Learning from the real estate profession: factors in industryded research

Past research have rarely combined what this book has identified as macro- and micro-factors so far. Due to the need to position the research within a complex

Factorisation of commercial real estate 29 theoretical framework (APT for macrovariables or HPM for micro-factors), academic studies customarily focus on one set of factors only. In that sense, Lecomte’s (2007) combinative and factor-based hedge presented before are innovative insofar they aim to combine both families of factors into one single derivative instrument.

The real estate industry has not waited for academia to conduct research on the nature of commercial real estate risk. This section highlights three such research conducted by industry researchers and academics on UK commercial real estate assets.


In 1999, the Investment Property Forum (London, UK) commissioned the Investment Property Databank (IPD) to undertake a major survey aimed at “identifying the ways in which risk is understood, assessed and managed within the UK property industry”. Based on 124 detailed responses from property investors and advisors, the research (IPF, 2000) identifies 1,590 controllable property specific risks, which are then narrowed down to 57 different aspects of risk at the asset level, at the portfolio level, and in a mixed asset portfolio. Similarly, 20 approaches to risk management are singled out. These 57 risk factors range from cyclical synchronisation through the presence of deleterious materials? Since many of them are not easily quantifiable, they cannot be included into a risk model. However, “this hardly counts as a simple model of risk”.

Indeed, IPD outlines that applying conventional models (i.e. CAPM’s dichotomous decomposition of risk into market risk and specific risk) does not work for managing the specific risks of commercial real estate. Actually the CAPM is only one of 23 ways for formally assessing or quantifying relevant risks (top 5) mentioned in the survey. Market risk can be disaggregated by component: national non-property, national property, and local property. IPD calls for the development of a property multi-factor risk model which encompasses “all three of the market risk sub-categories [...]”, knowing that “a local factor component would perhaps be of greatest value to property investors”. With respect to specific risk, IPD asks for “a much tighter measurement framework that is designated to operate [...] at the level of the individual asset rather than one drawn from conventional theory which operates primarily at the portfolio level”. This, IPD acknowledges, “remains in the land of property research science fiction”.


Under the umbrella of a joint LaSalle-IPD project, Blundell, Fairchild and Goodchild (2005) map out “a practical way for managing risk in property portfolios” by identifying factors causing volatility in property returns. These factors which have to be managed to control volatility of commercial property returns are categorised into two groups depending on their role in triggering risk for investors:

four fundamental factors and several modulators which have a dampening or intensifying effect on the former. Unfortunately for the analysis conducted here, the focus is on ex ante risk at the portfolio level, not single assets. Factors are “based on decomposing the causes of volatility”. They write:

Like attribution analysis of return, overall portfolio volatility can be broken down into component parts relating to changes in income (default, re-leasing and rent reviews) and capital (shifts in capitalisation rate and market rent). These changes are driven in turn by a range of factors partly shared in common.

Fundamental risk factors include: tenant credit worthiness (TICCS), exposure to volatile sectors (e.g. City offices), speculative development exposure, weighted lease length, expected income growth, yield level (low is riskier), void rate. By the same token, modulators are: number of properties/ lot size concentration, property type concentration, sector balance, location concentration, leverage, exchange rate mismatch, tenant concentration, lease expiry concentration. Factors can act both individually and in combination. Each fundamental risk factor is given a score which, correlated with the other risk factor scores, defines the Blundell Risk Web (equal weights).

Notwithstanding its focus on portfolios rather than individual buildings, the Blundell web framework epitomises a few interesting ideas for factor-based real estate hedges. First, it does not include any of the macro-factors customarily mentioned in the academic literature. Second, it eschews direct references to buildings’ physical dimensions, by focusing instead on their cash flow generating abilities in the money-time realm.


In 2015, the IPF commissioned a group of researchers to study “the measurement and explanation of investment risk at the individual property level in the UK commercial market”. The research analyses the performance records, property characteristics, and tenancy records of over 1,000 commercial properties held over the period 2002-2013 as well as detailed case studies of 88 commercial properties. Market risk dominates total risk in most individual properties except for a quarter of the properties. For the latter, risk is dominated by asset-specific truly idiosyncratic sources of risk. In particular, properties with relatively small lot sizes, higher yields, fewer tenants and greater exposure to the leasing market and to capital expenditure tend to show high specific risk.

The researchers note that “lease-related [i.e., tenant administration and vacancy following lease expiry] and, to a lesser extent, asset management-related factors are the predominant drivers of high specific risk in individual properties”. Macro-economic factors selected in the study (GDP growth surprises, inflation surprises, gilt total return) are not instrumental in explaining the performances and risk of individual properties. The IPF report (2015) concludes by stressing that “in pricing risk in individual properties, investors and researchers need to focus more heavily on the risks related to tenant default, lease events and asset management”.


Industry-led research and academic research are seemingly very different in their approaches to commercial real estate risk. In fact, it is difficult to reconcile the various sources and selection of risk factors presented in this section with those selected in academic papers. One striking dimension in industry-led research is the implicit reference to the urban land economics’ paradigm of real estate. As mentioned before in this book, urban land economists identify three dimensions to real estate: physical, economic, and legal. Macro-factors and micro-factors as initially thought out in Lecomte (2007) after Hoag (1980) cover a property’s economic dimension and physical dimension, respectively, whereas its legal dimension is mostly overlooked in academic models. Notable exceptions are Brennan, Cannaday and Colwell (1984) and Miles, Cole and Guilkey (1990), who explicitly refer to valuation rules.

Conversely, the legal dimension which encapsulates Ratcliff’s “legal content of ownership” is predominant among industry researchers’ recommendations for a multifactor pricing model of real estate. A bottom-up comprehensive approach to commercial real estate risk at the asset level supposes to start with the property, its physical and legal dimensions, and to progressively move up by extending the scope of the analysis all the way to its local, regional, and national environment. For instance, Brennan, Cannaday and Colwell’s (1984) analysis of Chicago CBD office rents is conducted at the office unit level and not at the building level. They posit:

The use of the building as the unit of observation effectively precludes including the date of lease transaction for each office unit within the building as an independent variable.

Therefore, most studies do not include a variable to account for the fact “transaction rental rates on which the average rate is based may have been negotiated at different points in time when market conditions were significantly different”. Brennan et al.’s hedonic model which explains 90% of the variation in the log of rent encompasses: lease features, occupancy rate at the time the lease was executed, physical characteristics of the building, physical characteristics of the unit, location of the building. The complete list of variables is presented in Appendix 1.3.

Considering a building as a set of individual units each with its own lease features and physical characteristics within a building is the ultimate bottom-up approach, which admittedly goes against the mammoth trend of aggregate thinking that gripped real estate finance researchers at approximately the same time. The legal dimension of a property is what modulates the myriad of dynamic relationships between a building and its environment. It acts as a bridge between Graaskamp’s two realms of real estate: space-time and money-time. So what is at play here? Has urban land economists’ legal dimension been willingly overlooked?

First of all, it is undoubtedly a matter of paradigm. There is no denying that urban economics-driven research cannot easily accommodate the messiness of countless risk factors at the property level. Basically, the scientific bias in real estate research and the tendency for aggregate thinking in real estate finance have resulted in a range of studies on macro-variables in line with seminal papers in financial economics. Meanwhile, an even smaller number of research using the HPM have aimed to identify property-specific factors or micro-factors. The HPT’s reliance on human subjectivity partly discredits hedonic variables in the eyes of quantitative research. In the process of defining factor models after the APT, real estate finance researchers have overwhelmingly forgotten about commercial real estate’s legal dimension which has been abandoned to appraisers’ and surveyors’ valuation rules. However, these rules position commercial real estate risk directly into the money-time realm at the most granular level which is exactly where an asset risk should be assessed. Secondly, it is a matter of access to data. Very few researchers have access to the wealth of granular data and property level information accessible to the IPF research team in their 2015 study of individual property risk. The lack of access to data serves as a reinforcing feedback to the scientific bias in real estate studies, which ends up promoting a fragmentary and reductionist view of real estate risk.

< Prev   CONTENTS   Source   Next >