Agents’ differing definitions of basic terms such as yields and vacancy rates make it hard for investors to compare market data, but there are moves afoot to standardise the terms used
Which is the bigger office market, London or Paris? As with many things in life, the answer is an unhelpful: “it depends”. The same answer could be given if one asks: “What is the net yield on prime offices in Milan?”. Variations in definitions and data consistency mean that property data in European markets often varies between providers, which adds layers of opacity to the market.
Difficulties with source data in turn make forecasting harder and pose barriers to entry for investors – at a time when more and more are keen to get into property. At last month’s IPD Real World research conference in Cambridge, Paul Kennedy and Aiko Sauer of Invesco Real Estate, Michael Haddock of CB Richard Ellis and Ben Sanderson and Kieran Farrelly of Prudential Property Investment Managers highlighted problems with data, looking at yield definitions and problems with transaction and valuation data.
In response to these difficulties, a group of four of the largest data providers have developed a first draft of some market definitions for a number of factors, which they aim to work towards.
Returning to the question of the relative sizes of the London and Paris office markets, if one takes Greater London and the Ile de France areas as the boundaries of London and Paris, then the Paris office market, at 47.5m m², is 80% larger than London’s. However, if the comparison is made instead between the 20 Parisien arrondissements, plus La Défense, and the CB Richard Ellis definition of central London (including Docklands), the office markets are both around 20m m² in size. Finally, comparing the Paris central business district with London’s core City and West End markets shows London’s office market to be 67% larger than Paris’s, at 14m m² and 8.5m m² respectively.
Potential investors comparing the benefits of London and Paris, or trying to fit them into a European or global asset allocation plan, could soon get confused!
Confusion over vacancy rates
Other confusions may arise when looking at vacancy rates in different European cities, CBRE research director Michael Haddock told the conference. In the first quarter of this year, a single agent reported Frankfurt’s and Amsterdam’s vacancy rates to be 16.8% and 10.9% respectively — a clear distinction. But “if we look at the definitions behind the numbers, the difference is less clear,” said Haddock.
In Amsterdam, space available to sublet is included in the vacancy rate, whereas in Frankfurt it is common practice not to include such space. That information is also available for the amount of space available to sublet in Frankfurt. If this is also included, the vacancy rate rises to 14.8% — still lower than in Amsterdam, but now by a much smaller margin.
And in Amsterdam it is also standard practice to include developments due to be completed within the next year but which have not already been prelet, whereas this space would not be included in Frankfurt’s vacancy rate. If speculative developments are also added, the vacancy rate for Frankfurt rises to 15.2% – not all that different from Amsterdam.
Property market data tends to come in two forms, based on either deals or valuations, both of which have particular problems. In the equities markets, transactions are frequent and take place in a public forum — major companies such as BP or Vodafone can see £500m of their shares traded each day in thousands of transactions.
In contrast, even in the largest and most liquid European property investment markets, there are only a few hundred deals in a year (see graph), so market pricing is much harder to calculate accurately and a few major deals can skew the data.
Haddock also pointed out that “the people who are putting individual transaction details into the market have their own agenda. When price information comes into the equation this agenda can become important. A vendor or landlord may want to present the facts to emphasise the price or the rent, whereas a buyer or tenant may want to minimise these figures.” He added that this means transaction data alone cannot be relied upon for analysis of property markets. Valuation and other estimated data — such as that for vacancy rates — is used as an adjunct.
Valuation smoothing occurs because the previous observation of a variable inevitably has an impact on the current observation. So series of valuation data are less variable than if the true values of each data point were known. And valuation data will also lag behind market data, making it tricky to identify the timing of turning points in a market.
Investors looking across a number of different markets in Europe are naturally concerned with investment yields. But comparisons are hampered by clear differences between prime market yield definitions used across European property markets.
Approaches vary by both data provider and local market. Invesco’s Kennedy told the conference: “Uncertainties in definitions are compounded by inconsistent, and sometimes inaccurate, interpretations of the costs associated with both acquisition and ownership.”
The most fundamental difference between yield definitions is at the level of “gross” and “net”. Gross yields, said Kennedy, can be broadly defined as the ratio of gross income (ie income before non-recoverable cost are allowed for) over net purchase price (ie net purchase price excluding other costs associated with the transaction).
In contrast, net yields can be broadly defined as the ratio of net income (ie, income after allowing for non-recoverable costs of ownership) over gross purchase price (ie, purchase price plus other costs of acquisition). In Europe, the difference between gross and net yields varies between markets and between different data providers in the same markets. And the non-specific ‘market yield’ which is often quoted may also differ from either figure.
Furthermore, the difference between the lowest and highest calculation of the net yield in the same market can differ by as much as 0.9%. This is because different agents and markets apply differing levels of the same costs.
In an effort to minimise these problems, a group of major data providers – CB Richard Ellis, Jones Lang LaSalle, DTZ and Cushman & Wakefield Healey & Baker – have worked on a set of definitions for a number of key terms used in the European property industry. They will work towards using these definitions for all data they produce, but will run them alongside existing definitions until the market is familiar with the new definitions.
Gradual switch of definitions
“For example,” says CBRE’s Haddock, “in our central and eastern European offices, lease renewals are included in take-up figures – this will not be allowed under the new definitions, but we will publish both sets of figures for a few years and then switch to only using the new definitions.”
In coming to the definitions, the agents consulted with groups such as IPD’s European research group, German research group GIF and pan-European research group PEPCIG. Definitions are provided for terms such as achieved rent, availability, demand, grade of office, prime rent and stock.
For example, their definition of triple net yield is: “triple-net yield is derived from the net operating income (after deducting all non-recoverable expenditure) divided by total purchase costs (including price, costs and taxes).
Take-up is defined as: “the total floor space known to have been let or prelet, sold or pre-sold to tenants or owner-occupiers during the survey period. It does not include space under offer. A property is deemed to be ‘taken-up’ only when contracts are signed or a binding agreement exists.
“Prelet refers to take-up that was either in the planning or construction stage. All deals (including prelets) should be recorded in the period in which they are signed. Contract renewals are not included. Sales and leasebacks are not included, as there had been no change in occupation.”
Invesco’s Kennedy suggested that a solution to the problem of differing data from different agents could be solved by setting up a central data collection agency to which agents all submitted their data.
However, Haddock argued that the provision of data, as well as its analysis, was still an important competitive advantage for agents with a significant research function. He also said the set-up and ownership of such an agency could cause difficulties.
For a complete list of definitions contact michael.haddock@cbre.com; hela-henrichs@eu.jll.com; joanna.tanno@eu.cushman.com
