Wake Robin Corp. v. Town of Shelburne

• Court: Vermont Superior Court
• Decided: January 14, 2013
• Docket: S0133
• Status: Published

Opinion

Wake Robin Corp. v. Town of Shelburne, No. S0133-11 CnC (Crawford, J., Jan. 14, 2013)

[The text of this Vermont trial court opinion is unofficial. It has been reformatted from the original. The accuracy of the text and the accompanying data included in the Vermont trial court opinion database is not guaranteed.] STATE OF VERMONT

SUPERIOR COURT CIVIL DIVISION CHITTENDEN UNIT DOCKET NO.: S0133-11 CnC

WAKE ROBIN CORP.

v.

TOWN OF SHELBURNE

DECISION ON PROPERTY TAX APPEAL

This is an appeal of the 2010 property tax valuation for the Wake Robin property in Shelburne, Vermont. The court held a trial in the matter from December 3 – 6, 2012.

FINDINGS OF FACT

I. Introduction

The Wake Robin property is a continuing care residential community (“CCRC”) located on a campus setting on 136 acres in Shelburne, Vermont. The owner Wake Robin Corp. is a non- profit corporation. The project was built in 1993 – 95 and substantially enlarged in 2007. It consists of a central dining hall and community center and surrounding cottages and buildings which contain apartments. There is also a health center which provides higher levels of care for residents. There is a maintenance building. There are recreational facilities such as tennis courts, an aquatic center, a library, and walking trails. Both sides agree that Wake Robin is an attractive place to live and an extremely well-run, successful community.

Wake Robin has 295 residential units. There are 114 apartments and 98 cottage units. These range in size from 557 square feet to 3,156 square feet. There are 35 assisted living and 48 skilled nursing units. The skilled nursing units include 18 “memory care” beds. The apartments and cottages are simple, attractive homes. Even the largest are relatively modest in scale. The assisted living and skilled nursing units are all private rooms.

II. Levels of Care

The distinguishing characteristic of a CCRC is that a resident can stay in the retirement community despite needing higher levels of care over time. There are many senior apartments and other communities which do not provide higher levels of care. A person who becomes incapacitated must move to a nursing home or some other “stand alone” facility. The CCRC model is designed to provide a full spectrum of care in one place so that spouses can live in the same community and a person can remain close to his or her friends and associations. There are essentially three levels of retirement care. The lowest level of care is “independent living” which is a level of care not unlike the care which a college provides to its students. At Wake Robin, a minimum of one meal per day, some housekeeping, grounds work, a wide variety of recreational activities, some transportation, and a low level of personal assistance are provided to adults who are still very much able to care for themselves.

As residents age, they may require “assisted living” which includes a higher level of service and care. At Wake Robin, a person receiving “assisted living” care will move from his or her independent apartment or cottage into the Linden Health Center. They receive a private room and bath. They remain welcome and included in the on-going life of the community.

Some residents will require skilled nursing at the end of their lives or following an episode of illness or injury. In some cases people will require special care for dementia. These residents also reside in the Linden Health Center on a floor where they can receive care equivalent to a nursing home. Subject to their limitations and desires, they also remain welcome in the larger community.

Wake Robin promises to provide all three levels of care for a resident’s lifetime. A typical resident or couple might move to Wake Robin in their late 70’s. They would be accepted only on condition that a medical exam shows that they are reasonably likely to live independently for at least two years. At any time during their residence, however, they are entitled to a higher level of care as they need it.

III. Cost of Care

Care of this nature is expensive. Wake Robin residents pay in three ways. They pay an initial entrance fee which ranges from $142,152 to $576,255 depending on the size of the independent living unit which they select. This entrance fee is treated as an insurance product. It functions in part as the purchase of long-term care insurance. For this reason, Wake Robin is regulated by the state insurance regulator. The entrance fees function as reserves which are amortized over the years a resident is projected to live at Wake Robin. The IRS treats 38 % of this payment as a health insurance premium for purposes of deductions from Wake Robin’s income tax.

Residents in the independent living apartments and cottages pay a monthly fee which ranges from $2,482 to $6,372 – again depending on the size of the independent living unit. These fees are also subject to the same 38 % deduction for health care costs.

When a person moves permanently to assisted living or another higher level of care, they move physically into the Linden Health Center. Since they have given up their independent living unit, the monthly fee drops to the lowest possible level with an additional meal adjustment.

These three charges and the investment income from the accumulated entrance fees are the primary sources of revenue for Wake Robin. Both the entrance fee and the monthly fee are

2 increased in the case of a couple. There are a few additional charges for extra services. In addition, Wake Robin receives a relatively small amount of Medicare and other reimbursements for patients recovering from events such as a hospitalization.

IV. Highest and Best Use and Date of Valuation

The parties agree that the highest and best use of the property is in its current use as a retirement community. They agree that the appropriate date of valuation for purposes of this case is April 1, 2010.

V. Recent Levels of Tax Appraisal

The Town’s appraisal of Wake Robin has increased rapidly over the period 2006 to 2010. The Town’s valuation of the real estate in recent years has been:

2006 -- $22,970,600. 1

2007 -- $30,845,900.

2008 -- $49,414,900.

2009 -- $49,328,200.

2010 -- $49,328,200.2

VI. Market Trends for CCRCs

Since 2008 Wake Robin and other CCRCs have experienced some downturn in demand. A drop in demand from consumers results in a decline in fair market value for the properties. The Town’s appraiser Courtney Lees described a significant decline in values for assisted living investments between 2007 and 2009:

The thinking by many investors had been that the assisted living industry was supposed to be resilient to economic downturns as the services are need driven just like the nursing levels and a slow-down in move-ins. With the apprehension in the lending arena and the higher costs of capital, the result was an average decline in prices of 21.0 percent from 2007 – 2008. For 2009, the first half of the year there was continued deterioration but by mid-year the capital markets began to open up as it appeared that properties had stabilized. The result was only a 9.0 percent drop in 2009. The average price [per unit] paid in 2008 was $124,900 and in 2009 it was $113,300. The median price paid was $98,400 in 2008 and in 2009 it was $102,400.

1 2006 was prior to the completion of construction in 2007of a new group of cottages, the aquatic center and other improvements. 2 Note: the 2010 tax bill also includes an unrelated exemption of $3,894,932 for land use reasons. This exemption changes the tax bill but does not affect the valuation of the real estate itself.

3 Town’s Exhibit C. Wake Robin currently has an occupancy rate of just over 90 %. The occupancy rate in past years was close to 100 %. Wake Robin now offers some modest discounts and inducements to new residents. The downturn is due to the general fall in the real estate market since 2008. Since many Wake Robin residents must sell their homes in order to pay the entrance fee, delays and lower sales prices for these homes reduce the number of people who can afford to move into Wake Robin and other CCRCs.

Although Wake Robin has not welcomed the decrease in occupancy, the community remains on a sound financial footing. Annual revenue from all sources has increased from $14.9 million in 2007 to $17.3 million in 2008 and 18.1 million in 2009. Wake Robin has substantial cash reserves primarily derived from the entrance fees which are amortized over the term of a resident’s stay. In 2010, total cash and investments excluding real estate were $26.3 million.

VII. Definitions

It is helpful to define a few terms at the outset of the discussion of valuation.

Cost approach to valuation. The cost approach requires the appraiser to calculate the cost of duplicating the improvements through current construction methods and then reduce that cost to reflect depreciation. The value of the underlying land and the depreciated cost of replacing the structures and other improvements should equal fair market value for the realty.

Income approach to valuation. The income approach is based upon the principle that the market of real estate investors demands a particular rate of return for various types of investments. A free market, filled with potential buyers, will establish a fair market value by evaluating the income generated by a property and the annual return expected from similar properties.

Sales approach to valuation. The sales approach is based upon a comparison of recent prices obtained for similar properties. By adjusting up or down for differences in size, condition, location and other variables, the appraiser can develop a prediction for a sales price for the subject property.

Net Operating Income (“NOI”) is the amount of money left over after the expenses of the business are subtracted from its revenues. It represents the cash return to the investor at the end of the year. It provides a principal element of the income method of valuation.

Business extraction rate describes the process of separating out NOI related to providing real estate to residents from the business-related income derived from providing services such as health care or meals. It is used to distinguish between the value of real estate owned by a business and the “going concern” value including intangibles such as staff and management skills, reputation, good will, and intellectual property.

4 Capitalization rate is the rate of return on capital – expressed in percentage terms – which the market demands for a particular business. It reflects the risk of failure or poor performance. A very safe business with little risk will have a relatively low capitalization rate. A riskier business will have a higher rate. As we divide ever lower capitalization rates into the same net income, the value of the business increases. Given the same rate of return, a safer business will be worth more to the investor.

Tax-loaded capitalization rate. Since NOI is calculated without subtracting the annual property tax bill, it is necessary to add the annual tax rate back into the capitalization rate to determine market value. In this case, the annual tax rate for the Town of Shelburne is 1.699 %.

Equalization rate. The equalization rate is a mandatory adjustment of assessed values for each municipality which is intended to achieve state-wide uniformity in valuation. The equalization rate varies from town to town depending upon the accuracy of each town’s assessment process. The rate is applied to the fair market value to produce the final assessed value for tax purposes. The equalization rate for the Town of Shelburne as of April 1, 2010 was 97.89 %.

Personal property. The tax value excludes personal property. Both appraisers have excluded personal property from their calculations.

VIII. Three Methods of Valuation

Both sides presented expert witness appraisals of fair market value for the Wake Robin property as of April 1, 2010. Both appraisers are experienced professionals in their field. Each took a different approach to the problem of valuation.

Wake Robin’s appraiser is George Silver. Mr. Silver employed all three methods of real property appraisal: the “income” approach; the “cost” approach; and the “comparable sales” approach. Of these, he determined the sales approach to be the most reliable. Mr. Silver calculated a fair market value of $40,000,000 for the Wake Robin property.

The Town’s appraiser was Courtney Lees. Ms. Lees employed the “sales” and “income” approaches. She found the “income” approach to be the most reliable. Ms. Less calculated a fair market value of $53,600,000 for the Wake Robin property.

The court’s task is to compare the reliability of the three approaches as they were applied by the appraisers. Most of this analysis appears in the “Conclusions” portion of this decision. Before reaching that discussion, it is necessary to make specific findings about the component parts of each approach.

5 IX. Findings Related to “Income” Valuation

A valuation based on income is a projection of what a willing buyer and seller would calculate as the value of the subject property based on its cash return as an investment. There are two variables in this equation: NOI and the capitalization rate. The value of any investment can be calculated as the net operating income to the investor divided by a capitalization rate.

The Wake Robin valuation presents a special problem in determining NOI. Wake Robin is engaged in three broad categories of business at the same time. These are the provision of housing and related amenities to residents, the sale of an insurance product for future health needs, and the provision of meals, health care and other services. Wake Robin charges for all three services at the same time through the entrance fee and the monthly fees. Since the property tax valuation applies only to the real estate, the appraisers must extract “business income” not related to the real estate from NOI before computing value.

All three categories depend upon a mixture of real estate and personal service. Providing an apartment or a cottage is almost entirely a “real estate” transaction. Providing meals and health care requires both a location – real estate – and a high level of employee service. Accepting a one-time premium and establishing a reserve to pay for future health care is largely a “non-real estate” transaction.

The two appraisers used different approaches to the problem of “separating the egg” – that is, isolating the income associated with a return on the real estate only. Both agree that it is a necessary step in the analysis. There is substantial disagreement about how to separate the elements and in what proportions.

A. Net Operating Income

The Town used gross revenues. For the year 2010, the Town’s appraiser Ms. Lees projected total revenues of $18,828,000. These are consistent with Wake Robin’s own budgeted revenues as well as the historical reports of revenues.

From gross revenues, the Town’s appraiser deducted expenses of $12,461,400 (including an imputed 5 % management fee). She also deducted an annual “replacement reserve” of $500 per unit for $147,500.

By subtracting expenses from revenues, Ms. Lees calculated net operating income (“NOI”) of $6,219,100. The court finds that Ms. Lees’ figures for gross revenues, total expenses and NOI accurately portray the financial operation of Wake Robin in 2010.

B. Average Rent Per Unit

Wake Robin’s appraiser George Silver developed a rent-based approach to revenue. Although the majority of CCRCs are owned by non-profits and are not leased, there are examples of

6 retirement communities which have been purchased by real estate investment trusts (“REITs”) and leased to operating companies. Since by design a REIT limits itself to ownership of real property only, rents received by the REIT from the operating company should reflect a return to the real estate only.

Mr. Silver was able to identify 18 examples of CCRCs leased by REITS to operating companies in the eastern half of the United States between 2005 and 2011. The retirement communities provided various levels of service but all provided some level of continuing care in addition to “independent living.”

The leases identified by Mr. Silver included smaller units with fewer than 100 units (3), 100 – 200 units (4), 200 – 400 units (5) and more than 400 units (6). Only 5 units were approximately the size of Wake Robin. The annual rents varied from $6,500 per unit to almost $15,000. Some of the leases had escalation clauses which would increase the rent from the date of signing.

In his report, Mr. Silver stated that

After careful analysis, the net operating income (NOI) for the subject property, based on market activity is estimated to be approximately $11,000 per unit/bed. The estimated annual net operating income for the subject real property assets is, therefore estimated to be ($11,000/unit x 295 units) $3,245,000.

The details of this “careful analysis” are not set out in his report. His testimony was that the estimation of an annual rent for Wake Robin was the product of professional judgment and cannot be shown with mathematical certainty.

The court rejects Mr. Silver’s “average rent” analysis as not credible. It is based on a small sample of transactions taken from widely varying communities across the eastern United States over a five year period. The total sample is very small. He was unable to adjust for differences between the communities or the circumstances, such as business failure, which brought them into the lease market. Even the process of averaging the 18 lease amounts to reach a figure of $11,000 per unit is not disclosed.

C. Capitalization Rate

On this issue, the two appraisers were in substantial agreement.

Mr. Silver used a capitalization rate of 8.75 % which he developed by analyzing six recent sales of CCRCs for which annual lease information was available. Dividing the rental income by the sales price provides a capitalization rate for each transaction. These averaged out to 8.47 % which he rounded to 8.5 %. As he noted in his report,

[s]ales data is somewhat limited for the analysis herein, and the necessary information on net operating incomes from actual sale data is somewhat conjectural. Hence the

7 estimation of a proper overall capitalization rate (R), is, as noted, subject to some appraiser judgment and therefore error.

Ms. Lees used a capitalization rate of 8.5 %. She found only one sales comparable which could provide a capitalization rate. That project had a capitalization rate of 9.0 % and included more than half nursing beds. Since nursing homes command a higher expectation of return to compensate for the higher level of risk when compared to independent living projects, the capitalization rate is higher.

Instead, she turned to published averages in two industry survey reports which collected capitalization rates for assisted and independent living communities and CCRCs. The rates ranged from 8.2 – 9.5 % for CCRCs and 8.3 – 9.8 % for assisted and independent living communities with a trend towards an increase in capitalization rates. She selected a capitalization rate of 8.5 % before real estate taxes. She adjusted the rate upwards by the tax rate of 1.6994 to reach a “tax loaded capitalization rate” of 10.1994 %.

The two capitalization rates are essentially equal. The court accepts as credible Ms. Lees’ capitalization rate for purposes of her income based appraisal. Mr. Silver’s capitalization rate is equally valid, but the court has found his income figures to be insufficiently credible.

D. Business Extraction Rate

Ms. Lees’ report had an additional feature not present in Mr. Silver’s report. Because Mr. Silver sought to identify comparable levels of rent paid to owners of CCRCs by operating companies, he did not have to separate the egg and distinguish between income related to the real estate and income related to the business. Since Ms. Lees used actual financial results from Wake Robin, she was compelled to separate these two elements. An income-based valuation which included revenue from sources other than real estate would skew the value for the real estate unfairly upwards.

It makes no difference to Wake Robin residents whether they are paying into the category of “rent” or “business income.” They receive a bundle of rights and services, including the right to live in their independent living unit, to receive health care, and to take meals together, when they pay their fees. So far as the record of this case shows, the Wake Robin Corporation itself does not allocate its income between rental and business revenues. Accordingly, it fell to Ms. Lees to make an allocation.

At its heart, the business extraction rate seeks to distinguish between the fair market value of the real estate and the “going concern” value of the business. The “going concern” value includes the real estate plus intangibles such as “marketing and management skill, an assembled work force, working capital, trade names, franchises, patents, trademarks, contracts, leases, and operating agreements.” Appraisal Institute of Real Estate, The Dictionary of Real Estate Appraisal, 3d (1966), p.44 (Plaintiff’s Ex. 34). For example, a successful restaurant in a leased

8 space may have minimal “real estate” value beyond the right to assign the lease. It will have significant “going concern” value based on the combination of management skill, employee selection and training, reputation and advertising and other factors.

All of these “going concern” values are also present in the case of Wake Robin which has been developing management and staff, advertising, and developing an intangible reputation since 1996. These values are as much a factor in generating NOI as the availability of acreage and buildings.

In an industry which is “service-rich” like the CCRC industry, the intangible “going concern” value is especially prominent. A warehouse or plain vanilla office suite may have little value to anyone as a “going concern.” These are highly fungible properties. A successful retirement community, filled with trained employees and their supervisors, which offers a wide variety of services generating NOI, will have a “going concern” value which is much greater than the real estate alone.

In seeking to extract the “going concern” value from Wake Robin, Ms. Lees turned to the HUD guidelines which apply to HUD-guaranteed financing of senior living projects. The HUD projects do not include CCRCs. HUD supports financing of projects for lower income projects. Nevertheless, the principle of extracting business revenue not related to real estate and excluding it from the appraisal process applies as much to a HUD project as it does to this municipal tax dispute.

Ms. Lees reduced NOI by 10 % to reflect the proprietary, non-real estate element of Wake Robin’s earnings. She based this reduction on a HUD memo which issued in 2006 and is in evidence. The memo states in relevant part:

[Another HUD publication] suggests Proprietary Earnings percentages between a low of 5 % to 10 % for Board and Care, up to 15 % to 25 % for a Nursing Home. However, it must be emphasized that these are only “general guidelines” and based on the level or levels of care to be provided, market conditions, etc. the actual proprietary earnings may be higher, or lower, than the general guideline range(s). While HUD offers several potential methods of determining the “Real Estate Only” value … it is the responsibility of the appraiser to determine and use the method that provides the most reliable, best supported value indication. (emphasis in original).

The HUD memo states the problem and looks to the appraisers for a case-specific solutions. As this case demonstrates, a case-specific solution is elusive. Ms. Lees’ appraisal avoids the problem by selecting 10 % -- the high end of the range for the now-defunct category of “board and care” homes for the elderly.

On its face, it is improbable that 90 % of fees of $4,000 – 5,000 per month and $400,000 – 500,000 at the time of entry pay for a relatively modest apartment or duplex home and that 10 %

9 is enough to pay for lifetime of health care, meals and other services. These income-producing activities depend heavily upon elements of “business value” such as staff training and experience. As business income is excluded from real-estate related revenue, NOI is reduced and the value for the real estate element of the property decreases. A 10 % extraction rate – at the low end of the HUD range – results in a higher value for the real estate than an extraction rate at the high end.

In the course of testimony, however, Ms. Lees was able to provide some practical support for her 10 % extraction rate. Her table of projected revenues and expenses at p. 95 of her report shows income ($68,119) and expense ($45,085) per unit. After all the health care ($11,939 per occupant), dining ($6,675), administration ($10,854) and other expenses are paid, the remaining unallocated income is $22,500 or $1,875 per month. In other words, on average each resident pays $1,875 per month for some value not otherwise covered in the itemized expenses. With the extraction of an additional 10 % from NOI, the final figure per occupant is $20,250 or $1,687 per month. One plausible explanation for what this pays for is “rent.” (Since Wake Robin is a not- for-profit which seeks to keep income and expense in balance, this money is not “profit.”)

Although this discussion demonstrates that the 10 % extraction rate is used in valuing retirement communities and nursing homes, the criticism remains that it is an essentially arbitrary figure which bears little relation to the actual experience of Wake Robin. It meets the needs of the HUD administrators, but it is hardly an intellectually satisfying answer to the question of how to separate real estate income from business income in this particular case.

X. Findings Related to Comparable Sales Valuation

A. Silver Appraisal

Mr. Silver cast a wide net in commencing his comparable sales valuation. He identified 116 senior housing facilities which have sold in recent years. They varied so much in size, quality, type of service, and conditions of sale (many sales occur in distressed circumstances) that they were almost useless as comparables. Thirty-three of these properties were CCRCs in one form or another. Mr. Silver adopted the unit/bed (accommodation for a person or couple) as the unit of comparison. Even within the CCRC group, the sale price per unit/bed ranged from $52,000 to $254,000. Clearly some additional filter would be required before a useful comparison could be made to Wake Robin.

Mr. Silver narrowed the CCRC properties to 19 and then to 8 properties which more closely resembled Wake Robin. Within each of these properties, he allocated the sales price between the land and the improvements and among the component parts such as independent living units, assisted living, memory care, and skilled nursing units. From the perspective of objectivity, the allocation of component values is one of the less credible steps in the analysis. Mr. Silver recognized this shortcoming in his report:

10 The contributory values assigned to the landholdings were made on the basis of economic potential, topographic features, locational influences and other variables as well as market reactions for unimproved land. Likewise, the allocations of value to the improvements were based on size, condition, utility and particularly unit mix. It is readily admitted that the following allocations appear to be arbitrary; however, in light of the collected data itself, the several contributory value analyses found herein are felt to be reasonable and realistic and to reflect typical buyer-seller reactions as found in the marketplace.

Mr. Silver valued the independent living units most highly. Component values declined as the level of care increased. Nursing units are both physically smaller and require far more expensive services from staff than an independent living apartment. Accordingly, they received lower values in any comparable sale. Mr. Silver then made a series of adjustments to each of the eight most comparable CCRC transactions. These included the quantitative adjustments described above and adjustments for conditions of sale such as date (distance in time from the 2008 market decline) and location (suburban and rural communities valued more highly than urban). In the end he developed values for five components as well as a value for the building site (the campus) and the surrounding undeveloped acreage.

The court finds that Mr. Silver’s process of abstracting eight comparable CCRC sales from a far larger body of retirement facility sales and then breaking down the component parts to identify values for each component is credible. He was candid about the difficulties in allocating the sales price among the component parts. These allocations, however, appear reasonable based on testimony about the relative value and risk of the units devoted to higher levels of care.

B. Lees Appraisal

Although Ms. Lees’ preferred method was her income-based valuation, she also performed a comparable sales valuation. She identified five CCRC properties which were sold in 2009 – 2010. (One transaction did not actually close but it was included because a second payer closed later for a slightly higher price.) Total sales prices ranged from $9.5 million to almost $70 million. Average price per unit (without allocation among the different types of units) ranged from $64,655 to $241,463. Ms. Lees recognized the diversity of her comparables with a certain amount of understatement:

As indicated, the overall degree of comparability of these sales is difficult due to variances in business operation.

She applied a series of adjustments to exclude personal property and to account for distressed sales (at least three of the five), location, age and condition. These adjustments brought the properties into a narrower band of values. Per unit/bed sales prices ranged between $123,653 and $213,038 with an average sales price of $170,000.

11 The percentage adjustments necessary to bring these divergent comparables into line with one another were relatively high. Total adjustments to the comparable sales prices were 40 %, 30 %, 95 % (!), 70 %, and -15 %. A high level of adjustment indicates a high degree of difference between the comparables and the subject property.

Multiplying 295 total units by an average sales price of $170,000 per unit resulted in a sales comparison value of $50,150,000.

Ms. Lees also calculated a sales price per square foot. She used $125 per square foot of building area which she drew from a range amongst her comparables of $60 to $197. This calculation gave her a value of $52,093,500.

Finally, Ms. Lees used a variant of the income approach by making use of a “gross income multiple” of 3.00 which was published in the Senior Care Acquisition Report (2010). Multiplying total revenues of $18,828,000 by 3 resulted in a gross income multiple value of $56,484,000.

Ms. Lees averaged the three figures to reach a sales comparison value of $52,900,000.

The court finds this sales comparison valuation to lack credibility. The essential shortcoming is that the five comparable properties are too different from Wake Robin to serve as meaningful comparisons.

Three of the comparables represent financial bloodbaths in which properties recently constructed for $51,000,000 (Capstone Village), $33,000,000 (Sherburne Commons), and $46,000,000 (Covenant of South Hills) sold for $23,500,000, $9,500,000 and $15,000,000 respectively. Due to problems with occupancy and other factors, these three properties required total adjustments of 40 – 95 % in order to present in a comparable manner to Wake Robin. The other two comparables required total adjustments of 70 % and -15 %.

The court finds that the lack of credibility is not due to any fault or shortcoming on the part of the appraiser. Ms. Lees struggled to provide guidance in a highly specialized market with few sales and too many variations in size and composition. She stated from the outset that the lack of strong comparables made it difficult to employ the “sales” approach to valuation.

XI. Findings Related to Replacement Cost

Neither side relied heavily on replacement cost as a measure of valuation. Ms. Lees did not find that approach to be useful. Mr. Silver provided a replacement valuation, but he testified that it was not as reliable as the cost approach in this case.

Mr. Silver calculated that the cost of replacing the improvements in new condition would be $62,200,000. He used a standard estimating publication provided for the appraisal industry by Marshall & Swift Publication Co., Los Angeles, California. The more subjective calculation

12 was depreciation, both physical and through obsolescence. He estimated total accrued depreciation to be $21,500,000 based on rates of physical depreciation which ranged between 5 % to 35 % for the structures, excluding the tennis courts. Total value of the improvements after depreciation was $40,702,512. To this figure he added $150,000 for improvements to the site such as roadways and parking lots and $1,360,000 for land value.

The court finds Mr. Silver’s opinion about the components of his replacement cost analysis to be credible. Although the land was originally purchased for more than $2 million, his opinion that the purchase price was unusually high for Shelburne at the time and remains ahead of the market was supported by other comparable sales of unimproved land. He derived the construction costs from a standard publication which is accepted in the field of appraisals. The court finds that the current replacement cost figure is credible.

The use of replacement cost as a primary measure of fair market value suffers from the arbitrariness of the allowance for depreciation. The replacement cost plus the land value is approximately $10 million more than the Town’s value and $20 million over Wake Robin’s figure. Obviously buildings which have been in place for 17 years are not worth as much as new buildings. Mr. Silver found that the buildings have been carefully maintained and that there are relatively few portions of the structure which are physically or functionally obsolescent. To compound the problem, Wake Robin added a substantial community of additional cottages in 2007 which are much less affected by depreciation.

Because of the variety of buildings, Mr. Silver was compelled to use a range of depreciation figures which ranged from 50 % to 15 %. At the same time, he estimated that the various elements had an economic life of 40 – 50 years, depending on the building and the date of construction. These adjustments create a gap between the cost estimate and fair market value. Mr. Silver agreed in his testimony that the cost estimate is most accurate when applied to new or almost new construction. The cost of the land plus the cost of the improvements should be very close to market value at the time a CCRC opens. Over the course of 17 years – the length of time the majority of the Wake Robin buildings have been in place – depreciation becomes an increasingly inexact measure of the decline in market value. Particularly with buildings which are maintained to a high standard and have many decades of remaining useful life, the original construction cost reduced by a relatively arbitrary depreciation schedule is a poor guide to current market value.

XII. Owner’s Opinion of Value

A representative of Wake Robin also provided an opinion as to the value of the property. Mr. Erdman is a skilled financial manager, but he has no particular expertise in property valuation. For this reason, the court has not found his opinion as to value to be credible and has turned instead to the testimony of the two experts provided by the two sides.

13 CONCLUSIONS OF LAW

SUMMARY

There are three principal methods for determining the value of real estate. These are the income capitalization method, the comparison of comparable sales of real property, and the determination of replacement cost less depreciation. Each has its own well-recognized shortcomings.

Both appraisers used the income and comparable sales method. Mr. Silver also used replacement cost. The court has considered the relative credibility of these estimates. “Credibility” in this context is derived from the reliability of the underlying data and the persuasiveness of the methodology. It has nothing to do with personal credibility. Both appraisers were honest and professional in their approach to a difficult valuation problem.

The court reaches the following conclusions:

1. Mr. Silver’s income analysis is not fully credible because the income figure is not derived from a reliable source.

2. Ms. Lees’ income analysis is not fully credible because the “business extraction” percentage is arbitrary and unsupported by the evidence.

3. Mr. Silver’s replacement cost analysis is not fully credible because the depreciation schedule is a poor predictor of fair market value.

4. Ms. Lees’ sales analysis is not fully credible because her 5 comparable properties are an inadequate basis for developing an estimate of fair market value.

5. Mr. Silver’s sales analysis is more credible than Ms. Lees’ sales analysis because he considered a wider range of possible comparables, developed a greater number of similar transactions, and applied more specific and detailed adjustments.

6. The court accepts Mr. Silver’s estimate of $40,000,000 as the most credible opinion of fair market value.

ANALYSIS

The court’s task in this de novo hearing is to compare and contrast the parties’ use of these three methods and to select the method which is analytically strongest. The valuation of a property which has not been sold recently is necessarily a process of estimation and projection. It is what historians call a “counter-factual.” Wake Robin is not for sale and because it is successfully fulfilling its not-for-profit mission, it is highly unlikely to be sold in the conceivable future. But just as we might ask what would have happened if Hitler had not invaded Russia, we are required

14 in this case to ask what would happen if the Board of Directors voted to sell Wake Robin to the highest bidder on April 1, 2010.

As the Findings show, the parties have provided two answers. These answers are about 20 % apart. It could be said with a high degree of certainty that the value of the real estate at Wake Robin is about $40 – 53 million. For purposes of tax valuation, a single number is required. The short-cut of averaging the two figures would only result in a third figure with no decision as to the relative credibility of its components. The court is required to pick the stronger method based on the reliability of the data relied upon by the appraisers and the inherent strengths and weaknesses of the three appraisal methods.

I. Use of Replacement Cost

Neither appraiser relied heavily upon the replacement cost of a 17-year-old facility as a measure of market value. Ms. Lees did not use the method at all. It is fair to say that Mr. Silver used it primarily as a check on the reliability of his other methods. The difficulty lies in the depreciation schedule which is inherently arbitrary in the sense that it is not derived from the experience of a particular property or a particular community. In isolation, the schedule leads to the conclusion that all real estate except unimproved land loses value steadily over time – a loss which is offset only by inflationary increases in replacement costs. This is not a compelling explanation of why property values change and often increase. In fact, neither side contends that Wake Robin has lost value at all. Even allowing for the construction of additional housing in 2006, both sides believe that the value has increased over time. The only issue in dispute is how much. The court rejects replacement cost as the basis for its de novo determination of value.

II. Use of the Income Capitalization Method

The income capitalization method depends upon a simple economic assumption. A market made up of rational investors will establish value by considering two factors: the net income generated by the property and the rate of return required by an investor at the level of risk associated with the investment. As the return on an investment becomes more uncertain, the capitalization rate increases because investors demand a ratio closer to 1:1 between income and property value. For less risky investments, the same investor pool will accept a more distant ratio and a lower capitalization rate.

This process for determining value has long been accepted by courts in Vermont. As the Vermont Supreme Court has explained:

The income approach is based on the proposition that a rational investor would pay the fair market value for a piece of property, which is the price (P) that, when multiplied by the rate of return available from alternative investments of comparable risk (the capitalization rate or R), is equal to the property’s expected net income (I). In other

15 words, if the known factors are capitalization rate and net income, the price of the property may be calculated by dividing the net income by the capitalization rate: P = I/R.

Beach Properties, Inc. v. Town of Ferrisburg, 161 Vt. 368 (1994). In order to be useful in determining market value, the income capitalization method requires a high degree of accuracy for the two variables in the equation: the capitalization rate and the net income figure. As the Vermont Supreme Court observed in Beach Properties, “[r]elatively small variations in the [capitalization] rate” and “relatively small adjustments to net income” will result in significant changes in fair market value. Id. at 373. If these values are wrong – even a little – the determination of fair market value will be significantly incorrect. Applying the formula of FMV = net income divided by capitalization rate is like driving a car with a small steering wheel at a high speed. A slight movement of the wheel results in a dramatic change in the location of the car.

A. Capitalization Rates

The two appraisers have no significant disagreement over their respective capitalization rates. Both have determined capitalization rates through a process of judgment and intuition which eludes objective analysis. Mr. Silver relied on a handful of reported transactions to compute a capitalization rate of 8.75 %. Ms. Lees relied upon an industry report (The Senior Care Acquisition Report 2010) and passed over an average rate for CCRCs of 8.2 % to settle upon 8.5 % based on a trend towards higher rates. If the capitalization rates had differed in some material way, the court would have great difficulty deciding which method provided the more reliable capitalization rate. As the case has developed, however, the two capitalization rates are within .25 % of one another. They are both estimates of equal reliability. The court will accept each for purposes of analyzing the respective results of the two appraisals.

B. NOI Attributable to the Real Estate

In the Findings section, the court has already criticized the construction of an income stream by Wake Robin’s appraiser Mr. Silver based on 18 leases of CCRCs by REITs to operating companies. The court found that the average rent paid per living unit for these properties did not provide a credible estimate of NOI for Wake Robin.

The Town’s appraiser Ms. Lees has used Wake Robin’s own income numbers and there is no substantial disagreement between the parties over the calculation of gross revenues or NOI at Wake Robin. Ms. Lees added an additional 5 % annual expense for a management fee. This brings the expenses for Wake Robin closer to the expenses which a rational investor would expect to bear if he or she purchased the property. It also results in a lowering of Ms. Lees’ calculation of market value and is therefore not contested by Wake Robin.

The area of disagreement concerns the “business extraction” rate. The Town taxes only the real estate portion of Wake Robin. Since Wake Robin makes money through renting homes to its

16 residents (real estate income) and providing benefits such as meals and health care (income derived from a mixture of real estate and employee services) and the promise of future health care, an adjustment must be made to remove the business-related income from the equation. If this is not done, we would be valuing the entire business as a “going concern” which would overstate its value for real estate tax purposes.

Ms. Lees’ selection of a 10 % business extraction rate is too arbitrary and unsupported by objective evidence to provide much assurance of its accuracy. It flies in the face of the experience of a person living at Wake Robin. The facts are undisputed: an elderly person arrives and pays several hundred thousand dollars in a lump sum and several thousand dollars more per month to live in a simple apartment or duplex. If asked why they would make such a decision, few would say, “Well, 90 % of the value for me lies in the chance to rent this attractive apartment in a semi-rural setting.” Many would point to the sense of community, established through meals and recreational activities enjoyed together, or to the security of guaranteed nursing care on-site. These are values derived from services and business activity as much as real estate. Meals, for example, require a physical location which has value as real estate. Skilled nursing also requires a physical structure. But both services require planning, hiring, direction and other entrepreneurial values associated with an ongoing entity which are entirely separate from the value of the real estate.

The percentage of income excluded from NOI because it relates to the non-real estate business activities strongly affects the valuation. A business extraction factor of 30 %, for example, would reduce “income to the real estate” from $5,597,190 to $4,353,380. Using Ms. Lees’ tax- loaded capitalization rate of 10.199 % (including the percentage for real estate tax), the market value falls to $41,564,380 (after subtracting personal property). This is a drop of over $12 million dollars or 22 %.

The evidence at trial included a less dramatic shift in the business extraction rate which highlights the arbitrary quality of the equation. In a preliminary version of her report, Ms. Lees employed a slightly lower figure for NOI ($5,991,250) and a 15 % business extraction rate. She used a 9.5 % capitalization rate. She made a computational error in reaching a tax adjusted valuation of $53,496,169 (before deducting for personal property). The actual figure was $45,473,372. Had Ms. Lees used the 10 % business extraction rate, she would have calculated a value of $48,150,000. Using 15 % instead of 10 % as the business extraction rate increased the valuation by nearly $3 million. At either a 10 % or a 15 % business extraction rate, the result was significantly below the 2008 - 2010 appraisal values already set by the Town. In her final version, issued 11 days later, Ms. Lees reduced the business extraction rate to 10 % and reduced the capitalization rate to 8.5 %. She increased NOI by $230,000. These adjustments increased the appraised value to $54,878,000 (before subtracting personal property).

This unsettling episode demonstrates the extent to which the results of the income method vary with small, essentially arbitrary changes in the variables in the equation. Every 5 % increase in

17 the business extraction rate results in an added value to the real estate of almost $3 million. There are no objective standards for determining this rate. The business extraction rate is lower in a “pure” real estate transaction such as the rental of an unfinished apartment. It increases with the level of service since more “business” is being conducted in addition to the rental of living space. Because the business extraction rate for CCRCs can be adjusted by the appraisal community within the broad limits set by the HUD memo of 5 – 25 % depending upon the degree of personal care provided, the income method is less reliable in this particular factual setting than the “market comparable” method.

III. Use of “Market Comparables”

The market comparable process as applied in this case also has a significant deficiency. Its shortcoming is the lack of strong market comparables and in consequence the need to adjust the few available comparables very highly. Instead of the relatively small adjustments for differences in square footage, condition or location which marks the appraisal process for single- family homes, the commercial appraiser looking at unusual properties in a market with few arms-length transactions has to apply considerable pressure to his comparables in order to bring them into line. Mr. Silver’s appraisal is such a case.

Despite these reservations, the court concludes that the “market comparable” approach is more credible in this case. It is based on actual sales of properties with functions which are very similar to Wake Robin’s. CCRCs come in different styles, different locations, and are marketed to different demographic groups. But the evidence in this case demonstrates that they have as many similarities as differences. The housing units are consistently modest in size. These residents have already raised their families and mowed their lawns. The expression of individuality through design and outward show is no longer so important. The selling points are security, ease of life, quality of care, and companionship. These produce real estate developments which are similar in function. Each CCRC is committed to some form of future care in the event of a decline in health. Each provides some form of communal dining. Like college dormitories, the CCRCs may look very different on the outside – and some have more appeal than others – but functionally each CCRC is intended to answer a very similar need for housing and care at a specific moment in a person’s life.

With these similarities in mind, the court accepts Mr. Silver’s survey of property sales within the industry as more reliable than either party’s income method of appraisal. It requires fewer loosely-justified assumptions. It solves the “business extraction” problem by relying upon sales of real estate only to REITs which can buy nothing else. It avoids the potential error inherent in estimating a capitalization rate. Most importantly, it is directly tied to the actual experience of buyers and sellers for the type of property at issue in this case.

Mr. Silver’s final cut of eight recent sales included prices of between $10,500,000 and $85,000,000 and between 91 and 760 units. (The figures are rounded for ease of discussion.)

18 The average price per unit fell into a narrower band (a low of $115,000 and a high of $150,000). None had exactly all of the features and elements of Wake Robin, but all were broadly similar in providing assisted living, skilled nursing or dementia care. By definition, all provided independent apartments, cottages, or both.

Mr. Silver went to considerable effort to break out 13 areas of adjustment including qualitative adjustments such as location and quantitative adjustments such as numbers of apartments or acres of land. The qualitative adjustments are the more subjective – these resulted in minimal adjustments. The quantitative adjustments were objective in the sense that they reflected the numbers of apartments, skilled nursing units, and other components of the property.

Exactly how much money each of these elements contributed to the sales price was less certain. By isolating and comparing the elements of the sales, Mr. Silver was able to identify a range of values contributed by the various types of housing and care units. There is considerable subjectivity in this phase of the analysis since buyers do not specify how much they are paying for each element. What, after all, do we pay for the engine in our new car? Or the trunk lid? But by comparing prices for properties with more or less of a given type of unit, Mr. Silver fairly identified the contributions to sales price of increases in the numbers of each class of units.

The eight comparables come as close as possible to establishing the price likely to be paid by the same market of investors if Wake Robin were to be placed on the market. There is risk of error in any one of the adjustment categories, but breaking the adjustments into so many specific areas reduces the risk of a large error. The estimated adjustment for one element of a single comparable property may be incorrect, but that error can be offset by the numerous other adjustments. The court concludes that Mr. Silver has fairly created an accurate picture of the market for Wake Robin by locating a range of comparable transactions and making multiple, incremental adjustments which are based on largely objective and measurable qualities characteristics such as the number of various types of units.

For these reasons, the court concludes that the evidence presented by Wake Robin is more credible than the evidence presented by the Town on the issue of valuation.

CONCLUSION

The court orders that the fair market value for property tax purposes of the Wake Robin property is $40,000,000 prior to tax equalization and $39,200,000 after tax equalization.

Dated: January 14, 2013

___________________ Geoffrey Crawford, Superior Court Judge