Using transaction yields to build a database that produces effective risk and return measures for the UK office market
MetadataShow full item record
This thesis creates a database based on transaction yields and transaction prices to construct metrics that enable effective investment research to be conducted on UK offices. Performance indices based on valuations, reflecting annual, quarterly or monthly frequencies have been used in the commercial property industry for decades. There have, however, been longstanding concerns with the use of such indices in the UK and beyond. These have been investigated in a substantial pool of academic literature that also dates back decades. One major issue identified in respect of these indices relates to the smoothing of returns. This reflects the concern that current and previous valuations, particularly in respect of monthly and quarterly frequencies, are highly correlated. Suggested answers produced by academia to date include mathematical tools employed to ‘de smooth’ valuation-based measures. An alternative approach mainly coming out of the USA has suggested using transaction prices, allowing for the small portion of a market that is traded during a measurement period. The literature relating to transaction-based measures, again, spanning decades, has largely settled on using a mathematical process based on interpolating price movements between a series of properties which have been sold more than once – ‘repeat sales’. The use of repeat sales brings its own issues, namely the reduction of the sample size of transactions plus the reliance on an established footprint of a building, that may have been altered, rather than issues relating to leasing terms, for example, being used as the final arbiter of the categorisation of property. This thesis takes headline transaction prices – yield and price – for UK office properties that have been collected on quarterly basis over the period between January 1999 and June 2018. A range of transparent filtration processes are then used to clean the data, producing a database that forms the basis of a range of relevant measures. Data is segmented into a range of geographic categories that the thesis defines as CUKOs (Category of UK offices). These definitions stand as Big Cities, London and Rest of UK. Having defined the CUKOs, the transaction-based data is used to produce variables including weighted yields and the (nominal and real) risk premium. These headline measures are subjected to various statistical tests to determine their validity. The headline measures are then converted into explicit performance metrics, based on the ‘point-to point’ yields adjusted for income growth’ as and measures based on repeat sales covering price change and total return. These are compared with existing appraisal-based counterparts. The metrics are then used practically in respect of quantifying risk and the effective application of portfolio theory to the UK commercial office market. The thesis demonstrates that it is possible to take relatively straightforward data and convert it into measures which are relevant and effective. Given the increasing risk of opacity produced by a combination of industry consolidation and proprietary restrictions on the availability of data, the thesis provides a straightforward and cost-effective method of addressing some of the major issues now facing investment researchers covering real estate in the UK and further afield. It also explicitly demonstrates some of the shortcomings of the repeat-sales methodology and its application to non-London offices.