Towards a common standard?

The development of regional and global real estate investment markets has created pressure for improved prime market data quality. Despite substantial investment and clear improvements, hurdles remain: data continue to be collected according to either local or company methodologies; and data pooling remains the exception, rather than the rule. Often, market conventions and business structures have frustrated change. In addition, most advisory firms have uneven market coverage; as a result data quality often varies. Further, the core role of data providers as market advocates may create at least a perceived conflict of interest.

Most fund managers and investors deal with these challenges through multiple data subscriptions, subscriptions to independent research companies and in-house expertise. As a result, data are typically subject to expert criticism and adjustment. However, although this improves quality, the collection of essentially identical data by multiple advisory firms leads to cost duplication, which is ultimately passed on to investors. This paper illustrates some of the challenges and suggests a potential solution.

On methodology there are several reasons for variation. For example, while some providers define prime rents as the highest rent achieved for a defined unit, others require a minimum number of transactions to justify a change in level. Further, while some adjust transactions with atypical incentives, others do not. Similarly, while some yields are quoted on a ‘gross' basis others are ‘net' or a hybrid of the two.

These methodological differences impact analyses. Typically, assessments of available future returns are based on analysis of longitudinal and cross-sectional relationships between variables such as prime rent, GDP growth, interest rates, employment, vacancy and development activity.

The importance of longitudinal consistency can be illustrated by considering approaches to the assessment of natural or ‘equilibrium' vacancy rates (NVR). The NVR is the rate at which rental values are expected to neither rise nor fall. However, the suggested NVRs are only valid when considered in the context of their source data. For example, while Jones Lang LaSalle (JLL) data suggest an NVR for Central London offices of 6%, DTZ data suggest NVRs of between 9% and 7.5%. Similar challenges are presented by other data differences.

Turning to cross-sectional consistency, by changing methodology it is possible to answer the question "Which city has the largest office stock, London or Paris?" in three equally valid ways: first, that Paris is 80% larger; second, that they are both the same size; and finally, that London is 67% larger. As with rents and yields, methods used vary by data provider.

A similar point can be made with vacancy data. According to CB Richard Ellis, at the end of 2004 vacancy in Amsterdam was 16.8% and in Frankfurt it was 10.9%. However, the Amsterdam data include sub-let space, whereas in Frankfurt the practice is to exclude such space. Adding sub-let space into the Frankfurt number would have increased vacancy to 14.8%. Amsterdam data also include developments due to be completed within the next year, while such space is excluded from the Frankfurt data. Adding this space increases Frankfurt vacancy to 15.2%. As a result the difference suggested by the original numbers has narrowed due to methodology, not market characteristics.

Turning to deviations due to opinion, it is clear that although initiatives such as ImmoStat in Paris and the Central European research forums have reduced scope for deviations due to access to data, imperfect information remains a challenge in most European markets. Often local businesses with a large market share are reluctant to pass information to their less successful competitors. While this is understandable from the perspective of local businesses, it may be less logical for European or global businesses where advantages with one market are likely to be offset elsewhere. We hope that, as businesses integrate regionally and globally, this issue will become less important.

Although difficult, the issues presented above can be solved by collective action. There are clear precedents for this. The formation of IPD in the 1980s led to high quality performance data for the UK and elsewhere. The formation of bodies such as the IPF, INREV and EPRA demonstrates collective action to promote common goals. Recent collaboration between the IPF, BPF and others on UK REITs is another example of the industry working together.

Perhaps most importantly, the growth of independent and in-house investment research, combined with improvements in the transparency of European real estate markets, has led to the increasing commoditisation of market data. As a result, advisers now add more value by data interpretation than possession. This raises the prospect that major stakeholders might be open to change.

One possible approach involves the pooling of data from a group of advisers based on common definitions. The inputs from all advisers as well as independent and in-house researchers would be collected, analysed and tested for consistency by an independent third party. The combined information will then form the basis of a single set of agreed outputs. Over the short term it is likely that data would be supplied according to both local conventions and an agreed common standard. In the medium term the importance of data based on local conventions should reduce.

This change would be a clear improvement, enhancing the quality of data while providing significant cost savings for both data consumers and producers.

By establishing a group to develop consistent definitions, advisers may have taken the first steps towards the changes outlined above. However, while their profits are supported by a booming investment market, it is unlikely that cost pressures will be sufficient to drive change. When the investment market turns down, as it inevitably will, these pressures are likely to alter.