BALMER, J.
**247The issue before us in this direct appeal from the Oregon Tax Court involves the proper valuation, for property tax purposes, of a shopping center that did not have an *299anchor tenant on the assessment date. The Tax Court accepted taxpayer's valuation that significantly decreased the value of the shopping center because it was missing an anchor tenant and was more than 50 percent vacant on the relevant date. Powell Street I LLC v. Dept. of Rev. , 22 OTR 423, 2017 WL 3310655 (2017). On appeal, the Department of Revenue contends that the Tax Court erred; according to the department, the shopping center was required to be valued the same as a shopping center that did have an anchor tenant and was only 8-10 percent vacant. For the reasons that follow, we reject the department's argument and affirm the Tax Court's judgment.
OVERVIEW OF LAW
We begin with an overview of several familiar property tax concepts: the definition of real market value, the requirement that the tax be assessed on the "fee simple interest" in the property, and the valuation methods used by appraisers to determine real market value.
In general, the calculation of the property tax levied on a particular property begins with the property's real market value (RMV). See ORS 308.232 ; Dept. of Rev. v. River's Edge Investments, LLC , 359 Or. 822, 825, 377 P.3d 540 (2016) (both so providing). The real market value is the amount that a hypothetical buyer would pay to a hypothetical seller on the assessment date. See Or. Const., Art XI, § 11 (11)(a)(A) (defining real market value as "the amount in cash that could reasonably be expected to be paid by an informed buyer to an informed seller, each acting without compulsion in an arm's length transaction occurring as of the assessment date for the tax year"); see also ORS 308.205(1) (using terms "informed buyer" and "informed seller"); ORS 308.205(2)(a) (referring to "amount a typical seller would accept or the amount a typical buyer would offer that could reasonably be expected by a seller of property"). The real market value is derived from the "highest and best use" of the property-the most profitable use, for which a buyer would be expected **248to pay the highest price. See Hewlett-Packard Co. v. Benton County Assessor , 357 Or. 598, 602, 356 P.3d 70 (2015) (so explaining).
When the property is subject to leases (as is the case for the shopping center here), the value for property tax purposes may differ from the price that the owner actually might receive for the property. That is because the property tax is assessed on the fee simple interest in the property, which is the value of all interests in the property, including those of the owner (ordinarily the lessor) and any lessees.1 That rule was relied upon and explained in Swan Lake Mldg. Co. v. Dept. of Rev. , 257 Or. 622, 478 P.2d 393 (1970), reh'g den , 257 Or. 622, 480 P.2d 713 (1971). In that case, substantial portions of the property being valued were subject to leases with remaining terms as long as 30 years. The question was whether the leases should affect the valuation. The taxpayer argued that they should because any buyer of the property would take the property subject to those leases.
This court disagreed. It held that the actual terms of those leases should not be used in calculating the value because the value of all of the interests in the property-including the lessee's interest-is to be taxed against the owner:
"In fixing the true cash value of land for property tax purposes[,] the effect of existing *300leases on the value to the owner is disregarded. The basis for such a principle is that the tax is levied upon the land and is a tax upon all the interests **249into which the land might be divided. Admittedly, a lease might decrease the price which the owner might receive; however, the tax is not merely upon the owner's interest; the tax is upon all the interests in the land, including the leasehold interest. This is so because of the corollary principle that taxes are assessed only against the one having title[.]"
257 Or. at 625, 478 P.2d 393. If a property owner leased a property at below-market rates, for example, then the value of the owner's interest in the property is lower, but at the same time the lessee's leasehold interest is that much more valuable and offsets the decrease:
"If the rent reserved in the lease is less than the property is capable of producing, the lessee's interest is more valuable and it is the entire group of interests in the property, lessor's and lessee's, that is valued."
257 Or. at 629, 478 P.2d 393 (denying rehearing).
A "fee simple valuation" thus may deviate from the ordinary concept of real market value. If the property is leased at nonmarket rates, and if the lease will not terminate on sale of the property, then the price that the owner could obtain in the market could differ from the value used for property tax purposes. See Joan M. Youngman, Defining and Valuing the Base of the Property Tax , 58 Wash. L. Rev. 713, 716 (1983) (fee simple valuation allows " 'value' to exceed the amount which the holder of a restricted interest could command in an actual sale").
Turning to valuation under Oregon law more generally, and broadly speaking, an appraiser determines the real market value of a property by considering three different approaches to valuation: the cost approach, the comparable sales approach, and the income approach. OAR 150-308-0240(2)(a) ; see River's Edge , 359 Or. at 827, 377 P.3d 540 ; Hewlett-Packard , 357 Or. at 603, 356 P.3d 70 (explaining approaches). The appraiser is not required to use all three approaches, but the appraiser must consider them. OAR 150-308-0240(2)(a) ; see River's Edge , 359 Or. at 827, 377 P.3d 540 ; Hewlett-Packard , 357 Or. at 603, 356 P.3d 70 (so stating).
We have outlined the three approaches in River's Edge and other cases. The comparable sales approach estimates the value of the property by extrapolating from the **250prices paid for similar properties in the area. The cost approach estimates the cost to recreate an equivalent property. If valuing a factory, for example, the appraiser would estimate the cost to build a similar factory. The income approach assumes that the hypothetical buyer would pay the present value of the stream of income that the property will generate. Accordingly, the income approach calculates that stream of future income, then discounts it by an appropriate capitalization rate. Seneca Sustainable Energy, LLC v. Dept. of Rev. , 363 Or. 782, 799-800, 429 P.3d 360 (2018).
If more than one approach is used, the appraiser will then reconcile the values obtained from the various approaches to obtain a final valuation. River's Edge , 359 Or. at 827, 377 P.3d 540 ; Hewlett-Packard , 357 Or. at 603, 356 P.3d 70. The weight to be given the different approaches is a question of fact that depends on the record developed in the case. Pacific Power & Light Co. v. Dept. of Rev. , 286 Or. 529, 533, 596 P.2d 912 (1979) ; Brooks Resources Corp. v. Dept. of Revenue , 286 Or. 499, 503-04, 595 P.2d 1358 (1979).
FACTS AND PROCEEDINGS BELOW
As noted, this case involves determining the real market value of a shopping center. The shopping center has one anchor tenant space, plus some 21 other spaces of various sizes. The building's total square footage is 118,000 square feet. The anchor tenant space is 54,000 square feet of that.
The tax year at issue is 2014-15, so the assessment date-the date as of which the property is valued-is January 1, 2014. See ORS 308.007 (defining assessment date and explaining relationship to tax year). On that date, the anchor space was vacant. A discount grocer previously had occupied the anchor *301space for 25 years, but it vacated the space in January 2013 and ended lease payments in May 2013. Three other spaces also were vacant on the assessment date. Those additional vacancies totaled 11,000 square feet. The Tax Court found that the property had a 51 percent vacancy rate on the assessment date, Powell Street I , 22 OTR at 426, and that finding is not at issue on appeal. The Tax Court also noted testimony that tenants in the other spaces had lease clauses that modified the lease terms if **251the anchor space was vacant. Such contractual provisions are typical of the marketplace and can include lower rent payments or even cancellation of the lease. 22 OTR at 425.
The Multnomah County Assessor initially found a real market value of $14.7 million. Taxpayer appealed to the Multnomah County Board of Property Tax Appeals, which found the same value. Taxpayer then appealed to the Tax Court, and the Department of Revenue intervened as a defendant.2 The parties petitioned for the matter to be specially designated to go directly to the Regular Division of the Tax Court, under TCR 1C, and that petition was granted.
At trial before the Tax Court, both taxpayer and the department presented appraisals, and their appraisers testified in court. The appraisers agreed that the fee simple estate is the basis for the valuation, which includes, when leased property is involved, "the lessor's interest (i.e. , the leased fee) and the lessee's interest (i.e. , the leasehold)." Appraisal Institute, The Appraisal of Real Estate 441 (14th ed. 2013). The two appraisers also reached similar conclusions on a number of facts that went into their appraisals.3 Both concurred that the highest and best use of the property was its current use as a shopping center. Both also reached roughly similar conclusions regarding the market rents per square foot of shopping center space: The appraisers agreed on a value of $9 per square foot for the anchor space, though they differed somewhat on the other spaces. Both appraisers also reached similar conclusions regarding what would be the market vacancy rate for a similar shopping center. Taxpayer's appraiser concluded that the market vacancy rate would be 10 percent. The department's appraiser concluded that the market vacancy rate would be eight percent.
The primary dispute between the appraisers was whether this shopping center should be valued as if it did **252have a market vacancy rate. Both appraisers referred to that concept as whether the property should be valued as being at "stabilized" or "non-stabilized" occupancy.4 As we discuss in more detail below, taxpayer's appraiser concluded-and the Tax Court found-that the property at issue here was "non-stabilized" because it was missing its anchor tenant and the property's overall occupancy rate of less than 50 percent represented a major deviation from market vacancy rates. In contrast, the department's appraiser concluded that the property should be considered "stabilized."
As will be seen, that dispute affected the appraisers' different applications of both the comparable sales approach and the income approach to determining real market value. With respect to comparable sales, taxpayer's appraiser examined comparable "non-stabilized" shopping malls-those exhibiting high *302vacancy rates-while the department's appraiser compared the subject property to "stabilized" shopping malls, which had market vacancy rates. Similarly, in applying the income approach, taxpayer's appraiser first estimated rental income from the shopping center using market vacancy rates and market rents, and then deducted $4.71 million in "projected stabilization costs" to reflect the costs that any owner would need to incur to bring the shopping mall to "stabilized" occupancy. Although the department's appraiser also based his income approach valuation on the assumptions of market vacancy rates and market rents, consistent with his view that the property should be viewed as stabilized, he did not make any deduction for the costs that needed to be incurred to achieve a market occupancy rate.
We return to the underlying dispute over whether the property was stabilized or non-stabilized. Taxpayer's **253appraiser explained that, because the property lacked an anchor tenant and was non-stabilized, the only potential buyers for the property would be "value add" buyers. Those are "a somewhat narrow range of opportunistic investors" who are willing and able to incur substantial risk and effort to bring the property back up to stabilized lease rates. Because of the money, time, effort, and overall risk involved, "value add" buyers will expect a higher return on their investment, and so they will pay substantially less for the property. The department's appraiser disagreed, contending that the property was stabilized. He considered tenant turnover as normal market behavior, and he noted that (as of the assessment date) the anchor tenant spot had only been vacant for a year, a relatively short period for a large commercial space. In effect, the department's appraiser argued that the anchor tenant vacancy here should be treated as a passing fluctuation in occupancy. In his view, only if an anchor tenant spot had been vacant for a substantial period would he consider the property to be non-stabilized. The department's appraiser agreed that, if a property was non-stabilized, then the sorts of adjustments to value made by taxpayer's appraiser-those costs necessary to bring the property back up to stabilized occupancy rates-would be appropriate.
The appraisers' factual disagreement over whether the property was stabilized ultimately led the two appraisers to different conclusions regarding value. As noted, to determine real market value, taxpayer's appraiser used both the comparable sales approach and the income approach. (He also considered, and rejected, the cost approach, but that decision is not at issue here.) His comparable sales approach, which considered "non-stabilized" shopping malls around the state, led him to find a value of $10.01 million. Using the income approach, he initially calculated a real market value of $14.95 million using market rents for the shopping center at market vacancy rates. That, however, was the value for the property if it were stabilized. Because the property was not stabilized, in his view, he concluded that the initial value should be reduced by $4.71 million. Those costs consisted primarily of tenant improvements that would be necessary to fill the anchor tenant space, **254leasing commissions, rent concessions, and a turn-around profit incentive reflecting the risks associated with bringing the property back to stabilized occupancy. That amount did not include-because they were not known as of the valuation date-additional substantial costs that likely would be imposed by the City of Portland for parking lot landscaping, bicycle parking, ADA upgrades, and storm water waste improvements, all of which would be triggered by the necessary tenant improvements. Those expenses were not specific to this particular owner of the property-they would have been required of any owner who wanted to increase the occupancy rate towards the market level. Accordingly, using the income approach, the appraiser found a value of $10.25 million. Taxpayer's appraiser then reconciled the two values, giving equal weights to the income approach and the comparable sales approach. His final valuation was $10.13 million.
The department's appraiser used all three approaches. Regarding the income approach, he calculated the value of the property by determining a projected income stream for the property using market vacancy rates and market rents, and then discounting it to present value, which resulted in a value of *303$17.3 million. Because he categorized the property as stabilized, his valuation did not reflect any costs that had to be incurred to bring the property to the market occupancy level, that is, to "stabilize" the property's occupancy. In applying the comparable sales approach, the department's appraiser rejected taxpayer's use of non-stabilized property sales, and instead looked at more local sales of shopping malls-all of which were stabilized. Based on those comparables, he found a value of $18.8 million. The department's appraiser also did a cost approach, which led him to a value of $19.8 million. He then reconciled the differing values based on the three approaches. He placed primary emphasis on the income approach and secondary emphasis on the comparable sales and cost approaches. His final valuation conclusion was $17.5 million.
The Tax Court ultimately accepted the $10.13 million value found by taxpayer's appraiser. The court agreed with taxpayer that the substantial vacancy and missing **255anchor tenant significantly affected the market for the property. See 22 OTR at 435. The court rejected the department's argument that the property should be considered "stabilized," which was the premise for the department's valuation. See 22 OTR at 434-35, 438, 440-41. The Tax Court also rejected the department's argument that taxpayer had not valued the fee simple estate, as required by Swan Lake . The court noted that taxpayer had used market rents in the income approach, but then took a "substantiated" deduction for the costs that would be needed to stabilize the property. 22 OTR at 435. The propriety of that deduction, the court concluded, had nothing to do with valuing the fee simple estate or using market rents. Id . at 435-36. In accordance with that ruling, the Tax Court entered a judgment that lowered the real market value of the property to $10.13 million, as of January 1, 2014.
The department (and the Multnomah County Assessor) appealed to this court.
ANALYSIS
We review the Tax Court's opinion for errors of law or for lack of substantial evidence in the record to support the court's decision. ORS 305.445.
The department contends that the Tax Court erroneously valued the property based on taxpayer's individualized ownership of the property-a valuation constrained by the high vacancy rate-rather than determining the market value of the property itself. As noted, the real market value is defined by looking at what a hypothetical buyer would pay a hypothetical seller. See Or. Const., Art XI, § 11 (11)(a)(A); ORS 308.205(1). The department, focusing on the "hypothetical seller" part of the definition, contends that valuations of leased property always must use market rents and market vacancy rates, even when a property's actual rent or vacancy rates are substantially different.
Taxpayer's appraiser did use both market rents and market vacancy rates, as required by Swan Lake , 257 Or. 622, 478 P.2d 393, to calculate the stabilized value of the property, and the department does not dispute that. The department **256maintains, however, that deducting stabilization costs from that value functionally leads back to a value peculiar to the particular owner, rather than the value to the hypothetical owner required by statute. Specifically, the department asserts that the substantial vacancy rate and missing anchor tenant here are the peculiar characteristics of the owner-taxpayer-and not of the property itself. As the department puts it, this court should "remove[ ] any consideration of the skill and luck of a particular owner in negotiating contracts, and keep[ ] the focus rightfully on the property, not the owner." The department adds, "A typical owner would not have had the same abnormally high vacancy that Powell Street had." The use of market rents and market vacancy rates consistent with Swan Lake , the department argues, would have removed the owner's characteristics from the valuation equation.
The department is largely correct about the underlying legal principles. Both the statutory and constitutional definitions of "real market value" are based on what "an informed buyer" would pay to "an informed seller" on the assessment date.
*304Or. Const., Art XI, § 11 (11)(a)(A); ORS 308.205(1). It is undisputed that both "an informed buyer" and "an informed seller" refer to hypothetical buyers and sellers. See River's Edge , 359 Or. at 825, 377 P.3d 540 ; Boise Cascade Corp. v. Dept. of Rev. , 12 OTR 263, 266, 1991 WL 434542 (1991) (to estimate real market value, "the appraiser attempts to replicate the positions of a hypothetical buyer and a hypothetical seller"). The peculiarities of the actual owner of the property are not considered. Boise Cascade , 12 OTR at 266 (in determining real market value, appraiser "ignore[s] the specific circumstances or peculiarities associated with specific owners of property"). The buyer and seller must be hypothesized to be "typical." See ORS 308.205(2)(a) (department's rules for determining real market value must accord with principle that real market value will look to "[t]he amount a typical seller would accept or the amount a typical buyer would offer that could reasonably be expected by a seller of property").
But the department then takes another step that is not supported by the record in this case. Its conclusion that the vacancy rates here are a characteristic of taxpayer , **257rather than of the property, appears to be based on the following syllogism:
• a property's actual vacancy rate depends on the owner's skill in negotiating leases;
• the owner's skill in negotiating leases is a characteristic of the owner and not the property;
• therefore, valuing the property based on the actual vacancy rate is impermissibly valuing a characteristic of the owner.
Beyond the mention of the owner's negotiating "skill," the department does not otherwise discuss what test, factors, or criteria would separate characteristics of the property from characteristics personal to the owner.
We accept for purposes of argument the department's premise that, if an owner has an unusual skill level either above or below that of a typical owner, then that might be a personal characteristic of the owner rather than a characteristic of the property, and thus should not to be used in determining real market value, which is based on a transaction between a hypothetical seller and a hypothetical buyer. Using market rents, as this court did in Swan Lake, has the effect of eliminating from the valuation equation the owner's personal skill in leasing the property.5 Using market rents means that the real market value will not be affected by a property owner whose poor negotiating skills led to below-market rents or whose extraordinary negotiating skills led to above-market rents.
We also will assume that the same logic applies to vacancy rates, which is the primary issue in dispute here. If the owner of a shopping center has driven out tenants by mismanagement, then that mismanagement is a characteristic **258of the owner and generally should not affect the property's real market value for ad valorem tax purposes.6
In this case, however, the Tax Court rejected the factual underpinning on which the department's syllogism rested: that the anchor tenant vacancy was, in fact, a characteristic of taxpayer and not of the property. The department had not presented evidence that either the departure of the prior anchor tenant, or the fact that a new anchor tenant had not filled the space by the assessment date, had anything to do with the personal characteristics of this taxpayer. Rather, the Tax Court found that:
"[N]othing in the record suggests that the approvals needed, the nonconforming upgrades *305required, or estimated funds and time necessary to place an interested anchor tenant into the subject property are unusual or outside market norms. They cannot accordingly be deemed to be attributable to poor property management, which would go to actions or skills of the particular owner and would accordingly not be relevant for purposes of property assessment and taxation."
22 OTR at 437.
We agree. To the extent that the department disputes the Tax Court's factual findings on this issue, we conclude that they were supported by substantial evidence that the former operator of the anchor tenant space vacated for reasons that had nothing to do with taxpayer. The record shows that the operator "was 70, he was ill, [and] he had another store that was struggling that had closed and he still had to pay on it." The department does not point to contrary evidence, nor does it argue on appeal that a typical owner would have been able to fill the anchor tenant space by the assessment date, even though the owner here did not. Indeed, while the anchor tenant space had been vacant 12 months at the assessment date, both the department's and the taxpayer's appraisers were in general agreement that that period was well within market norms (the department's **259appraiser testified that it could take up to 18 months to find a new anchor tenant; taxpayer's appraiser estimated 24 months). Thus, we must accept the Tax Court's finding that the property value would have been burdened by a substantial vacancy regardless of the seller's identity. The vacancy was not a characteristic of this particular owner's skill level; it was a characteristic of the property.
The department makes a separate legal argument based on First Interstate Bank v. Dept. of Rev. , 306 Or. 450, 760 P.2d 880 (1988). It contends that the value that the Tax Court ascribed to the property is the investment value for this particular taxpayer, rather than its value to a hypothetical buyer, a result that the department asserts is prohibited by First Interstate . We turn to that decision.
First Interstate involved the valuation of multiple lots in a subdivision that had a single owner. See 306 Or. at 452, 760 P.2d 880. The department had sent a memorandum to county tax assessors approving valuation using a "developer's discount." Id . at 453-54, 760 P.2d 880. The taxpayer sought to take advantage of that concept, arguing for a "valuation method [that] reduces the market price of the properties by a rate of return based on expected profit, taking into account the expected time necessary to sell the lots." Id . at 454, 760 P.2d 880 (footnote omitted). This court rejected the taxpayer's argument. Id . at 455, 760 P.2d 880.
The department reads First Interstate to prohibit a valuation methodology that takes into account the property owner's cost to sell or improve properties, on the ground that doing so would result in an investment value for that particular owner rather than the real market value of the property, which assumes a hypothetical seller of the property. The department relies on the following quote from that opinion:
"Reduction by this method results in a determination of the properties' value to the current owner or their value as an investment. This is not the market value, which is the price that each property would receive on the open market. While in certain circumstances the value to the owner might equal the market value, the value to the owner cannot be equated with the market value."
306 Or. at 454, 760 P.2d 880 (citation omitted).
**260First Interstate does not support the broad rule of law for which the department advocates. In First Interstate , the court held that the unit of property being assessed was not the subdivision as a whole, as the taxpayer had argued, but each individual tax lot within the subdivision. Id . at 453, 760 P.2d 880. The highest and best use for each of those lots was for "construction of a single-family residence." Id . at 454, 760 P.2d 880. The "developer's discount," then, was problematic for two reasons. First, the discount did not value the individual lots; it valued the lots as a single package. Id . at 454, 760 P.2d 880. Second, the discount "does not assess the *306value of the properties if put to their highest and best use, but reduces their value to arrive at the value of the properties considered as an investment. Investment is not the highest and best use of the properties." Id . at 455, 760 P.2d 880. The court added that, if each lot's highest and best use had been as "part of a group of lots," then "it would be appropriate to assess that lot based on its value as part of a group." Id . at 453 n. 2, 760 P.2d 880.
The factors on which the court relied in First Interstate , simply are not present here. In contrast to that case, the property here is being valued at its highest and best use; the appropriate unit is being valued; and the non-stabilized occupancy of this property would impair its value to any seller, not simply to this taxpayer. As we have explained, the Tax Court found that the anchor tenant vacancy was a condition of the property and not of this particular taxpayer, and that the vacancy would burden the property even if owned by a hypothetical seller.
For the reasons discussed above, we reject the department's argument that the Tax Court committed legal error by failing to correctly apply this court's decisions in Swan Lake and First Interstate.
We return to the underlying question of the real market value of the property at issue-what a hypothetical buyer would pay a hypothetical seller for the property on the assessment date. Here, that question reduces to the central dispute between the parties over whether the property should be considered stabilized or non-stabilized. With respect to appraisal methodologies or other means of determining value, this court has held that, in the absence of a **261statute or regulation that requires a particular valuation method, the determination of the appropriate methodology is a question of fact. Bylund v. Dep't of Revenue , 292 Or. 582, 585, 641 P.2d 577 (1982) ("methods of accounting or valuation, unless these are prescribed by law or regulation," are determined "by evidence in the record"); see also Bend Millwork v. Dept. of Revenue , 285 Or. 577, 586, 592 P.2d 986 (1979) (judicial adoption of a method of valuation in a particular case "was not the statement of a 'legal principle' or a rule of law," but "merely a finding of fact on the record before the court that the method should be applied in that case and a disposition of the factual controversy by application of that appraisal method"). The Tax Court recognized that principle here, stating that, "unless otherwise prescribed by law or regulation, methods of valuation are analyzed in light of the evidence introduced by the parties." Powell Street I , 22 OTR at 434 ; see also id. at 436.
No statute or regulation prescribes whether or when a leased property should be considered stabilized or non-stabilized for valuation purposes. The Tax Court correctly treated that valuation methodology issue as one to be resolved based on the evidence. Taxpayer presented evidence that, because the property was non-stabilized, potential buyers of the property were limited to "a somewhat narrow range of opportunistic investors" who would be able and willing to incur the time, effort, and risk to bring the property back to stability. That is the set of "informed buyer[s]" whose purchase prices will determine real market value for property tax purposes. See Or. Const., Art XI, § 11 (11)(a)(A); ORS 308.205(1). Taxpayer presented evidence that such buyers would only buy the property at a substantial discount that would allow them to recover their costs and compensate them for market risk. Those findings were supported by substantial evidence, and the department does not contend otherwise. The Tax Court resolved the factual dispute over whether it was appropriate to consider stabilization in valuing the property, and, if so, whether the property was stabilized, by agreeing with taxpayer's appraiser as to the appropriate methodology and also with that appraiser's calculations. That finding, too, and the Tax Court's ultimate conclusion on value, was supported by substantial evidence **262in the record. Accordingly, the Tax Court did not err in finding that the real market value of the property was $10.13 million, as of January 1, 2014.
The judgment of the Tax Court is affirmed.