erroneous and none were supported by market data in the record.
It has become well settled that there is no single authoritative approach to the valuation of real property. Much depends on the character of the property and the market data available. Here, plaintiff asserts that the market data approach to value as applied by its expert Parkhouse is the most effective method. Plaintiff argues that the six sales of improved industrial properties utilized by Parkhouse were, in fact, comparable to the Ford facility, and thus, after appropriate adjustments provide the surest guide to valuation in this case. Defendant contends, and plaintiff does not dispute, that in order for sale properties to be comparable they must be “[substantially similar] to the subject property so as to admit of reasonable comparison.” Congoleum Corp. v. Hamilton Tp., 7 *170N.J.Tax 436, 446 (Tax Ct.1985); Venino v. Carlstadt, 1 N.J.Tax 172, 175 (Tax Ct.1980), aff’d 4 N.J.Tax 528 (App.Div.1981).
This comparison, however, is not limited to the physical characteristics of the properties but also includes their economic aspects, e.g., highest and best use. As previously indicated, in the discussion relative to the highest and best use of the subject, there are two reasons underlying a highest and best use analysis. The Appraisal of Real Estate, supra at 68, 273-274. The first is to identify the most profitable use of the property, while the second is to assist in identifying comparable properties in a market data approach to value. Ibid. The appraisal text notes that “comparable improved properties should have the same or similar highest and best uses as the improved subject property.” Id. at 68. It would be inappropriate to use, as comparable, a property that does not have the same highest and best use as the subject. Id. at 68, 274. All of the special features of the improvements are not given full consideration if, in fact, the allegedly comparable properties were actually liquidations of industrial buildings which were either vacant or no longer being used for their originally designed highest and most profitable use. The validity of any market value estimate is fundamentally and directly related to an appraiser’s conclusion regarding the highest and best use of the property under consideration. Thus, since I have previously concluded that the highest and best use for the subject is as an automobile assembly facility, the sale properties offered as comparable must have the same or similar highest and best use.
Here, with the exception of two of the sales offered by Parkhouse (sale number one [Volkswagen of America, Inc. to Chrysler Corp.] and sale number four [Ford Motor Co. to Northrop Corp.]), none of the sale properties were substantially similar to the subject in economic characteristics because they were not of the same highest and best use, i.e., an *171automobile assembly plant.7 Therefore, I am of the opinion that these sales save for Parkhouse’s sales number one and four do not materially assist in any value determination for the Ford facility.8 They are simply not comparable.
It becomes necessary then to consider the two sales of industrial facilities which did involve the same or a similar highest and best use as the subject in question, i.e., Park-house’s sales number one and four.
Parkhouse’s first sale was that of an automobile assembly plant in Sterling Heights, Michigan by Volkswagen of America, Inc. to Chrysler Corporation in September 1983. The total sale price was $194,756,780 which included machinery and equipment in addition to the real estate. Parkhouse maintained, based on the deed, and what he had been told by others, that the parties allocated $72,975,365.46 to the real estate. Defendant argues that this was an arbitrary allocation that was not related to the fair market value of the real estate but rather to a number of other unrelated factors, primarily federal tax consequences. See in this regard the discussion of nonmarket “conditions of sale” in The Appraisal of Real Estate, supra at 322. Thus, defendant urges that the allocation set forth in the deed between the parties should be disregarded because of its arbitrary nature.
*172As support for the allocation, Parkhouse indicated that the allocated price was arrived at by the parties by means of appraisals. However, Parkhouse did not produce, nor did he have access to, the purported appraisals. In response to this, defendant’s expert reported, in his attempt to verify the allocation, that he was given four different allocations by four different appraisal firms ranging from $49 million to $72 million for the real estate. Polayes testified that, as related to him by a tax representative of the seller, the allocation was actually based on the seller’s book value at the time of sale. Thus, Polayes was of the opinion that the allocated price as presented was not reliable as a meaningful indication of the fair market price for the real estate alone.
Clearly, the allocation made by the parties as reflected in the deed cannot be ignored. It was significant to the parties for some reason. On the basis of the record in this case, however, I cannot ascertain what the purpose of the allocation was or whether a different allocation could have been made. No competent proof was submitted with respect to the value of the real estate alone or of the machinery and equipment involved in that transaction. Indeed, the appraisals upon which the allocation was purportedly made were not available nor were they reviewed by any of the appraisers in this case. Furthermore, no competent proof was offered with respect to the intention of the parties to that transaction. Thus, I am not satisfied by a fair preponderance of the evidence that the allocated price represents a negotiated price solely for the real estate, motivated by the real property value considerations of the parties. See McGinley Mills v. Phillipsburg, supra, 9 N.J.Tax at 516; cf. N.J.A.C. 18:12-1.1(a)23 (sales of industrial real property which include machinery and equipment are not usable in an equalization context when the value of the items are indeterminable).
While my rejection of Parkhouse’s sale comparable number one is premised essentially upon the failure to objectively demonstrate that the parties to that transaction allocated the sales price to the real estate based on fair market value considerations, I note that there are other deficiencies which *173make the sale as presented an unreliable indication of value for the subject.
For example, plaintiff did not persuade me by a fair preponderance of the evidence that Parkhouse’s cash equivalent discount from the allocated sales price for a preferential tax abatement to the purchaser was appropriate. See The Appraisal of Real Estate, supra at 318-322, for a discussion of cash equivalence. Parkhouse indicated that the seller had a tax abatement which it could and did transfer to the purchaser, and thus, the allocated sales price required a discount because it included an item that was over and above the value of the real estate. The proofs simply are insufficient to convince me that the tax abatement granted by the municipality was an item that the seller could transfer to the buyer and that it formed a part of their negotiations as to the ultimate purchase price for the real estate.
Another reason for eliminating the sale from consideration as a market value indicator was the appraiser’s combined adjustment for location and land-to-building ratio. Parkhouse was of the opinion that the dissimilarities in location, and the variables attendant thereto, such as availability of labor, labor costs, transportation costs, zoning, tax burdens, energy costs, availability of utilities and market for a particular product, cannot be adjusted and quantified separately but that it is not necessary to do so. This is because all of these variables are reflected in the land value or the prices that purchasers are willing to pay for the land at each of the locations. Parkhouse noted that “typically the ground values will [reflect] those pressures.”
In further explanation, Parkhouse related that there is not need to segregate the variables that comprise the locational differences since these are all reflected in the values of the underlying land. For example, if the sale property has fewer utilities available than the subject, the market will recognize this by producing a greater price for the parcel that does have a greater number of utilities.
*174It is readily observable that if Parkhouse’s hypothesis is correct then it is incumbent upon the appraiser to have a firm and factually supportable grasp of the market values of the land at the site of the comparable as well as the subject. Here, Parkhouse relied primarily upon four land sales in the area of his comparable sale number one which were never viewed nor inspected by him. The issue, thus, is not whether Parkhouse’s adjustment philosophy is appropriate, but whether his estimated land value at the site of the comparable property is reliable. Based on the record in this case I am not persuaded that it is.
When an expert sidesteps a detailed locational analysis involving a consideration of all the locational variables by “lumping” everything in the land value it is clearly incumbent upon that expert to produce a manifestly supportable land value for the comparable sale land. That was not provided in this case for comparable sale number one.
Parkhouse’s fourth sale was that of an automobile assembly facility in Pico Rivera, California by Ford Motor Company to Northrop Corp. in November 1982 for $65,000,000 which represented $32.17 a square foot for the improvements with the land merged. Defendant contends that Parkhouse’s analysis of this sale is deficient for a number of reasons. The first of which was Parkhouse’s inability to inspect the interior of this sale property. As defendant points out, in WCI-Westinghouse, Inc. v. Edison Tp., supra, which also involved a very large industrial facility, this court stated:
It is difficult to understand how an expert can make an appropriate comparison without viewing the property used for comparison. In the absence of an onsite inspection, there is no way for this court (or the expert for that matter) to determine whether the interior of the allegedly comparable property required an adjustment for interior finish or physical amenities. This alone causes little weight, if any to be assigned to [the] sale.... [7 N.J.Tax at 623]
Plaintiff argues that, even though its expert could not gain access to the allegedly comparable property, this does not detract from the reliability of the sale as a comparable because Parkhouse received information about the interior of the plant from representatives of the parties and the brokers negotiating the sale. It was plaintiff’s position that, with industrial buildings, it’s not that important for an appraiser to inspect the *175interior because industrial buildings have a degree of uniformity. I am not persuaded by plaintiffs arguments. In any appraisal analysis the task of an appraiser is to examine a property through the eyes of an informed, prudent purchaser. Here, first of all, an informed and knowledgeable purchaser would not purchase a large industrial facility without an inspection which would confirm the factual data furnished by others and reveal the observable condition, amenities and finish of the improvements.
Secondly, if plaintiff is correct, inspections would not be necessary to perform any appraisal of industrial property. An appraiser could simply go through the necessary appraisal procedure at his desk using only information supplied by others either in the mail or over the telephone without personal verification. This eliminates a great deal of an expert’s experienced appraisal judgment. See Industrial Real Estate, supra at 456. Even if I were to be persuaded to accept the pragmatic proposition that an interior inspection of an allegedly comparable industrial property is not always necessary — because appraisers do need some latitude with regard to the practicalities of the use of a market data approach — I am not convinced, by the appropriate standard, that Parkhouse was in sufficient command of the details of the interior of the Pico Rivera facility to permit a rational comparative analysis to be made in this case.
Defendant has another objection to this sale and contends that it should be eliminated from consideration because of the deficiency in Parkhouse’s combined adjustment for location and land-to-building ratio. Parkhouse made a 65% adjustment for these items based on essentially the estimated land value of the subject as compared to the land value of the allegedly comparable property. As with Parkhouse’s sale number one, the expert did not perform a detailed locational analysis, but, again, relied solely upon the respective land values to reflect dissimilarities in location. Here, also, however, Parkhouse used a compilation of land sales for the comparable property that were provided to him by others. He did not see or inspect the vacant land sale *176properties. Essentially, he relied upon an estimate of land value provided by other appraisers.
Initially, the magnitude of the location adjustment (65%) causes one to wonder whether the properties can be compared. See Owens-Illinois Glass Co. v. Bridgeton, supra, 8 N.J.Tax at 512 (a “63% adjustment casts grave doubt upon the reliability of any conclusion reached concerning [a comparable] sale”). Even more fundamental, however, is the fact that the location adjustment rests upon the validity and reliability of Park-house’s land value at the comparable site. The supporting land values in Pico Rivera, California were not only important but were crucial. If that land value is suspect or unreliable then the adjustment is just as suspect or unreliable and thus, the sale is not usable. Inasmuch as Parkhouse relied on others for the land value estimate at the site of the comparable it is not acceptable. For this reason also, comparable sale number four cannot be utilized for a value determination. . '
Lastly, defendant contends that the sale must be ignored because plaintiff's appraiser did not adjust the sales price for the fact that the seller, as part of the price, was able to use two buildings “rent free” for an unspecified period of time. Defendant also argues that the sales price did not reflect a sale of all of the property rights but rather a sale of the seller’s leased fee inasmuch as an 18-acre portion of the property was subject to a lease with which Parkhouse was not familiar. All of these factors in combination militate against the suitability of comparable sale number four as reflective of fair market value of the subject.
Based on the record in this proceeding, I conclude that there is an absence of sufficient sales data having probative value, and thus, I must, of necessity, reject the market data or direct sales comparison approach as developed in this case.
Cost Approach.
A. Land Value.
The first essential aspect of the cost approach is the determination of land value. The most relevant method to determine *177this value is a straightforward sales comparison approach. The Appraisal of Real Estate, supra at 446.
With this method, sales of similar parcels of land are analyzed, compared and adjusted to provide a value indication for the land being appraised. [Ibid.]
The amount of land to be valued at the Ford facility site is 76.54 acres as of the assessment dates for each of the tax years of 1983, 1984 and 1985. Plaintiffs expert, Parkhouse, valued the land at $47,500 an acre for 1983, $50,000 an acre for 1984 and $65,000 an acre for 1985 while defendant’s expert, Polayes estimated the land value to be $83,000 an acre for 1983, $91,000 an acre for 1984 and $100,000 an acre for 1985.
For tax years 1983 and 1984 Parkhouse relied upon ten sales of allegedly comparable vacant land which produced an unadjusted sales price range of $11,150 an acre to $75,954 an acre. Adjustments were made for differences in location, zoning, time, size and topography with size and topography considered most important. These adjustments produced an adjusted sales price range of $14,495 an acre to $64,561 an acre. Parkhouse selected $47,500 an acre for 1983 and a 5% increase to $50,000 an acre for 1984 feeling that the value of the subject was within the price range but closer to the upper adjusted limit.
For tax year 1985 he relied upon seven additional sales of vacant land which had an unadjusted price range of $55,172 an acre to $70,441 an acre. After the same series of adjustments that were made for 1983 and 1984 his price range limits remained the same, i.e., from $55,172 an acre to $70,441 an acre. Parkhouse concluded from this data that $65,000 an acre represented the value of the subject land for 1985.
Defendant’s expert, Polayes, relied upon five sales of vacant land to support his value estimate of $83,000 an acre for tax year 1983. Four of the five sale properties were located in an LI (light industrial) zone as is the Ford facility while the fifth sale property was located in an ROL (research, office and laboratory) zone. The four sales with the same zoning, which ranged in size from seven acres to 16.48 acres, produced an unadjusted sale price range of $53,498 an acre to $74,229 an *178acre while the fifth sale, a 24.79 acre parcel in the ROL zone sold for $232,957 an acre.
Contrary to the usual appraisal concept that the larger the parcel of land, the lower the unit or per-acre value, Polayes indicated that unit prices or per-acre prices and values increase as the size of the parcel itself increases, and thus, the subject 76.54-acre parcel enjoyed a plottage or assemblage value because the smaller parcels have less utility. Although Polayes indicated that his market analysis supported this concept there were no land sales offered to confirm the theory. Based on this plottage or assemblage philosophy Polayes, instead of adjusting smaller parcels downward in comparison to the much larger subject, adjusted the smaller comparables upward and thus his adjusted price range for four of his sales with the same zoning increased from a low of $81,400 an acre to $104,663 an acre with his fifth sale in the ROL zone producing an adjusted value of $128,270.
For 1984 Polayes relied upon the same sales to support his estimated land value of $91,000 an acre, a 10% increase over 1983. He simply modified his time adjustment to reflect a 1984 assessment date value. With this change his adjusted price range for 1984 became $81,250 an acre to $113,570 an acre for the comparably zoned LI land and $142,104 an acre for the ROL zoned parcel.
He added two additional land sales to his analysis for 1985, one of which reflected $65,349 an acre for 10.9318 acres and fell within his unadjusted sales price range for 1983 and 1984. He, again, applied his plottage or assemblage concept to produce an upwardly adjusted sale price for the comparable property of $92,131 an acre.
The second additional sale was of land in an LI zone but was sold contingent upon the ability of the buyer to develop the land for apartments. This contingency was satisfied and apartments are presently being constructed. This sale cannot be considered as it clearly is not reflective of industrial land values. As previously indicated, based on his sales and plot*179tage or assembly precept, Polayes concluded a land value of $100,000 an acre for the subject for tax year 1985.
The major difference in the value estimates of the experts is one involving the operation of competing appraisal principles. Parkhouse adjusted the smaller comparable land sale parcels to reflect the fact that unit values or per-acre values for smaller parcels are typically higher than similar unit values for larger parcels. See Newark v. Vernon Tp., 1 N.J.Tax 90, 94 (Tax Ct.1980), aff’d 179 N.J.Super. 332 432 A.2d 106 (App.Div.1981); Alpine v. Alpine Hills, Inc., 1 N.J.Tax 136, 141 (Tax Ct.1980), aff’d 2 N.J.Tax 391 430 A.2d 641 (App.Div.1981); Romulus Dev. Corp. v. West New York, 7 N.J.Tax 305, 311 (Tax Ct.1985), aff’d 9 N.J.Tax 90 (App.Div.1986); Red Devil, Inc. v. Union Tp., 5 N.J.Tax 1, 8 (Tax Ct.1982). This is normally true because there is a greater market for the smaller parcels; hence they tend to produce greater unit prices.
On the other hand, Polayes advanced the theory that the smaller parcels sold for less per acre than would the larger subject because the larger parcel would have greater utility, and thus, there exists plottage or assemblage value. The primary land sales proffered by the experts do not persuasively support Polayes’ value notion of assemblage, nor am I convinced of plottage or assemblage value by Polayes’ brief reference to two land sales in South Brunswick Township and an improved sale in Mahwah. Rather, the primary land sales utilized by both experts generally support the declining-unit-value phenomenon, i.e., that smaller parcels do sell for greater unit values than larger parcels. Size in and of itself is no guarantee of a plottage increment in value or that a market exists for a large parcel of industrially zoned land. Therefore, Parkhouse’s negative adjustments for size are more appropriate than Polayes’ positive adjustments for assemblage.
There were other deficiencies noted in Polayes’ market data approach as well. As indicated, he utilized a sale of property located in an ROL zone which permits considerably greater office-building development, and thus, is not, in fact, compara*180ble. He also utilized a sale of property located in an LI zone that was contingent upon development for a multi-family residential use, clearly not reflective of industrial values. It is difficult to know the extent to which Polayes relied upon these sales in deriving his estimated land values.
Lastly, Polayes utilized a multiplication or factoring system of adjustment for the land sales that is no longer included in modern appraisal texts because it is generally not currently considered a valid adjustment technique. See Eaton, Real Estate Valuation in Litigation (1982) at 153-154; see also Boyce and Kinnard, Appraising Real Property (1984) at 196-197 (“Cumulative percentages are valid only if a causal interrelationship is shown to exist and this is not often the case”) and Ring, The Valuation of Real Estate (2 ed. 1970) at 140-141.
With these considerations in mind, I find that Parkhouse presented a sounder and more persuasive analysis of the land sales than did Polayes, and therefore, adopt Parkhouse’s unit values of $47,500 an acre for 1983, $50,000 an acre for 1984 and $65,000 an acre for 1985. Application of these unit values to the subject tract of 76.54 acres as calculated by Parkhouse produced the following rounded land values for the subject: 1983 — $3,650,000, 1984 — $3,825,000 and 1985 — $5,000,000.
B. Improvement Value. ■
Both plaintiff and defendant developed improvement values by means of the cost approach methodology. As a matter of fact, defendant insists that, while recognizing its limitations, the cost approach “must be the predominant mode of analysis” in this case because this approach is “the only way to rationally” value the Ford facility.
Plaintiffs expert, Parkhouse, developed a cost approach value estimate for the subject essentially as a backup for his market data approach estimate of value. This, however, was not to detract from his cost approach value. He explained that *181he went to great lengths to estimate replacement cost new,9 had good land sale support for his estimated land values and his estimate of accrued depreciation based on an adjusted age-life concept was market supported by the depreciation reflected in the sales he relied upon in his market data approach. Therefore, his cost approach, although not his primary approach, was given suitable weight in his final valuation estimates.
Parkhouse utilized the Marshall Valuation Service, a nationally known building cost manual published by the Marshall and Swift Publication Company. He applied the manual’s calculator cost method for each portion of the improvements. This method, essentially, requires a segregation of the different portions or sections of the improvements and placement of those segments into an appropriate classification as set forth in the manual. The manual’s building descriptions or classifications are not precise building specifications but are rather meant to provide a description of the major elements of the many buildings utilized as a basis in formulating each classification. The selected classification reveals, after necessary adjustments and mathematical computations, the approximate cost to replace the improvement as of a particular time. Parkhouse, based upon the building data available to him, his experience and observation, was of the opinion, that the Ford Plant was an average industrial complex which was most appropriately identified as a blended class C and class S average manufacturing building.
Parkhouse estimated that the replacement cost new of all the building improvements was $32,788,728 for 1983, $34,975,982 for 1984 and $37,448,419 for 1985. The next step was to estimate accrued depreciation in all forms. Parkhouse employed what he termed an “adjusted age-life basis” to derive appropriate depreciation estimates. An age-life concept has two components, the effective age of the improvements and *182their total economic life. The latter divided into the former produces a percentage reflective of total accrued depreciation. Parkhouse derived the total economic life component from the Marshall manual tables which provided that industrial structures similar to the Ford facility typically have an economic life of approximately 40 years. He estimated that the average age of all the Ford improvements was 25 years but because of the alleged deteriorated condition of the roof on the main assembly plant and the alleged deteriorated condition of the free-standing metal-clad warehouse he determined that the average effective age of the Ford facility was approximately 28 years. Thus, this produced total estimated depreciation for tax year 1983 of 70% (28 -h 40 = 70%). This estimate was increased by 2x/2% for 1984 (29 -h 40 = 72.5%) and 5% for 1985 (30 -h 40 = 75%) to reflect the increase of each additional year in the effective age numerator.
In support of his adjusted age-life methodology, Parkhouse derived a range of depreciation from the sales used in his market data approach. As stated in Parkhouse’s appraisal report, his market depreciation analysis was described as follows:
This analysis begins with the deduction of land and site improvement value from the sale price of each transaction. The result of this calculation is depreciated building value. Replacement cost new of the building at the time of sale is then estimated, using the same technique employed in arriving at replacement cost new for the subject. The depreciated building value is subtracted from replacement cost new, with the residual reflecting total dollar depreciation. Total depreciation, as a percentage of replacement cost new, is then calculated.
Employing this technique with each of the six sales in his market data approach produced a range of overall depreciation from 57.2% to 84%. Since Parkhouse’s depreciation estimates of 70% for 1983, 72.5% for 1984 and 75% for 1985 fell within the parameters indicated by the sales, he was of the opinion that his estimates were market supported.
Parkhouse then deducted the estimated depreciation from his replacement cost for each tax year and derived replacement cost estimates of $9,836,618 for 1983 ($32,788,728 — 70% or *183$22,952,110 = $9,836,618), $9,618,395 for 1984 ($34,975,982 - 72.5% or $25,357,587 = $9,618,395) and $9,362,105 for 1985 ($37,448,419 - 75% or $28,086,314 = $9,362,105).
To these amounts Parkhouse added the depreciated cost of the various site improvements such as asphalt and concrete paving, fencing, rail, water and fuel tanks which costs were obtained from various sections of the Marshall manual. These costs were depreciated at 40% based on his observation, experience and judgment. These depreciated site improvement costs were $1,150,000 for 1983, $1,200,000 for 1984 and $1,300,000 for 1985.
Lastly, the land values of $3,650,000 for 1983, $3,825,000 for 1984 and $5,000,000 for 1985 were added for each of the indicated years which produced rounded cost approach values of $14,600,000 for 1983 and 1984 and $15,650,000 for 1985.
The cost approach, as previously noted, was also employed by Edison. Its experts, however, did not employ the Marshall Valuation Service to estimate the cost of the improvements. Edison’s appraisal expert, Polayes, was of the opinion that the Marshall manual, when dealing with improvements as large and complex as the Ford facility, has a number of limitations which render the cost conclusions derived from its use highly subjective, and therefore, speculative.
He indicated, for example, that the calculator cost sections provide only a very general description of a particular classification, and thus, it is extremely difficult to take the Ford plant and place it within the generalized classifications of the manual. Additionally, the classifications do not account for the special features in the Ford plant such as the electrical power capacity, waste treatment systems, air compressors and the elaborate plumbing, all necessary to accommodate the automobile assembly process. Polayes was further of the opinion that even resorting to the segregated cost sections in the manual would not permit use of the manual because the descriptive categories in those sections are also too general.
*184Lastly, Polayes questioned the ability of a real estate appraiser, lacking an engineering background, to properly identify the various components of a structure and place them in the appropriate classification as set forth in the Marshall manual.
As a consequence, he felt that the cost analysis should be more specific and should also be accomplished by one who could more accurately identify the structural components of a facility such as the Ford plant. Therefore, Oliver B. Spence, a civil engineer, was selected to perform a cost analysis utilizing a more detailed cost service, to wit, the R.S. Means manual. Polayes’ function would be to estimate the accrued depreciation of the buildings and site improvements and the value of the land.
Spence indicated that he used the R.S. Means cost service because it was more “definitive” than the generalized categories or classifications found in the calculator cost sections of the Marshall Valuation Service.
In general, Spence’s cost analysis required that the improvements be divided into building components such as floor, roof, walls, structural steel, electrical, etc., to which he applied a square foot or lineal foot cost derived from the R.S. Means cost service.10 After a total cost per square foot was obtained a size modifier was applied to reflect the fact that the manual normally provides unit costs for buildings smaller than the Ford facility. The size modifier reflects the “economy of scale” that can be achieved with larger buildings.
This procedure produced a total cost of $37.92 a square foot for 1,106,521 square feet of the Ford main assembly building. Spence indicated that he checked this square foot cost estimate by using another method which produced $38 a square foot, hence he was confident of his figure of $37.92 and saw no need to change it.
*185Thereafter, using the R.S. Means manual and the same procedure, he performed separate calculations for the office and garage and the miscellaneous buildings.11 Then he calculated the cost of the site improvements and a number of special features. Lastly, he computed the replacement cost of such items as tanks, piping and ducts, electrical support systems and fan rooms.
After the application of cost conversion factors to bring the derived costs to the appropriate assessment dates, the resulting cost estimates were $50,851,005 for 1983, $56,419,614 for 1984 and $58,321,596 for 1985.
Plaintiff’s appraisal expert, Polayes, then applied his accrued depreciation estimates. The estimates of depreciation for the improvements, in general, were derived from the depreciation schedules in the Marshall Valuation Service and amounted to 35% for 1983, 37% for 1984 and 39% for 1985. The site improvements were depreciated based on observation at 50% for 1988, 55% for 1984 and 60% for 1985 while the special items such as tanks and related systems, piping and ducts, air handling units, electrical support systems and fan rooms were depreciated at 60% for 1983, 64% for 1984 and 68% for 1985.
An application of the stated depreciation estimates to the replacement cost new with the addition of Polayes’ estimated land values produced total values for the Ford facility by the cost approach of $36,988,988 for 1983, $40,338,794 for 1984 and $40,797,263 for 1985.
As plaintiff has pointed out, the Tax Court has accepted the cost approach procedure using recognized cost services such as the Marshall Valuation Manual, R.S. Means cost service and the Real Property Appraisal Manual for New Jersey Asses*186sors. See, e.g., ITT Cont’l Baking Co. v. East Brunswick Tp., 1 N.J.Tax 244 (Tax Ct.1980); Pennwalt Corp. v. Holmdel Tp., 4 N.J.Tax 51 (Tax Ct.1982), Coca-Cola Bottling Co. of N.Y. v. Neptune Tp., 8 N.J.Tax 169, 174 (Tax Ct.1986). Acceptance of this methodology is premised, however, on reliable proofs being adduced in support of the method and an appropriate procedural application of the method. WCI-Westinghouse, Inc. v. Edison Tp., supra, 7 N.J.Tax at 619.
As defendant has noted, the cost approach is particularly significant when the subject property is one which is “not frequently exchanged in the market.” The Appraisal of Real Estate, supra at 349. Absent sufficient comparable sales the cost approach can provide “the best reflections of market thinking and, thus, of market value.” Ibid.
Neither plaintiff nor defendant contends that the cost approach is not an acceptable approach in this case. Rather, each attacks the way in which the method was utilized by its adversary. Defendant argues that the approach by Parkhouse is flawed because: (1) the calculator cost method of Marshall Valuation Service is not only too generalized but can only be used for substantially smaller facilities, (2) it omitted the cost of a number of items such as.electrical wiring, the paint facility and related piping, extra boilers, air handling units and air conditioning, etc., and (3) the depreciation analysis was deficient and produced a “worthless” estimate.
Plaintiff asserts that the cost methodology employed by defendant was unreliable and should be rejected: (1) because of Spence’s failure to follow the instructions in the R.S. Means manual relative to the use of the size modifier and (2) the depreciation estimates were simply numbers taken from a national cost service unsupported by “market data, observation, experience and explicable analysis and calculation.” WCI-Westinghouse, Inc. v. Edison Tp., supra, 7 N.J.Tax at 629.
My review of the evidence relative to the cost approach as employed by both parties leads to the conclusion that the criticisms of both parties are right. I am not convinced by a *187fair preponderance of the evidence that either party provided a credible estimation of fair market value for the subject by means of a cost approach methodology. Further, the nature of the proofs do not permit me to determine an appropriate value through the cost data provided by both parties. See Samuel Hird & Sons, Inc. v. Garfield, 87 N.J.Super. 65, 74 208 A.2d 153 (App.Div.1964) (the court has an obligation to apply its judgment to the valuation data and determine value provided there is enough substantial and competent evidence to enable the court to do so).
In reviewing the cost approach as performed by plaintiffs appraisal expert, Parkhouse, I have noted a number of items which when considered on an individual basis might not cause me to reject his ultimate cost calculations. In combination, however, there is sufficient doubt in my mind as to the results derived by Parkhouse to conclude that plaintiff has not met its burden of proof as to the reliability of the cost approach as applied by plaintiffs expert.
Initially, defendant argued that the Marshall Valuation Service was not an appropriate pricing mechanism for the Ford facility because Marshall was not designed for buildings that exceed 400,000 square feet as does the subject. Relative to this point, a good deal of the testimony in this case focused on the fact that the Marshall manual does not provide area perimeter multipliers for buildings in excess of one million square feet. As a matter of fact, the area perimeter table provided in the manual only goes to 400,000 square feet.
Clearly, Parkhouse had to make some assumptions in order to extrapolate a perimeter multiplier. He explained that the perimeter table is a mathematical table which can be projected out beyond the numbers reflected in the table — even though the Marshall manual has not provided any figures for buildings in excess of 400,000 square feet. The problem is that I cannot check to see if Parkhouse’s assumptions and ultimate selection of an area multiplier are, in fact, accurate. The manual is *188silent in this regard. Thus doubt as to the appropriateness of Parkhouse’s cost approach procedure begins.12
Next, as was previously noted, Parkhouse’s classification within the manual for the main assembly area was a blend of two classifications, a class C average manufacturing at $18.24 a square foot and a class S average manufacturing at $16.38 a square foot, and thus, his base cost of $17.36 a square foot was also the result of a combination of the two classifications. The difficulty with this classification selection, however, is that it presupposes equal parts of class C and class S. On cross-examination Parkhouse could not relate what portions of the Ford facility were class C or class S, yet he conceded that the percentage of each class and the accuracy of selection could affect the ultimate conclusion. Thus, doubt continues and grows.
The Marshall manual distinguishes between “heavy steel frame” and “steel frame” in its class descriptions for manufacturing buildings. The difference in cost can be substantial. In this matter I have sufficient doubt as to whether the Ford facility is totally or substantially “heavy steel frame” within the meaning of the Marshall classification to question Park-house’s selection of “steel frame.” This doubt is based on Parkhouse’s inability to indicate what measure of thickness, weight or strength would be considered “heavy steel” as opposed to simply “steel frame” and what measure would apply to the Ford facility. Moreover, the proofs demonstrate that the Ford structure was supporting craneways, automobile assembly conveyors, mezzanines with offices, cafeteria, dining rooms, locker rooms, a medical clinic and fan rooms as well as lighting, electrical, plumbing systems and compressor lines. I do not *189conclude that this demonstrates “heavy steel” construction. Based on the proofs I simply cannot say.
There were a number of items that Parkhouse considered as part of the base classification cost, and therefore, did not provide any additional segregated cost. The Ford facility enjoys electrical service of 26.4 KV power. The record demonstrates that this electrical capability exceeds that found in the typical general purpose manufacturing building, and therefore, leads to the conclusion that some segregated cost calculation should have been made for this item. Additionally, Parkhouse did not calculate and include the cost of the secondary electrical wiring in the plant nor the power panel. His explanation that the cost of many of these items is not calculated because the appraiser makes subjective mental adjustments and allowances for items such as these in his final selection of a base cost is simply not persuasive. In order to determine the appropriateness of an appraiser’s cost procedure a court must require explicable calculation and analysis.
In similar fashion Parkhouse neglected to provide a cost for the air handling units because he assumed these were included in the base cost. He did not provide cost figures to the existing air conditioning for the offices, cafeteria and medical clinic in the mezzanines. Nor did he estimate the cost of the heavy drainage sewer lines, the extra boilers and the air compressors.
Next, Parkhouse used a base cost of $18 a square foot for the offices, cafeteria, dining rooms, etc., in the mezzanines in the main assembly area while providing a base cost of $37.65 a square foot for the offices in the administration building. His explanation for the variance was that he could not find a specific category for a suspended structure such as the mezzanines in the Marshall manual therefore he used the cost on other descriptions for industrial type space. This unpersuasive explanation lends additional doubt to Parkhouse’s cost approach.
The foregoing are not intended to provide an exhaustive laundry list of items that individually result in a rejection of *190plaintiff’s cost approach procedure but are sufficiently representative of the deficiencies noted that, in combination, render this court unable to conclude a justifiable replacement cost.
In regard to the depreciation estimates advanced by Park-house, suffice it to say that the estimate of functional obsolescence is deficient because it is premised on the fact that the present use (automobile assembly) will not continue based on his opinion that it was not the highest and best use. Thus, his measure of functional obsolescence, although incorporated in his age-life formula, was based on the suitability of the improvements for alternative general purpose industrial uses. The existing use, however, is the highest and best use for the Ford facility and thus takes maximum advantage of the facility’s existing features, and thus, the functional obsolescence would be less than if the plant were characterized as general purpose industrial.
The deficiencies in defendant’s cost approach methodology can be set forth in less detail. The first is Spence’s failure to follow the instructions of the R.S. Means manual relative to the use of a size modifier. As previously indicated, the size modifier is provided in order to account for the affect that size has relative to square foot cost. In the words of the manual,
In general, for buildings built to the same specifications in the same locality, the larger building will have the lower [square foot] cost. This is due mainly to the decreasing contribution of the exterior walls plus the economy of scale usually achievable in larger buildings.
The manual then provides a conversion procedure which gives factors that will “convert costs for the typical size building to an adjusted cost for the particular [building].” The instructions of the manual direct the cost estimator to divide the building afea (in square feet) of the subject by the typical size (again in square feet) for the particular building type shown on a table. This results in a factor which when entered in the area conversion scale will provide an appropriate cost multiplier for the subject’s building size. Spence simply did not use the typical size figure shown on the table but instead used the number reflected as the upper limit of the typical range. The difieren*191tial produced by correctly following the directions was strikingly demonstrated in the cross-examination of Spence. Spence, on direct-examination concluded a unit cost of $37.92 a square foot for the main assembly area of some 1,106,521 square feet. On cross-examination when required to follow the manual’s instructions precisely, the result was a cost of $22.84 a square foot or a total difference of approximately $16,686,337. Spence concluded that this was an absurd result because he had estimated the unit cost by another means and it had produced $38 a square foot. Thus, it would seem that if a correct application of the manual’s procedure produced an untoward result it was to be ignored. Spence’s explanation simply does not persuade me that he may ignore the manual’s specific directions and yet achieve the correct result. I am not certain, therefore, by the appropriate standard that either conclusion, i.e., $22.84 a square foot or $38 a square foot, is correct.
The second deficiency in defendant’s cost approach methodology relates to the estimates of accrued depreciation. In estimating depreciation Polayes conceded that he relied primarily upon the depreciation schedules set forth in the Marshall Valuation Service although he checked the estimates with the depreciation schedules in the Real Property Appraisal Manual for New Jersey Assessors. There is no indication in the record that the depreciation estimates were based on anything other than references to the tables in the cost services. As this court indicated in WCI-Westinghouse, Inc. v. Edison Tp., supra:
No nationally standardized table or chart ... can effectively replace an appraiser’s justification of accrued depreciation in a particular property based upon market data, observation, experience and explicable analysis and calculation. [7 N.J.Tax at 629]
Since the probative value of an expert’s opinion must stand or fall upon the facts and reasoning offered in support of that opinion, Dworman v. Tinton Falls, 1 N.J.Tax 445 (Tax Ct.1980), aff’d o.b. per curiam 3 N.J.Tax 1, 434 A.2d 1134 (App.Div.1981), certif. den. 88 N.J. 495, 443 A.2d 709 (1981), I am not convinced by the appropriate standard of proof that defendant’s estimates of accrued depreciation were appropriately derived and adequately supported in this matter. Based on the fore*192going, I am constrained to reject the cost approach methodology as presented by each of the parties.
Income Approach.
In spite of the fact that Edison’s appraisal expert, Polayes, determined that the highest and best use for the Ford facility was for a continuance of its existing use i.e., automobile assembly, and provided an appraisal to reflect the value of the subject for that use, he also considered the property for an industrial subdivision into a number of multi-tenant industrial uses.13 Polayes, in effect, did two appraisals, valuing the property for each of two suggested highest and best uses. Presumably, this was done on the basis that this court might reject his assertion that the highest and best use of the subject was as an automobile assembly plant. It was his opinion that the retrofitting or subdivision of the subject property into a multi-tenant industrial and office complex would be a higher use in a hierarchy of highest and best uses than the use proposed by Ford’s appraisal expert, Parkhouse, viz., a general purpose manufacturing facility-
Polayes was of the opinion that the industrial market revealed a demand in Edison and specifically in the area of the Ford facility for multi-tenant industrial use. This was based on the conversions that he saw taking place in the New Jersey industrial sales market, specifically, the Fedders plant in Edison, Township, Middlesex County, the Owens-Illinois facility in North Bergen Township, Hudson County, the Lilly-Tulip plant in Holmdel Township, Monmouth County and the Firestone warehouse building in Cranbury, Middlesex County.14
*193In order to construct an economic approach to value, Polayes advised defendant’s engineering expert, Oliver B. Spence, as to the criteria necessary to create marketable units in the Ford facility. Spence was instructed to maximize the utilization of space while subdividing the facility in a manner that would provide for industrial units or segments of under 200,000 square feet. The latter size was selected because it was felt that it would be easier to market industrial units in a size range of 50,000 square feet to 150,000 square feet and still keep the actual cost of readaptive use as low as possible.
Accordingly, Spence subdivided the plant into 12 units ranging in size from 63,360 square feet to 205,038 square feet which Polayes felt were marketable based on his observation of the demand for industrial space in the relevant market. Spence’s estimated cost of conversion was $4,465,966.15
Utilizing Spence’s projected subdivision, Polayes then analyzed nine industrial lease comparables for 1983, 12 for 1984, 14 for 1985 and six office lease comparables for 1983 and 1984 and eight for 1985 in order to estimate fair market rents for the units as designated by Spence. After making adjustments for time, location, physical features, age and condition, he estimated ranges of industrial fair market rentals of $1.50 a square foot to $3.75 a square foot for 1983, $1.65 a square foot to $4.10 a square foot for 1984 and $1.80 a square foot to $4.50 a square foot for 1985. Polayes’ adjustments for office space produced office fair market rentals of $8.00 a square foot for 1983, $8.80 a square foot for 1984 and $9.25 a square foot for 1985.
*194Polayes then ascribed a potential annual rental for the entire facility. Thereafter, he estimated and deducted an allowance of 5% for vacancy and credit losses and 10% for management, leasing, reserve for replacements and repairs and miscellaneous expenses which produced a net operating income that was converted to a capital value with an overall capitalization rate of 12% for 1983, 11.7% for 1984 and 11.6% for 1985. From the capitalization value thus derived, he deducted Spence’s estimated cost of conversion to provide a value as of each of the assessment dates. Although it was estimated that it would take approximately one year to complete the conversion and fully rent the facility, no deduction was taken for time because Polayes was of the opinion that this would be “offset by [a] rising rental market and [a] strong demand for industrial facilities within the subject market area.”
Polayes’ income approach analysis produced rounded values of $27,140,000 for 1983, $30,795,000 for 1984 and $34,011,000 for 1985.
Polayes also utilized a discounted cash flow analysis 16 with an assumption that it might take up to two years to fully lease the facility. This was based on the fact that an owner would be leasing portions of the facility periodically. It was assumed that during the first 90 days there would not be any tenants but that preleasing negotiations would be taking place. During the next 90 days some of the space would be occupied, and thus, begin to produce rental income or cash flow. This would progress until fully rented. Polayes estimated the rentals and discounted these for time in a discounted cash flow analysis. The value results produced by this analysis, according to Polayes, supported his income approach valuations.
Plaintiff’s appraisal expert, Parkhouse, on the other hand, considered, but did not develop, an income approach based on the Ford facility’s potential for subdivision. Although Park-*195house recognized that there is a market for large industrial facilities composed of investors contemplating purchase for conversion to multi-tenant use, he was of the opinion that a subdivision approach, properly applied would yield “a lower value than a sale to a single user when costs of effecting the subdivision, proper discounting for marketing time, and weak economic demand [were] taken into account.”
Plaintiff argues that Polayes’ income approach to value must be rejected because any reliance on defendant’s income methodology would, at best, be fanciful speculation. Plaintiff relates that there are so many inconclusive and uncertain components of defendant’s income approach that it cannot be given serious consideration. These variable factors include “the amount of rentable square feet for multi-tenant use, the size and configuration of the various components, the rental rates for those components which depend on factors such as size, length of leased term, type of utilization and the like and rent up time.”
I find that I am in agreement with plaintiff’s contentions regarding the speculative nature of the income analysis developed by defendant’s experts. The income approach to value, in general, must be developed with care and as our Supreme Court has noted, “one should hesitate to accept its answer without checking against all available data.” New Brunswick v. Tax Appeals Division, 39 N.J. 537, 544, 189 A.2d 702 (1963). Here, however, because of the added possible conjecture of the subdivision methodology an added cautionary note must be sounded.
This court has indicated that the subdivision or anticipated use technique, while not a traditional valuation procedure, is not to be summarily rejected but rather must be viewed with a great deal of caution. See Genola Ventures v. Shrewsbury, 2 N.J.Tax 541 (Tax Ct.1981) (the Tax Court will not summarily reject the subdivision or development method of valuing vacant land, but an expert will not be permitted to roam at large in the realm of fancy); Center-Whiteman Corp. v. Fort Lee, 4 N.J.Tax 153 (Tax Ct.1982) (the proponent of a vertical subdivision [conversion from rental apartments to condominiums] ap*196proach to value bears a very heavy burden of proof); Berkeley Arms Apartment Corp. v. Hackensack, 6 N.J.Tax 260, 297-298 (Tax Ct.1983).
Additionally, this court has specifically rejected the utilization of the income approach in dealing with large (645,062 square feet) owner-occupied industrial property. Judge Crabtree of this court put it this way:
The income approach posited by defendant’s expert fares no better. Dealing as he was with owner-occupied property, he was impelled to construct an amalgam of attenuated hypotheses concerning economic rent, projected expenses, whether the leases would be gross or net, whether the property is best suited for a single tenant or multiple tenancies and the anticipated return on investment. As stated in one authoritative work:
The difficulties and disadvantages of the income approach are numerous. They all revolve around the fact that a large number of estimates must be made; therefore, in most cases a complex set of relationships must be developed. The appraiser may have to inject his own analytical judgment and skill into the analysis more than is desirable. [Industrial Real Estate, supra at 478]
The sheer number of interdependent assumptions, none of them fortified by hard data, precludes acceptance of defendant’s expert’s income approach as a viable indicator of value. [Shulton, Inc. v. Clifton, 7 N.J.Tax 208, 217 (Tax Ct.1983), aff’d 7 N.J.Tax 220 (App.Div.1984)]
Both plaintiff and defendant presented expert testimony relative to a conceptual plan for readaptive purposes for the Ford facility for multi-tenant industrial uses. Plaintiffs architectural expert, Walter, set forth a conceptual plan which subdivided the existing plant into ten areas ranging in size from approximately 25,000 square feet to 129,000 square feet each with its own office and loading zone. He, however, concluded that only 1,039,170 square feet would be rentable (an 80% efficiency ratio)17 due to discounting for walls, stairs, circulation and materials-handling space.
Spence, defendant’s engineering expert, on the other hand, did not agree with Walter’s conceptual use plan because it *197permitted excessive common area space (nonleasable space) and, also, because Walter’s plan called for an interior traffic arrangement which Spence felt would generate excessive congestion, noise and pollution. Essentially, however, Spence’s objection focused on "too much lost space” or the nonmaximization of leasable area. Spence’s conceptual use plan subdivided the Ford facility into 12 areas with a sum total of subdivided leasable space of 1,255,699 square feet (93% efficiency ratio) or 216,529 square more rental area than Walter’s plan for readaptive use. Spence conceded, however, that Walter’s plan was not unreasonable but that he simply disagreed with it.
Both Walter and Spence, however, agreed that a prudent investor-purchaser of the Ford facility would not convert or retrofit the plant first. Rather, one would wait for the prospective tenants to advise of their specific needs and space allocations. In this manner one avoids unnecessary cost. Size, configuration and use of the subdivided areas then would be dependent upon market demand not anyone’s conceptual plan.18 Neither expert could say with any degree of certainty that his plan would ultimately become fact. As a matter of fact neither believed that his plan would become fact. Both plans were simply reasonable concepts.
This demonstrates the speculation19 that is inherent in the income approach formulated by defendant’s expert. The starting point for the income approach, i.e., fair market or economic rent would be dependent upon the size of the space required by the tenant, the term of the lease, and the investment to be made by the owner to make space suitable for a potential tenant. There are no typical or normal space requirements that *198can be set forth with any degree of certainty. Rather, the facility would be developed as the experts have indicated — on a wait and see who needs space basis. The facility would then be developed or rehabilitated by an investor-purchaser predicated on known tenant demands and requirements. In addition to a speculative estimation of income, the expenses to be deducted from any estimate of income are also speculative because these would depend upon the ultimate uses that would be made of the property and the nature and extent of the actual renovations.
After an estimated value is achieved by using hypothetical income and hypothetical expenses which are both subject to many variables, one would have to deduct the actual cost of altering the facility based on its actual rehabilitation cost determined for and to the actual specifications of the individual tenants, not based on some conjectural conceptual plan.
Aside from all of the speculation that is inherent in the income analysis of defendant’s appraisal expert, there is also a major omission: Polayes did not allow or account for entrepreneurial profit.
As explained by this court in Lawrence Assocs. v. Lawrence Tp., 5 N.J.Tax 481, 534-535 (Tax Ct.1983), entrepreneurial profit is “adequate compensation to the entrepreneur to induce him to organize the entire project.” Without it, no prudent investor would undertake a development project with all of inherent risks and hazards without seeking adequate remuneration. See also The Appraisal of Real Estate, supra at 359-361, 373. In this case the value projected by Polayes’ income approach makes no allowance for the entrepreneur. It assumes that an investor will buy the property in its undeveloped state for a price that would be equivalent to its value after development less only the projected physical costs of retrofit. No one would be motivated to pay that price. A prudent investor would, of necessity, deduct from the price or value after conversion, not only the rehabilitation costs, but also a profit — suitable to cover the business risks. See American Institute of Real Estate Appraisers, Subdivision Analysis (1978) in which the authors *199indicate that one specific unacceptable technique in a subdivision analysis is the “[fjailure to recognize and account properly for developer’s profit separately from return to or value of land,” id. at 2, and one of the most common errors “is failure to allow for the full influence of the land developer’s profit motive.” Id. at 4. As noted in Subdivision Analysis, the typical developer depends on a reasonable profit as a reward for the risks inherent in, and efforts required by, the project. Ibid.; see also Boykin, “Developmental Method of Land Appraisal,” The Appraisal Journal (April 1976) at 191 (unless profit is recognized as a deduction the present worth of the property will be overstated).
In this case, the failure to account for entrepreneurial profit by defendant’s appraisal expert renders an otherwise speculative estimate of value totally deficient.20 I therefore reject it. Compare Simmons Co. v. Linden, 190 N.J.Super. 448, 451, 464 A.2d 300 (App.Div.1983) in which the Appellate Division criticized the use of a multi-tenant income approach in determining the value of an industrial complex currently in use as a manufacturing center.
As previously indicated, defendant filed an appeal from the 1983 original assessment with the Middlesex County Board of Taxation. The county board increased that assessment. Inasmuch as defendant was not satisfied with the extent of the increase it filed an appeal with this court and plaintiff filed a counterclaim seeking a reduction. For the tax years 1984 and 1985 both parties filed direct appeals with this court seeking adjustments in the original assessment for each tax year. In this proceeding, the county board judgment for the tax year of 1983 and the original assessment for 1984 and 1985 are entitled *200to a presumption of correctness. Pantasote Co. v. Passaic City, 100 N.J. 408, 412, 495 A.2d 1308 (1985); Aetna Life Insurance Co. v. Newark, 10 N.J. 99, 105, 89 A.2d 385 (1952). This presumption can be overcome by cogent evidence which is “definite, positive and certain in quality and quantity.” Pantasote, supra at 413, 495 A.2d 1308.
Ford presented evidence in its affirmative case which, standing alone, was sufficient for this court to determine value. As such, I denied the motion by Edison to dismiss Ford’s complaints at the conclusion of plaintiff's case. Overcoming the presumption of correctness, however, in a Tax Court proceeding does not entitle a taxpayer to a judgment in its favor. The ultimate fact to be determined is the true value of the subject property. The burden of establishing this ultimate fact is met only when a court can conclude by a fair preponderance of the evidence that the entire record presents sufficient credible evidence to establish that the assessments are incorrect and what the correct assessments should be. In this case, each party has the burden of proof on its respective affirmative case. Plaintiff has the burden of proof in seeking a reduction while defendant has the burden of proof in seeking an increase.
As the foregoing readily reveals neither party has sustained its requisite burden of proof. A review of the voluminous record developed in this case at the initial hearings does not permit me to conclude an appropriate valuation by a fair preponderance of the evidence.
I am cognizant of the fact that the parties have expended a great deal of time, money and effort in the preparation and presentation of this local property tax case. I am also mindful of my obligation to utilize my knowledge and expertise, see Glenn Wall Assoc. v. Wall Tp., 99 N.J. 265, 280, 491 A.2d 1247 (1985), in conjunction with the valuation data submitted by the parties in an effort to ascertain an appropriate value for the property in question. An appropriate value and tax assess*201ment, however, can only be deduced when there is sufficient substantial and competent evidence in the record to support that determination. See Samuel Hird & Sons, Inc. v. Garfield, supra, 87 N.J.Super. at 74, 208 A.2d 153; cf. Glenwood Realty Co., Inc. v. East Orange, 78 N.J.Super. 67, 70, 187 A.2d 602 (App.Div.1963) (the Appellate Division affirmed the conclusion of the Division of Tax Appeals “that each party had failed to sustain the requisite burden of proof to alter the judgment of the board.”).
In this proceeding each party has exceeded itself in demonstrating the deficiencies in the valuation approaches as advanced by its adversary. I find that each of the valuation methodologies employed in this case has sufficient deficiencies that dictate my rejection of the valuation estimate derived by that method. Furthermore, I am not able, based on the record, to make supportable adjustments to the data submitted to fix an appropriate valuation.
Although, at the outset of this opinion, I indicated that additional proofs would be necessary in order to determine both the character and value of certain items and systems that Edison contends became real property subsequent to the initial hearing, due to the Legislature’s enactment of c. 117 following the trial, because of the determination I have made that now becomes unnecessary. The value conclusions to be produced by such a supplemental trial could only be considered as an addition to (or subtraction from) that value or those values concluded as a result of the initial hearing.
Absent such initial value conclusions, any supplemental hearings, to cover items that should have been considered previously, simply, are unnecessary.
Therefore, the Clerk of the Tax Court is directed to dismiss all complaints and counterclaims filed in this matter for failure to sustain the requisite burden of proof necessary to alter the existing assessments.
One of defendant's experts, Gennaro C. Macri, a certified manufacturing engineer, who had worked for Ford Motor Co. from 1950 to 1980, indicated that the reason for the extensive number of loading docks and amount of rail entering the facility was due to the fact that automobile parts were not manufactured at the subject but rather a final product — an automobile — was assembled at the plant, and thus, there was a need for extensive materials-receiving facilities.
Defendant also maintains, as an alternative argument, that conversion of the Ford facility into a number of multi-tenant industrial uses would produce a higher value for the subject by way of an income approach than would use as a general purpose manufacturing facility for a single user. It is defendant’s *162position that as such, if this court rejects its contention that the present use, Le., automobile assembly, is its highest and best use then conversion to multi-tenant industrial use would produce the highest value or highest and best use for the subject. Defendant's alternative argument is more fully considered in the section of this opinion entitled "Income Approach,” infra.
It is also of interest to note that one of the reasons why Parkhouse selected his first comparable (Volkswagen of America, Inc. to Chrysler Corporation) was the fact that it had the same use as the subject.
As a general matter, valuation principles and techniques utilized in eminent domain valuations should be applicable in local property tax cases inasmuch as *166the objective of both proceedings is to ascertain the fair market value of the property in question.
This Connecticut case involved the condemnation of a building that was being used as a poultry slaughtering house. Relative to use the court stated:
The business of slaughtering poultry is not one which may properly be conducted in every neighborhood. Locations suitable for it are relatively few. The fact that the slaughtering business was established on the property in question was evidence that the property was suitable for that use. It indicated what was an advantageous use for this particular building. The property could well have been found to be worth more to a potential buyer because it had been devoted to that use. Moreover, in view of the difficulty which the owner would be likely to have in finding another location suitable for his operations, he would be reluctant to sell, and that would be a factor in determining the fair market value of the property. [Ibid.; emphasis supplied]
The Supreme Court of Minnesota came to a similar conclusion in McCannel v. Cty. of Hennepin, 301 N.W.2d 910 (Minn.Sup.Ct.1980). In that case the trial court held that airport facilities owned by Northwest Airlines were to be valued by means of a cost approach with no consideration given to the expense that a potential buyer would incur in converting the property to a different use. In its approval of the trial court’s valuation method, the Minnesota Supreme Court explained as follows:
Northwest argues that the trial court’s method of valuing its property as unique property violates the general rule that property should be valued at its market value rather than its intrinsic value. Although the concepts of intrinsic value and unique property are closely parallel in cases such as this, the trial court did value the property by determining its reproduction cost, an accepted method of estimating market value. To state it differently, the trial court determined the value of the property according to its highest and best use as an airport facility without regard to who might own it. The fact that its intrinsic value to Northwest Airlines might be equal to its value to a hypothetical buyer as an airport facility does not render the trial court’s method of valuation invalid, [at 924-925; footnote omitted; emphasis supplied]
It should be noted and emphasized that this approach is not based on a "value-in-use" concept. The inquiry is not with respect to the value of the property to the owner, but rather to its market value at its highest and best use regardless of who might own it.
Simply put — it does not make sense to estimate the value of say, a gasoline station, by sales of gasoline station properties that were converted to another use either at the time of sale or shortly thereafter. The sale price in those instances would not be indicative of the value of a gasoline service station but rather of its value for some alternative use. Any buyer for an alternative use would consider the cost of conversion or remodeling to meet his specific requirements in any price to be paid the seller. See Brick, "Gasoline Service Station Appraisal,” Friedman, Encyclopedia of Real Estate Appraising (3 ed. 1978) at 952; see also Society of Industrial Realtors, Industrial Real Estate (4 ed. 1984) at 499-500 ("... different pieces of industrial real estate may have differing values depending upon the highest and best use of the property as well as the definition of the value to be estimated.").
Parkhouse indicated on direct examination that the “use to which [a] property was put was not necessarily a [criterion]” in the selection of a comparable property.
Parkhouse defined replacement cost new as used in his analysis to mean "a replacement of the same space and size configuration of a building, using current materials which is in conformance with the Marshall Valuation Service."
It should be noted that the record reveals that the square foot or lineal foot costs were derived from the R.S. Means service but does not show the method by which these costs per square foot or lineal foot were derived.
Not noted at the trial of this matter was an error in Spence’s multiplication for the item designated as “Boiler House (# 4) + Boilers” as shown in defendant’s appraisal report at 58. The calculation provides for $476,240 as the result when multiplying 8,156 square feet by $40 a square foot. The correct result is $326,240 or $150,000 less than the figure shown in the report. This error does not affect the ultimate conclusion in this case.
This is not to say that the Marshall manual cannot be used to estimate the cost of a building in excess of 400,000 square feet, but rather to say that since the Marshall Valuation Service provides the standard, it should provide a suitable explanation as to whether, and how, an area multiplier can be extrapolated beyond the limitations of its area perimeter table.
The appraisal experts for both parties were in agreement that an income approach could not be employed for a single-user tenant of the Ford plant in an unsubdivided state because there was insufficient market data.
What is curious about Polayes' value analysis is that if the sales of these industrial facilities for multi-tenant conversions demonstrated a market for that particular use why then did not Polayes develop, or at least explain why he did not develop, a direct sales comparison approach to value using the recited *193sales in light of the fact that these plants were physically similar and, more important, were economically similar since the highest and best use, i.e., conversion to multi-tenancy, was the same? See Westinghouse Elec. Corp. v. Bloomfield Tp., supra, 9 N.J.Tax at 101.
Although plaintiff eschewed an income approach analysis because of its speculative nature, it produced an architect, Carl F. Walter, III, who testified as to the suitability of the Ford facility for alternative uses. His projected cost for conversion of the subject plant to multi-tenant use was $5,151,269 based, of course, on his specific plan for retrofit to multi-tenant occupancy.
See The Appraisal of Real Estate, supra at 489-491 for an explanation of a discounted cash flow analysis.
"In appraising, [an efficiency ratio is] the ratio between a building’s net rentable area, i.e., the space used and occupied exclusively by tenants, and its gross area, which includes the building’s core.” American Institute of Real Estate Appraisers, The Dictionary of Real Estate Appraisal (1984) at 104.
As Spence put it:
"... you cannot go in and start all this work until you have somebody committed for it.”
Even defendant’s brief reflects the speculative nature of the income approach of its experts. Defendant states, ”[t]he hypothetical nature of the exercise conducted by the experts for Edison is not so speculative as to render its results weightless.” Emphasis supplied.
I am aware of the adoption of a figure of 10% for entrepreneurial profit by the Tax Court in Twin Oaks Assoc. & Health Resources of Morritown, Inc. v. Morristown, 9 N.J.Tax 386, 397 (Tax Ct.1987), however, I find nothing in the record of this case that permits such a conclusion here. To adopt such a percentage, absent some evidence, would be pure conjecture.