This paper aims to analyse the appraisal of a specialised form of real estate - data centres - that has a unique blend of locational, physical and technological characteristics that differentiate it from conventional real estate assets. Market immaturity, limited trading and a lack of pricing signals enhance levels of appraisal uncertainty and disagreement relative to conventional real estateassets. Given the problems of applying standard discounted cash flow, an approach to appraisal is proposed that uses pricing signals from traded cash flows that are similar to the cash flows generated from data centres. Based on the law of one price, it is assumed that two assets that are expected to generate identical cash flows in the future must have thesame value now. It is suggestedthat the expected cash flow of assets should be analysed over the life cycleof the building. Corporate bond yields are used to provide a proxy for the appropriate discount rates for lease income. Since liabilities are quite diverse, a number of proxies are suggested as discount and capitalisation rates including indexed-linked, fixed interest and zero-coupon bonds. The application of conventional discountedcash flow approaches to the appraisal of data centres requires information about a wide range ofinputs that is difficult to derive from market signals or estimate analytically. In practice, discounted cash flowappraisals of data centre facilities are forced to incorporate non-market assumptions that are inevitably subjective. Cash flows from data centres tend to be more complicated because ofthe high levels of current and capital expenditure compared with conventional assets. This requires an understanding of the appraisal ofliabilities as well as income. Whilst the use of price information from similar cash flows such as corporate bonds is helpful, there are rarely assets and liabilities that have identical cash flows and risks and some approximation isnecessary. The digitalisation of business and society has produced a different form of real estate asset that is specialised, physically complex and whose operation, maintenance and management require specialist staffing and high levels of current and capital expenditure. The complexity of the cash flows and the lack of active trading create significant appraisal problems. This paper proposes an alternative approachto appraisal that utilises external pricing signals to appraise the various incomes and costs.
In the late 1990s, the boom in the Telecommunications, Media and Technology (TMT) sectors raised theprofile and supply of data centres . The surge in the TMT sectors was mirrored by an expansion in thesupply of data centres, where operators of facilities leased cabinet, cage and floor space. As the TMTbubble burst, many data centre operators and providers went into liquidation resulting in an oversupply of space.The subsequent period has witnessed market recovery, associated with consolidation within the sector and an increased recognition of the importance of business continuity and security. A large proportion of data centres have some of the characteristics of conventional investmentproperties in that they generate rental income flows that are underpinned by occupational agreements. However, the buildings are specialized and require significantly more current and capital expenditure over their life cycle compared to conventional assets. This presents a number of appraisal problems which are the focus of this paper.
The remainder of this paper is organised as follows. Firstly, itprovides an analysis of data centres as a specialised form of real estate and discusses its unique attributes as an asset class. The appraisal issues for specialized or limited market properties are discussed. Particular attention is paid to the informationrequirements of discounted cash flow methods and the problems of identifying defensible inputs from market activity and analysis....