CCP Crestview 1505 LLC v. Multnomah County Assessor

IN THE OREGON TAX COURT MAGISTRATE DIVISION Property Tax

CCP CRESTVIEW 1505 LLC, ) ) Plaintiff, ) TC-MD 220292N ) v. ) ) MULTNOMAH COUNTY ASSESSOR, ) ) Defendant. ) DECISION ) Plaintiff appealed the real market value of property identified as Account R328941 for

the 2021-22 tax year. A trial was held via Webex on July 12, 2023. Alex C. Robinson, an Oregon attorney, appeared on behalf of Plaintiff. Jordan Lee (Lee), MAI, testified on behalf of Plaintiff. Carlos A. Rasch, Assistant County Counsel, appeared on behalf of Defendant. Aaron J. Brown (Brown), MAI, testified on behalf of Defendant. Plaintiff’s Exhibit 1 and Defendant’s Exhibit A were received without objection.

I. STATEMENT OF FACTS

The subject property is a skilled nursing facility in Portland. (Ptf’s Ex 1 at 2.) It was built in 1964 with subsequent additions and remodeling in the 1970s and 1980s. (Id.) The site is 2.33 acres with good access and visibility. (Id. at 2, 51.) The subject property improvements total 43,257 square feet and include what Lee described as all the usual amenities for the property type: 65 rooms, offices, a reception desk, lobby, commercial kitchen, dining area, therapy spaces, nurse stations, shower rooms, laundry, storage, maintenance, and more. (Id. at 65, 67.) Lee testified that the subject property was a fairly standard nursing home, similar in age DECISION TC-MD 220292N 1 to others, which were mostly built in the 1960s.1 He characterized the improvements as average condition. (Id.) Brown’s description of the subject property was similar to Lee’s, though he observed the property had a lot of common area, which provides more flexibility in utilization and functionality.

The appraisers agreed that the subject property was licensed for 127 beds but had an operating capacity of 105 beds as of January 1, 2021, due to government restrictions on occupancy.2 (See Ptf’s Ex 1 at 2; Def’s Ex A at 4, 35.) Stabilized occupancy was 77 percent, or 96 beds, due to residents’ preferences for private rather than shared rooms and short-term stays resulting in beds changing frequently. (Id.) The subject’s actual occupancy as of the assessment date was 67 beds due to staffing shortages. (Def’s Ex A at 4, 35.) Other properties in the market were similarly impacted. (See Ptf’s Ex 1 at 38-40; Def’s Ex A at 58-62.) Brown testified that the subject’s mix of private versus shared units was “fairly decent” for its vintage.

As of January 1, 2021, Avamere Living operated the subject property. (Ptf’s Ex 1 at 2.) Avamere operates numerous other facilities in Oregon and Washington that provide independent living, assisted living, memory care, and skilled nursing. (Id.) Brown reviewed the ownership and sale history of the subject property: it sold in October 2006 for $10,619,500 as part of a four-facility portfolio sale; it sold again in April 2011 for $15,104,078; it was transferred in August 2015 with no recorded sale price; and it was allocated a value of $17,385,000 in a 2017 merger, likely an assigned value. (Def’s Ex A at 4-5.) The subject property was subject to a 20-year lease signed February 1, 2022, that is part of a master lease with other properties. (Id.)

1 The appraisers agreed a newer property would have more private rooms. 2 Lee testified that Oregon no longer allows three- and four-person rooms, so that reduced the beds to 105. (See also Def’s Ex A at 4.)

DECISION TC-MD 220292N 2

A. Senior Housing Market Generally, Subject Property Market, Covid-19 Impacts

Lee testified that senior housing properties are grouped by acuity of care, ranging from age-restricted, multi-family properties (least care) to skilled nursing facilities (most care).3 (See Ptf’s Ex 1 at 18-19.) As the acuity of care level increases, more value derives from operations rather than real estate. (Id. at 20 (chart).) Income from a particular facility depends on the census mix of private pay, Medicaid, and Medicare patients. (See, e.g., id. at 88-89.) A potential buyer in this market typically owns and operates facilities at all levels of care; the property types vary in income source and risk. Buyers consider price on a per-bed basis.

Numerous states – including Oregon and Washington – have a “certificate of need” (CON) program that aims to restrain health care facility costs and price inflation by preventing excess capacity. (See Ptf’s Ex 1 at 45.) CON laws require “coordinated planning of new services and facility construction” based on actual need or demand. (See id.) Lee testified that each facility has one certificate, and the certificates are cumbersome to transfer. Brown testified that no new skilled nursing facilities have been built in Oregon in the past 10 years.

The shares of the population 65 years and older and 85 years and older increased from 2010 to 2020. (Ptf’s Ex 1 at 30; Def’s Ex A at 58.) The appraisers agreed that is a generally positive trend for senior housing demand. Lee testified that those broad trends are balanced against the fact that most revenue is from government sources and there are countervailing trends to move patients to lower cost facilities.

Lee testified that the subject property’s “primary market area” is a three-mile radius around the subject property from which most patients are drawn. (See Ptf’s Ex 1 at 25.) He

3 “Continuing care retirement communities” provide a variety of care levels, allowing “older adults to remain in the same community with the same provider, even if their future care needs change (age-in-place).” (Id.)

DECISION TC-MD 220292N 3 found the primary market area adequately supported the subject property’s current use based on average incomes and access to hospitals and transportation. (See id.) Indeed, average incomes in the primary market area exceeded the Portland market overall. (See id.) Lee testified that a nursing home does not necessarily benefit from an affluent location because its primary income is Medicaid – a fixed amount – but wages are based on the market. Brown countered that a more affluent neighborhood tends to lower risk and the related capitalization rate because “people always want to own properties in nice areas.” (See Def’s Ex A at 22 (higher real estate values).)

The appraisers agreed that senior housing, especially skilled nursing, was impacted by the Covid-19 pandemic as of January 1, 2021. (See Ptf’s Ex 1 at 26-28, 33-35; Def’s Ex A at 16, 49-52.) Lee noted direct impacts included reduced occupancy, staffing shortages, increased short-term expenses for personal protective equipment and testing, and increased long-term expenses for staff and insurance.4 (See id.) Prior to Covid-19, the subject property’s annual expenses increased by 3.3 and 3.4 percent in fiscal years (FY) 2019 and 2020. (Def’s Ex A at 110.5) By contrast, its expenses increased by 10.5 percent and 9.3 percent in FY 2021 and 2022. (Id.) Lee wrote that, based on his discussions with operators, increased expenses were not expected to be permanent. (Ptf’s Ex 1 at 96-97.)

Investor surveys in 2021 indicated that market participants expected Covid-19 impacts to be short-lived and anticipated normal occupancy in 2022. (See Ptf’s Ex 1 at 34; Def’s Ex A at 51.) Lee wrote that he concluded a three percent market inflation rate based on the expectations of buyers and sellers as of December 2020, as revealed through the CPI, surveys, and other data.

4 Both appraisers acknowledged that one-time government subsidies alleviated some of the Covid-19 impacts, but neither attempted to quantify how those subsidies impacted value. (See, e.g., Def’s Ex A at 51-52.) 5 The subject property’s fiscal year is July 1 to June 30. (See Ptf’s Ex 1 at 85.)

DECISION TC-MD 220292N 4 (Ptf’s Ex 1 at 64.) Brown noted that the overall trend before Covid-19 “was relatively stable with interest rates close to zero.” (Def’s Ex A at 119.) Home values appreciated “at an accelerated pace” during the pandemic, but FED interest rate hikes in 2022 and 2023 placed downward pressure on those values. (Id. at 118-120.)

B. Parties’ Appraisals, Requested Values, Tax Roll Values

The appraisers agreed that the subject property’s highest and best use as improved was continued use as a skilled nursing facility. (Ptf’s Ex 1 at 61; Def’s Ex A at 65.) Lee used each of the three approaches to value: the cost approach, the sales comparison approach, and the income approach. (Ptf’s Ex 1 at 59.) Brown used only the sales comparison and income approaches, concluding that the cost approach was too speculative based on the subject property’s age and several remodels. (Def’s Ex A at 66.) The appraisal reports are discussed in detail below.

The subject property’s 2021-22 tax roll real market value, sustained by the board of property tax appeals (board), was $10,561,070. (Compl at 3.) Its maximum assessed value and assessed value were each $7,435,440. (Id.) Lee concluded a real estate value of $9,450,000, and Plaintiff requests that the 2021-22 real market value be reduced accordingly. (See Ptf’s Ex 1 at 3.) Brown concluded a real estate value of $11,640,000, but Defendant requests that the 2021-22 tax roll real market value be sustained. (See Def’s Ex A at 129.)

II. ANALYSIS

The issue presented is the subject property’s 2021-22 real market value. “Real market value” is “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 DECISION TC-MD 220292N 5 of the assessment date for the tax year.” ORS 308.205(1).6 The assessment date for the 2021-22 tax year was January 1, 2021. ORS 308.007; ORS 308.210. Real market value “shall be determined by methods and procedures in accordance with rules adopted by the Department of Revenue * * *.” ORS 308.205(2). Three approaches to value must be considered but may not be applicable in every case: the cost approach; the sales comparison approach; and the income approach. Oregon Administrative Rule (OAR) 150-308-0240(2)(a); see also Dept. of Rev. v. River’s Edge Investments, LLC, 359 Or 822, 827, 377 P3d 540 (2016).

Plaintiff bears the burden of proof by a preponderance of the evidence, which means “the greater weight of evidence, the more convincing evidence.” Feves v. Dept. of Revenue, 4 OTR 302, 312 (1971); see also ORS 305.427. “[T]he court has jurisdiction to determine the real market value or correct valuation on the basis of the evidence before the court, without regard to the values pleaded by the parties.” ORS 305.412.

As noted, both appraisers considered all three approaches to value, with Lee developing an opinion of value under each approach and Brown developing an opinion only under the sales comparison and income approaches. Both appraisers placed primary, if not sole, weight on the income approach. Both appraisers have significant experience valuing senior housing and gave credible testimony in this case. They broadly followed similar methodologies, with discrete differences in the selection of comparable sales (Oregon vs. western region); the consideration of post-assessment date income and expenses; the treatment of specific expenses (reserves for replacement and property taxes); the selection of a capitalization rate; and the allocation of the going concern value between real estate, personal property, and intangible assets.

6 The court’s references to the Oregon Revised Statutes (ORS) are to 2019.

DECISION TC-MD 220292N 6 A. Approaches to Value

1. Cost approach

“The cost approach estimates value from the cost that would be needed to construct a similar property * * *.” Rivers Edge, 359 Or at 827. Lee testified that one benefit of the cost approach is that it does not capture intangible value, only “sticks and bricks.” A drawback of the approach is calculating depreciation, which is especially difficult with older properties like the subject. Ultimately, the cost approach is useful as a check on the other approaches. In addition to the difficulty accurately calculating depreciation, Brown noted a lack of good cost data because no new nursing facilities had been built in Oregon in the past 10 years. Lee agreed it is difficult to build in Oregon due to costly materials and the price per square foot compared to the return on investment; it’s usually more cost effective to update an old facility.

For his cost approach, Lee found a land value of $2.5 million based on comparable sales. (Ptf’s Ex 1 at 65-69.) To determine the improvements value, he used both the Marshall and Swift cost guide and budgeted construction costs for other skilled nursing facilities, albeit not in Oregon. (Id. at 70-72) Lee calculated depreciation using the age life method and concluded a total value of $10 million under the cost approach. (Id. at 73-75.) The court accepts Lee’s value estimate under the cost approach, subject to the limitations of that approach in this case.

2. Sales comparison approach

The sale comparison approach “estimates value from the prices paid for similar properties.” Rivers Edge, 359 Or at 827; see also Powell Street I, LLC v. Multnomah County Assessor, 365 Or 245, 249-50, 445 P3d 297 (2019) (explaining the sales comparison approach “estimates the value of the property by extrapolating from the prices paid for similar properties in the area”). “In utilizing the sales comparison approach, only actual market transactions of DECISION TC-MD 220292N 7 property comparable to the subject, or adjusted to be comparable, may be used. All transactions utilized in the sales comparison approach must be verified to ensure they reflect arms-length market transactions.” OAR 150-308-0240(2)(c). Generally, the fewer adjustments required the better to ensure properties are truly comparable. Allen v. Dept. of Rev., 17 OTR 248, 262 (2003).

a. Appraisers’ evidence under sales comparison approach

Lee testified that the typical buyer of a skilled nursing facility is a regional operator with a portfolio of properties or a REIT (real estate investment trust); the buyer is looking for an income stream. He identified the relevant regions as Oregon, Washington, and Northern California, though he did not find any good comparable sales in Washington: two were losing money and not stabilized, and others had a different resident census mix. Brown agreed that the market for skilled nursing facilities is regional, but he chose to focus on Oregon sales because of the difference in Medicaid reimbursement rates in California. Lee acknowledged that Medicaid payments vary state by state: Oregon uses a flat rate, whereas California varies by facility.7

Lee selected seven comparable sales that occurred between 2016 and 2020. (Ptf’s Ex 1 at 76.) Two sales were in Oregon, one was in Nevada, and the other four were in California. (Id.) Lee testified that the comparable sales were all individual sales of skilled nursing facilities. He made time adjustments consistent with inflation, reflected in the CPI, Medicaid and Medicare rates, and some location adjustments for adjusted values ranging from $99,726 to $141,207 per unit with an average of $113,873 per unit. (See id. at 80-81.) Lee testified that he placed the most weight on sales one and two, in Oregon, and seven, in California, because they required the fewest adjustments. Sale one was a property in Ashland that sold in 2016 for an adjusted price

7 Medicare is subject to a national standard.

DECISION TC-MD 220292N 8 of $141,207 per unit. (Id. at 76, 81.) Sale two was a property in Portland that sold in 2016 for an adjusted price of $105,499 per unit. (Id.) Sale seven was a property in Stockton, California, that sold in February 2020 for an adjusted price of $112,880. (Id.) Lee found a value of $114,000 per unit for the subject property, or $11,970,000. (See id. at 84.)

Brown selected four comparable sales, all in Oregon, though sales one and three are the same property located in Portland: sale one was in August 2022, and sale three was in October 2016.8 (Def’s Ex A at 68-69.) He testified that the pair of sales showed the price increase over time and the market for skilled nursing facilities after Covid-19. Sale two was a property in Keizer that sold in November 2021, and sale four was a property in Milwaukie that sold in November 2015. (Id.) Brown adjusted the sales for time by reference to Medicaid and Medicare reimbursement rates, and he made a downward adjustment of 10 percent to each sale for “area per bed” because the subject property had 451 square feet per bed whereas the comparable sales had 313 to 335 square feet per bed. (Id. at 74-76.) Brown found an adjusted value range of $110,602 to $218,036 per bed with an average of $151,999 per bed.9 (Id. at 80.) He testified that sale 2 at $218,036 per bed might be high due to more Medicare patients in the census mix. Brown concluded $150,000 per bed for the subject property, or $14,400,000. (Id.)

b. Appraisers’ income multipliers

Both Lee and Brown performed additional analyses of revenue per unit, which they used to form their opinion of value under the sales comparison approach: net operating income (NOI)

8 Brown’s sale 3 is the same as Lee’s sale 2. (Compare Ptf’s Ex 1 at 76 with Def’s Ex A at 80.) 9 Brown analyzed his sales both on price per square foot and price per unit or bed. The court here selects price per bed as the relevant unit of value because that is most commonly used by market participants. (See Def’s Ex A at 83 (giving “primary emphasis” to price per bed as “the base unit of revenue at a nursing home”).

DECISION TC-MD 220292N 9 per unit and a comparison of the effective gross income multiplier (EGIM) to sale price. (See Ptf’s Ex 1 at 81-84; Def’s Ex A at 80-83.) Their conclusions are as follows:

Comparable Sales NOI/unit EGIM Conclusion Plaintiff $11,970,000 $13,125,000 $12,881,818 $12,700,000 Defendant $14,400,000 $12,000,000 $12,000,000 $13,200,000

With respect to the NOI per unit method, Lee and Brown both found a value of $125,000 per unit, but Lee multiplied that value by 105 beds whereas Brown multiplied by 96 beds. (Compare Ptf’s Ex 1 at 84 with Def’s Ex A at 82.)

c. Court’s analysis and conclusion under sales comparison approach

The court agrees with the appraisers that comparable sales data is lacking with respect to the January 1, 2021, assessment date. As Brown noted, the best sales will be from Oregon to reflect similar Medicaid reimbursement rates. Unfortunately, the only Oregon sales are somewhat remote in time from the assessment date – most in 2015 or 2016. Brown identified one sale in 2021 in Keizer, Oregon, but it was able to achieve significantly more income than other properties, perhaps due to a different census mix. Looking to Lee’s three best sales (one, two, and seven) and Brown’s three best sales (excluding sale two), the court finds an indicated value of $125,000 per bed for the subject property, or $13,125,000.

The appraisers also presented various income multiplier methods in the context of the sales comparison approach. Because income multipliers such as EGIM “are mathematically related to direct capitalization,” those methods are “properly considered as a subset of the income approach * * *.” Allen, 17 OTR at 253; see also Appraisal Institute, The Appraisal of Real Estate 415, 417, 473 (15th ed 2020) (income multipliers are relevant to investor behavior based on income potential of property; considered under direct capitalization due to mathematic relationship). Further, income multiplier methods reliant on weak comparable sales are less DECISION TC-MD 220292N 10 helpful than a “traditional sales comparison” approach, which can better account for differences between the subject property and comparable sales. See Allen, 17 OTR at 264-65 (placing no weight on income multipliers). Given the lack of comparable sales in Oregon close to the assessment date, the court places little weight on the income multipliers.

3. Income approach

“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.” Powell Street I, 365 Or at 249-250. The appraisers each used the direct capitalization method, which utilizes the formula NOI divided by cap rate equals value. (See Ptf’s Ex 1 at 85, 106-107; Def’s Ex A at 84.) The direct capitalization method “focuses on two key components: (1) the capitalization rate (cap rate) and (2) net operating income (NOI).” Allen, 17 OTR at 253.

a. NOI

NOI, which “is the currently expected net income of a property after all operating expenses are deducted from gross income[,]” is determined by reference to the subject property’s historical income and expenses, “adjusted by reference to market data.” Allen, 17 OTR at 254.

(1) Effective gross income (EGI)

The subject property’s revenue streams are from private pay, Medicare, Medicaid, and other income. (See Ptf’s Ex 1 at 85 Def’s Ex A at 103.) Lee analyzed the subject property’s operational history, looking for rates in each category as of January 1, 2021, and compared those to other Portland area nursing facilities.10 (Ptf’s Ex 1 at 89-92.) For instance, Lee used the 2020

10 Lee concluded daily unit rates as follows: $350 for private pay; $625 for Medicare; $377 for Medicaid; and $484 for other. (Ptf’s Ex 1 at 89-91.)

DECISION TC-MD 220292N 11 actual Medicare rate and increased it by 2.2 percent based on the Centers for Medicare and Medicaid Services projections for FY 2021, which were released in July 2021. (Id. at 90) Similarly, he calculated a blended Medicaid rate based on the basic and complex reimbursement rates and the 2020 data, indicating that 18 percent of days were complex and the remainder basic. (Id. at 91-92.) To develop a census mix, Lee considered data from the subject property and comparable properties from 2018 through 2022, selecting figures in between the 2020 and 2021 actual census. (Id. at 93) Lee applied a vacancy rate of 23 percent, reflecting 81 occupied beds, and calculated (EGI) of $12,268,398. (Id. at 94-95.)

Brown considered the same mix of income types,11 but he used the actual Medicare reimbursement rates from FY 2021 and 2022; the FY 2021 rate increased by 24.1 percent over FY 2020 and declined 4 percent in FY 2022. (Def’s Ex A at 95.) Similarly for Medicaid, Brown developed a “blended rate” including parts of FY 2021 and FY 2022, which were “higher than the published [basic] rate * * * during this time period.” (Id. at 97.) Brown testified that he focused on post-assessment date data given the extraordinary impact of the pandemic on skilled nursing facilities. He noted that most data from FY 2020, ending in June 2020, did not reflect the pandemic, whereas the period from July 2020 to June 2021 entirely reflected the pandemic’s impact. Brown selected a census mix giving most weight to FY 2021 and 2022 figures and applied a vacancy rate based on 80 occupied beds for EGI of $12,620,000. (Id. at 100-103.)

The subject property’s actual EGI was $12,104,937 in 2020 and $12,143,454 in 2021. (Ptf’s Ex 1 at 95; Def’s Ex A at 91.) Brown wrote that the subject’s 2021 “stagnating” EGI

11 Brown concluded daily unit rates as follows: $348 for private pay; $705 for Medicare; $385.50 for Medicaid; and $484 for other. (Def’s Ex A at 103.)

DECISION TC-MD 220292N 12 reflected significantly lower occupancy, common for skilled nursing facilities nationwide. (Def’s Ex A at 91.12) The subject’s EGI declined to $10,815,750 in FY 2022. (See id.)

The court accepts Lee’s EGI of $12,268,398 as most reflective of market participants’ expectations as of January 1, 2021. As this court discussed in Pacific Ethanol Columbia, LLC v. Morrow County Assessor, TC-MD 200380N, 2023 WL 3297100 at *1 (Or Tax M Div, Apr 20, 2023), the key inquiry with post-assessment date evidence concerns the similarity of conditions with the assessment date. For instance, in a period of “intense change,” the court declined to consider two transactions 22 and 46 months after the assessment date. Level 3 Communications, LLC v. Dept. of Rev., 23 OTR 87, 102-103 (2018). Brown’s reliance on post-assessment date income data reflects significant inflation that was not known or anticipated as of January 1, 2021; indeed, the Fed’s first interest rate hikes were 15 months after the assessment date. (See Def’s Ex A at 120.) Investor surveys revealed an expectation that Covid impacts would be short-lived, and occupancy would return to normal levels in 2022. Accordingly, the court declines to adopt Brown’s EGI for the January 1, 2021, assessment date.

(2) Expenses and NOI conclusion

For expenses, Lee reviewed the subject property’s actual expenses from 2018 through 2020, and those of comparable properties, concluding expenses of $356.33 per resident day, or $10,522,389 in total, for an expense ratio of 85.77 percent. (Ptf’s Ex 1 at 95-95, 110.) His expenses included reserves for replacement of $350 per unit, or $36,750 in total. (Id. at 85-86.) Lee wrote that “the reported and comparable datasets have been adjusted * * * to include the market-level replacement reserves.” (Id. at 103.) In his initial analysis, Lee included property

12 Occupancy declined from 60.9 percent in 2021 to 53.8 percent in 2022. (Def’s Ex A at 101.)

DECISION TC-MD 220292N 13 taxes of $187,202 for an NOI of $1,746,009. (Id. at 97, 110.) In a later part of his report, Lee recalculated the NOI without property taxes to be $1,933,211 and loaded the leased fee cap rate to include the tax rate of 1.03 percent. (Id. at 122-123.)

Brown considered similar data as Lee and concluded expenses of $375.98 per resident day, or $10,980,000 in total, for an expense ratio of 87 percent. (Def’s Ex A at 104-115.) As with his EGI calculation, Brown focused on FY 2022 data, explaining that it “is more removed of the COVID pandemic” and so a “better indicator of ‘typical expenses.’” (Id. at 106.) He concluded a lower expense ratio than the subject’s actual expense ratio, finding the high ratio was “likely related to the low occupancy” in FY 2022. (Id. at 112.) Brown’s primary calculation did not include reserves for replacement because such expense items are not typically included in “the elder care industry” when calculating the cap rate, and the cap rates extracted from his sales were not adjusted for reserves. (Id. at 113.) He gave an alternate calculation that included reserves of $50,000, or 0.4 percent, increasing expenses to $11,030,000 and noting that the cap rate should be reduced by 0.25 to 0.50 percent if reserves are included. (Id.) Brown included property taxes of $190,000. (Id. at 115.) He concluded an NOI of $1,640,000 without reserves and $1,590,000 with reserves. (Id.)

Including property taxes but excluding reserves, the subject property’s expenses were:

Per resident day Total expenses Expense ratio NOI 7/1/19 – 6/30/20 $337.65 $10,012,279 82.7% $2,092,658 7/1/20 – 6/30/21 $373.24 $10,536,557 86.8% $1,606,897 7/1/21 – 6/30/22 $408.10 $10,171,420 94.0% $644,330

(Def’s Ex A at 105.)

For the same reasons as discussed above, the court finds Lee’s expense estimate more reflective of what market participants anticipated as of January 1, 2021. Two discrete expenses require further comment: reserves for replacement and property taxes. Including reserves for DECISION TC-MD 220292N 14 replacement in expenses is “an accepted practice” but is “proper only if the comparables used to develop a cap rate did so as well.” Allen, 17 OTR at 257, 262; see also The Appraisal of Real Estate at 462 (“[i]ncome and expenses must be estimated on the same basis for the subject property and all comparable properties[,]” listing reserves as an example) and at 466 (when considering cap rate data from surveys, appraiser may consider whether reserves were included in survey data). Lee wrote that he adjusted his “comparable datasets” to reflect a market rate for reserves, so the court accepts his expense estimate including reserves. Further discussion of how reserves may impact the cap rate is included in that section below.

Excluding property taxes from expenses and adding the effective tax rate to the cap rate is “an accepted method” in a value appeal “where the amount of the taxes is a function of the ultimate value that has yet to be determined.” See Allen, 17 OTR at 259 n12 (but noting the complication of Measure 50, which caps property taxes, and suggesting an “iterative approach” may be required); see also NYEI LLC v. Umatilla County Assessor, TC-MD 110340D, 2012 WL 114204 at *12 (Or Tax M Div, Jan 13, 2012) (using loaded cap rate method); see also The Appraisal of Real Estate at 259-60, 449-50 (approving method). In Allen, the court ultimately accepted the county’s method of adding the maximum property tax burden to expenses because it was adverse to the county and potentially beneficial to the taxpayers. Allen, 17 OTR at 259 n13. For the same reason and for the sake of consistency between the income approach and the going concern allocation,13 the court accepts Lee’s NOI of $1,746,009 including property taxes. ///

13 Lee recalculated his “real estate only cap rate” for purposes of allocating the going concern value between real, personal, and intangible property, but he made no adjustment to the cap rate he used in his income approach to calculate the going concern value. (Compare Ptf’s Ex 1 at 110 with 123 (going concern value based on NOI including property taxes of $1,746,009 and cap rate of 13.5 percent, whereas “inferred real estate value” based on NOI of $1,933,211 and loaded cap rate of 12.03 percent).

DECISION TC-MD 220292N 15

b. Cap rate and direct capitalization conclusion

“A cap rate is generally calculated using market sales. Slight deviations in cap rates profoundly change the estimated value of a property, making the proper calculation of the rate of paramount importance.” Allen, 17 OTR at 260. Investor surveys are a “secondary” source or “support” for the cap rate. The Appraisal of Real Estate at 465-66. The appraiser may consider “[t]he fact that surveys almost always represent leased fee [cap] rates.” Id. at 466.

To determine the relevant cap rate, both appraisers considered their comparable sales and investor surveys. (Ptf’s Ex 1 at 107-109; Def Ex A at 116-121.) Lee also considered the “band of investment technique” as “a crosscheck” on the other techniques. (Ptf’s Ex 1 at 108.) Those results are summarized in the following table:

Sales (fee simple) Survey (Class B Core14) Band of Investment Conclusion Plaintiff 10.41 – 16.57% 10.0015 – 14.00% 13.4% 13.5% Mean 13.31% Mean 12.00% Defendant 12.29 – 16.94% 9.0 – 14.0% 12.25% Mean 14.06% Mean 11.5 – 12.25%

To select a cap rate, Lee “considered recent events and prevailing market conditions” including “a combination of inflationary pressures, higher cost of capital (considering interest rates as well as risk spreads), and the recent geopolitical events.” (Ptf’s Ex 1 at 109.) He acknowledged that “the overall long-term outlook for commercial real estate remain[ed] positive” but the “uncertainty and the higher cost of capital [had] an upward influence on [cap] rates * * *.” (Id.) Lee concluded an indicated value of $12,900,000. (Id. at 109-110.)

14 Lee wrote that the subject property is a Class B facility in a primary market. (Ptf’s Ex 1 at 109.) Brown concurred in his testimony. 15 In his report, Lee summarized the investor surveys as indicating range of 12.00 to 13.60 percent. (Ptf’s Ex 1 at 109.) However, he did not explain how he determined that smaller range from the larger range of data he reported. (See id. at 108.)

DECISION TC-MD 220292N 16

To explain his selection of a cap rate below the range indicated by his comparable sales, Brown testified that he considered additional data, such as the income multipliers, and concluded that the subject property is underperforming and has potential for growth. He noted the subject property’s higher than average common area might be converted to rehab space and generate additional revenue. Brown adjusted his cap rate down to 11.9 percent for reserves and concluded an indicated value of $13,350,000. (Def’s Ex A at 121-122.)

As noted above, extracting cap rates from comparable sales is the preferred method. Here, the comparable sales each suffer some weaknesses, either because they are remote in time from January 1, 2021, or located outside of Oregon, thus reflecting different Medicaid rates. Considering survey data as a secondary source is appropriate. Survey data also suffers some weaknesses, including how respondents calculated NOI. Here, the comparable sales indicate a cap rate in the range of 13.3 to 13.5 percent.16 That is within the range indicated by investor surveys, albeit at the higher end. The appraisers agreed that the subject property was relatively average in most respects. Although the subject may have potential to generate additional income in the future from its common area, the court received no evidence suggesting that potential was likely as of January 1, 2021, particularly given staffing shortages, reduced occupancy, and unknown costs. The court finds a cap rate of 13.25 percent is supported, indicating a value of $13,177,426 under the direct capitalization method.

c. Discounted cash flow (DCF) analysis

Lee also performed a DCF analysis because buyers and sellers typically consider that method in addition to direct capitalization. (See Ptf’s Ex 1 at 111-115.) He found a value of $13

16 Brown’s average cap rate of 14 percent is reduced by 0.5 percent for reserves. (Def’s Ex A at 113.) Lee’s sale 3 financial data included reserves of $350 per bed, which Lee considered the market rate for the subject property. (Ptf’s Ex 1 at 78, 85-86, 103.) He did not address whether other comparable sales included reserves.

DECISION TC-MD 220292N 17 million under that method. (Id. at 116.) The court accepts Lee’s value estimate under the DCF but places limited weight on it. (See id. (placing primary weight on direct capitalization); see also Def’s Ex A at 84 (noting DCF “is appropriate if net income is irregular).)

4. Reconciliation and going concern value conclusion

Where multiple approaches to value are used, “the appraiser will then reconcile the values obtained from the various approaches to obtain a final valuation.” Powell Street I, 365 Or at 249-50 (citations omitted). The weight given to each approach “is a question of fact that depends on the record developed in the case.” Id. (citations omitted). In reconciliation, both appraisers placed primary, if not sole, weight on the income approach, treating the other approaches as a check. (Ptf’s Ex 1 at 117; Def’s Ex A at 123.) Lee concluded a value of $12,900,000 and Brown concluded a value of $13,350,000. (Id.)

To recap the values found by the court: $13,125,000 under the sales comparison approach and $13,177,426 under the income approach (direct capitalization). Lee’s DCF indicated a value of $13,000,000. The court accepts the appraisers’ opinion that primary, if not sole, weight should be given to the income approach and concludes a going concern value of $13,170,000. B. Allocation of Value to Real Estate

The going concern value of a property “includes the incremental value associated with the business concern, which is distinct from the value of the real estate.” Deschutes County Assessor v. Broken Top Club, LLC, 15 OTR 231, 237 (2000). Generally, Oregon only taxes real and personal property. Id.; see also Tetherow Golf Course LLC v. Deschutes County Assessor, 20 OTR 554 (2012) (affirming principle in Broken Top). The court has recognized that senior housing and care facilities often involve services and amenities beyond housing, such as “food services, housekeeping, assisted living or nursing care and other amenities provided the DECISION TC-MD 220292N 18 elderly.” Polk County v. Dept. of Rev., 14 OTR 566, 570 (1999) (applying ORS 308.490, which provides special appraisal procedures for “nonprofit homes for elderly persons”); see also Hope Village, Inc. v. Dept. of Rev., 17 OTR 370, 380 (2004) (expenses “unique to housing for the elderly, such as housekeeping or assisted living care, should be eliminated” under ORS 308.490).17 With properties that sell as a going concern, “it may be possible to extract [cap] rates from sales of going concerns using a residual method.” The Appraisal of Real Estate at 675.

Both appraisers allocated their value conclusions between real, personal, and intangible property. (Ptf’s Ex 1 at 118; Def’s Ex A at 124.) Lee used both a cost residual method and a lease coverage ratio (LCR) method, whereas Brown used only the LCR method. (Id.) For his cost residual method, Lee compared his value under the cost approach with his value under the income approach to determine a business value of $2,900,000 and a personal property value of $930,000, leaving a real estate value of $9,070,000. (Ptf’s Ex 1 at 118.)

The appraisers agreed that senior housing market participants prefer the LCR method to determine market rent and isolate the real estate component of value. (Ptf’s Ex 1 at 118; Def’s Ex A at 124.) For example, an investor might purchase a going concern and then lease the building to an operator, using the LCR to determine rent for the building. (Def’s Ex A at 125.) Lee described the concept as bifurcating NOI into real estate and operations by looking at leases of nursing facilities. The method indicates real estate value by applying a “market derived lease coverage ratio” to net operating income. (Ptf’s Ex 1 at 118.) Lee testified that skilled nursing

17 The court does not mean to suggest that ORS 308.490 applies in this case and neither party has argued it does; rather, the court cites those cases for their general descriptions of typical senior housing amenities. The Appraisal of Real Estate states “a senior care facility” is likely to include intangible assets. Id. at 663. Properties like assisted living facilities “rarely sell[] independently of personal property and intangible property.” Id. at 664.

DECISION TC-MD 220292N 19 facilities generally have higher LCRs than other types of facilities due to the higher acuity of care and related costs, as well as the risk and complexity of the operation.

Before addressing the details of the LCR method, the court must address an apparent difference of opinion between the appraisers in the order of operations. Lee divided his NOI by the LCR to arrive at a figure he described as “cash flow to the real estate,” subtracting that from the value of the going concern, and then separately subtracted the value of personal property from the value of the going concern. (See Ptf’s Ex 1 at 122-124.) By contrast, Brown divided his NOI by the LCR to arrive at a figure he described as “Real Estate/Personal Property Rent” and subtracted the value of personal property from that figure. (See Def’s Ex A at 127-128.) Brown testified that most leases he’s seen have clauses stating that personal property stays with the real property when the tenant vacates. He’s also seen leases stating that the landlord supplies the personal property. So, Brown concluded that the rent is for the real and personal property, indicating that personal property value should be deducted after rent is calculated. The court accepts Brown’s method as reasonable and received no evidence to rebut his testimony.

1. Selection of the LCR

The appraisers each considered several sources of LCRs: 1) REITs; 2) survey data from market participants; and 3) sales. (Ptf’s Ex 1 at 119-120; Def’s Ex A at 125-127.) Lee gave data from a variety of senior housing types, whereas Brown focused exclusively on skilled nursing. Lee testified that he provided broader datasets to give context and support the ratio range. /// /// /// /// DECISION TC-MD 220292N 20

REITs SNF only18 Investor surveys Leased fee sales SNF only19 Plaintiff 1.07 – 2.59 1.24 – 2.59 1.30 – 1.90 1.60 – 2.00 1.21 – 2.00 Mean: 1.38 Mean: 1.67 Mean: 1.41 Mean: 1.59 Defendant 1.1 – 1.8 1.08 – 1.80 Mean: 1.45

Lee selected an LCR of 1.70 for the subject property. (Ptf’s Ex 1 at 120.) Defendant questioned why Lee selected an LCR at the higher end of the range given that the subject property is typical and functional. Lee testified that many sales are part of portfolios so that diversifies and mitigates some risk compared to a standalone sale. Brown testified that his sales data reflected a mix of portfolio and standalone sales. (See Def’s Ex A at 126.) He selected an LCR of 1.45 for the subject property. (Id. at 127.) The court finds an LCR of 1.55 is supported by the evidence presented. Applying that to the NOI of $1,746,009 indicates rent or cash flow to the real and personal property of $1,126,457.

2. Selection of the leased fee cap rate

The next step in the LCR method is to determine a leased fee cap rate. (Ptf’s Ex 1 at 120-121; Def’s Ex A at 127.) Each appraiser provided a nationwide dataset of leased fee sales. Lee included a variety of property types, whereas Brown focused exclusively on skilled nursing facilities. Lee relied on the “basis point spread” between the fee simple and leased fee cap rates, whereas Brown considered the reported leased fee cap rates. That data is as follows:

Fee simple SNF only Lease fee SNF only Basis point SNF only cap rates cap rates spread Ptf 6.05-13.51% 9.52-2.72% 4.90-9.74% 7.24-9.74% 47 – 510 228 – 510 Mean: 8.4% Mean: 257 Mean: 353 Def 8.0-9.7% Mean: 9.1%

18 Lee’s SKN only responses: 1.24, 1.30, 1.54, and 2.59. (Ptf’s Ex 1 at 119.) 19 Lee’s SKN only sales: 1.21, 1.31, 1.31, 1.42, 1.67, 1.67, 1.81, 1.82, and 2.00. (Ptf’s Ex 1 at 120.) Brown testified that a few of his sales were campuses with primarily skilled nursing facilities. (See Def’s Ex A at 126.)

DECISION TC-MD 220292N 21

Lee testified that he used basis point spread instead of leased fee cap rates because the dataset reflected superior and portfolio sales. For instance, he noted two skilled nursing facilities in Arizona and Colorado that each sold in July 2018 and were newer than the subject property. (See Ptf’s Ex 1 at 121.) The Colorado sale reported a leased fee cap rate of 7.24 percent with a basis point spread of 228, whereas the Arizona sale reported a leased fee cap rate of 7.58 percent with a basis point spread of 510. (Id.) Lee selected a basis point spread of 250 based on the average of all property types “and the most similar standalone SNF transactions * * *.” (Id.) Using his basis point spread of 250, Lee calculated a “real estate only cap rate” of 11 percent for the subject property. (See id. at 122.)

Brown testified that he had never seen a basis point spread analysis, but he did not see any problem with it, other than Lee’s consideration of various property types rather than only skilled nursing. Brown testified that his dataset included both portfolio and standalone sales, though he did not see much variation between the two types. He noted a few standalone sales: Meadows on University with an 8.9 percent cap rate, Oakview with a 9.5 percent cap rate, and an Indiana facility with a 9.5 percent cap rate. (Def’s Ex A at 127.) Brown selected a leased fee cap rate of 9.0 percent. (Id.)

The court agrees with Brown that skilled nursing facilities provide the best evidence, particularly given the risk and complexity of skilled nursing facilities as compared with other types of senior housing. Upon consideration, the court finds a leased fee cap rate of 9.5 percent is supported in this case. That rate is at the higher end of the sales provided but supported by the standalone sales that Brown identified. It suggests a basis point spread of 375 when compared to the overall cap rate of 13.25 percent, which is slightly higher than the average for skilled nursing. A 9.5 percent leased fee cap rate indicates a value of $11,857,442 for real and personal property. DECISION TC-MD 220292N 22

3. Estimate and subtraction of personal property; indicated real estate value

Lee estimated the value of personal property using a depreciated cost method, concluding a value of $930,000. (Ptf’s Ex 1 at 74.) Brown looked to actual costs for new nursing facilities in the western U.S. between 2011 and 2016, finding a range of $15,000 to $20,000 per bed. (Def’s Ex A at 128.) He selected $15,000 per bed for the subject property and depreciated it by 50 percent, indicating a value of $540,000 for personal property. (Id.) The court accepts Lee’s value of $930,000 as a reasonable estimate based on the actual personal property. That indicates a real estate value of $10,927,442.20 The real estate only value is very close to the 2021-22 tax roll real market value of $10,561,070. Defendant did not request an increase in the real market value and the court finds no reason to order a change. See Price v. Dept. of Rev., 7 OTR 18, 25 (1977) (observing that value is a range rather than an absolute).

III. CONCLUSION

Upon careful consideration, the court finds the 2021-22 tax roll value of $10,561,070 is supported for property identified as Account R328941. Now, therefore, /// /// /// /// /// /// ///

20 $11,857,442 - $930,000 = $10,927,442.

DECISION TC-MD 220292N 23

IT IS THE DECISION OF THIS COURT that Plaintiff’s appeal is denied.

Dated this ___ day of March 2024.

ALLISON R. BOOMER PRESIDING MAGISTRATE

If you want to appeal this Decision, file a complaint in the Regular Division of the Oregon Tax Court, by mailing to: 1163 State Street, Salem, OR 97301-2563; or by hand delivery to: Fourth Floor, 1241 State Street, Salem, OR.

Your complaint must be submitted within 60 days after the date of this Decision or this Decision cannot be changed. TCR-MD 19 B.

This document was signed by Magistrate Allison R. Boomer and entered on March 28, 2024.

DECISION TC-MD 220292N 24