The taxpayer appeals under RSA 76:16-a, V (Supp. 1985) from an order of the New Hampshire Board of Tax and Land Appeals partially denying its application for abatements of taxes for the tax years 1983 and 1984. The taxpayer assigns error to the board’s failure to make adequate findings of fact, to its use of the 1983 equalization rate to calculate an assessment for tax purposes for 1984 as well as for 1983, and to its treatment of two properties as one for valuation purposes. On the first issue, we reverse the board’s order and remand, but we find no merit in the second and third assignments of error.
At a foreclosure sale in 1983, the taxpayer paid $2.6 million for land and two office buildings in Concord, know as 6 and 8 Loudon Road. The taxpayer allocated $2,075,000 to the land and building at 6 Loudon Road, and the remaining $525,000 to the other property. The city assessed 6 Loudon Road at $1,727,100 for 1983, and at $1,476,440 for 1984. It assessed 8 Loudon Road at $526,900 for each year.
Although the taxpayer accepted the assessment placed on 8 Lou-don Road, it sought a reduction of the assessment on number 6 by request for abatement, to which the city responded by offering to reduce the assessment to $1,378,850 for each year. In the subsequent proceeding before the board, the city defended this figure by reference to the cost, market and income approaches to valuation, with emphasis on the income approach in light of the building’s ability to generate cash flow for its owner. The taxpayer contested the figure by presenting evidence addressed to the income approach.
The parties’ evidence diverged principally on the issues of the applicable economic rent, see Demoulas v. Town of Salem, 116 N.H. 775, 780-82, 367 A.2d 588, 592-93 (1976), and the vacancy rate in the market area. The board made the following determination on these issues:
*626“In regard to the income approach, the Board rules the proper measure to be used is economic, not contract rent, . . . and taxes are properly an adjustment to the capitalization rate, not an expense, in tax abatement cases. The Board finds the City’s income figures and calculations to be credible. The Board notes that its expense figures and capitalization rate are close to those of the Taxpayer’s. The Board finds the City’s figures on gross income and the vacancy factor to be more creditable than the Taxpayer’s in light of the purchase price, comparable sales, and the cost approach to valuation.”
(Citation omitted.) The board accepted the city’s treatment of 6 and 8 Loudon Road as one parcel for purposes of its valuation, and concluded that a
“market value of $3,025,000 on April 1, 1983, [is not] unreasonable in light of the foreclosure sale price of $2,600,000 for the total property. The City’s figure is supported by credible evidence in regard to the market and cost approaches to value. . . . The Board concurs with the conclusion of value presented to the Board by the City at the hearing held April 8, 1985.”
The board then applied the stipulated 1983 equalization rate of sixty-three percent to this market value to derive an assessment value proportional to the assessed value of other property in the city. See Milford Props., Inc. v. Town of Milford, 119 N.H. 165, 167-68, 400 A.2d 41, 43 (1979). It employed the resulting figure of $1,905,750 as the tax assessment for both 1983 and 1984.
As we take up the taxpayer’s first assignment of error, there is no need for extended analysis to demonstrate that the board failed to make adequate findings of fact on the points in issue. Although the board devoted some time to summarizing evidence presented by the contending parties and describing their opposing views, it made virtually no specific findings in support of its ultimate conclusions.
On the issue of the vacancy rate for commercial buildings in Concord, for example, the taxpayer offered evidence that the rate was fifty percent or higher; the city’s evidence indicated it was as low as two to five percent, although the city apparently used fifteen percent in its actual calculations. The board’s closest approach to resolving this factual dispute was the general conclusion that “the City’s figures” were “more creditable ... in light of the purchase price, comparable sales, and the cost approach to valuation.” This is not a finding of fact; it is a conclusory statement that reveals very *627little. Although the purchase price and comparable sales may well tend to confirm one vacancy rate figure as against another, the board makes no specific finding to justify the broad claim and does not even state which of the city’s figures they are supposed to confirm. Presumably the board found fifteen percent to be correct, although the evidentiary basis for selecting that figure is not obvious. Rather, we are left to infer, or guess about, what the board found, and the board is left without the self-discipline imposed by the process of making explicit findings.
Thus, the opinion below does not meet the requirement to make “specific, although not excessively detailed, basic findings in support of [its] ultimate conclusions,” Appeal of Portsmouth Trust Co., 120 N.H. 753, 759, 423 A.2d 603, 607 (1980), an obligation now imposed by statute. RSA 541-A:20 (Supp. 1985). As a consequence, it is either difficult or impossible for us to tell what facts found by the board support its conclusions that the city’s revised assessment was proper and the limited abatement was warranted. See Vickerry Realty Co. Trust v. City of Nashua, 116 N.H. 536, 538-39, 364 A.2d 626, 629 (1976). Accordingly, the order will be vacated and the petition remanded for hearing de novo.
Although this ruling disposes of the appeal, the two remaining issues will doubtless arise again at the new hearing, and we will therefore deal with them briefly. The taxpayer claims that the board erred in calculating the 1984 tax assessment by employing the 1983 equalization rate of sixty-three percent, as against the stipulated 1984 rate of fifty-nine percent. There was no error, however, because the board applied the 1983 rate to the 1983 market value in calculating an assessment for each year. Use of the higher 1983 rate for 1984 would have led to error only if the board had applied the 1983 rate to a 1984 market value reflecting inflation in real estate values during the year or some change in the condition of the property that otherwise affected its value.
Finally, the taxpayer challenges the board’s treatment of 6 and 8 Loudon Road as one parcel. The taxpayer recognizes that it is not entitled to an abatement unless the “aggregate valuation placed on all of. [its] property is unfavorably disproportionate to the [general] assessment of property ... in the town,” Appeal of Town of Sunapee, 126 N.H. 214, 217, 489 A.2d 153, 155 (1985), but it argues that this rule does not authorize a town or the board to assess all of the taxpayer’s real estate as if it were one parcel. This is true, but it is equally true that the board is not necessarily precluded from aggregating adjoining parcels for unitary assessment. Fearon v. *628Town of Amherst, 116 N.H. 392, 393-94, 360 A.2d 127, 128 (1976) (citing RSA 75:9).
“There is no hard and fast rule that can be applied universally to guide assessors in determining whether [adjoining] parcels of land are to be assessed separately or together. . . . [N]o single factor is decisive of the issue.” Lenox v. Oglesby, 311 Mass. 269, 271, 41 N.E.2d 45, 46-47 (1942), quoted in Fearon v. Town of Amherst, supra at 393-94, 360 A.2d at 128. In such a state of the law, there must be good reasons to support a unitary assessment. The following suffice in this case: the properties were historically part of one large parcel and were conveyed as one parcel to the preceding owners, against whom the bank foreclosed. Although the preceding owners treated the properties as two units, and the city has accordingly prepared separate tax bills for two units, there was evidence that the zoning ordinance would legally preclude subdivision into two parcels.
Reversed and remanded for hearing de novo.
All concurred.