Portfolio: Asset-specific risk

New research highlights the need to price individual property risk more systematically and with a greater range than in the past, writes Paul Mitchell

New research undertaken for the Investment Property Forum’s 2010-2015 Research Programme provides investors with insights on how risk in individual properties should be priced, on the structuring of property portfolios, and for investment processes and strategies in general.

The analysis draws on the detailed records of 1,000 UK properties held by eight big investors since 2002 and on case studies of 88 of these properties. 

The approach distinguishes between specific risk in individual properties and systematic risk. Specific risk is unique to the asset and independent from one property to another and hence is a risk which, when combined with other assets in a portfolio, can be diversified away. The primary concern of the investor, therefore, is to have enough assets in the portfolio to diversify away much of this specific risk.

By contrast, systematic risk relates to the tendency for individual properties to move together and to be exposed to driver of this correlation. Such risk is inescapable, being part and parcel of investing in a risky asset class. In contrast to specific risk, the primary concern of the investor is to get a premium return in order to compensate for this inescapable risk.

The ‘market’ is often perceived to be the main systematic risk. The IPF’s research confirms that the market – represented by IPD segments such as City of London offices, shopping centres, South East industrials, and so on – is the predominant risk in most UK commercial properties.  

While this risk in most properties is proportionate to the market, the research also finds that some properties have either accentuated or dampened sensitivities to changes in the market’s return – they are correspondingly high or low ‘beta’ properties (figure 1). To ensure that their returns are commensurate with this accentuated (or dampened) risk, such properties need to be priced with a higher (or lower) risk premium than the market average.  

By contrast, specific risk for most properties tends to be low. Large numbers of tenants, low yields, and relatively long unexpired lease terms all keep specific risk in individual properties on the low side. The implications for portfolio construction are outlined later.

A minority of properties have high levels of specific risk. The case studies indicate the most common sources are associated with lease events – in particular, vacancies at the end of a lease, tenants falling into bankruptcy and so on. Furthermore, there is a tendency for investors to start pricing prospective lease events (for example, what is expected to happen on and after a lease expiry) very cautiously to the extent that the actual outcomes on average represent pleasant upside surprises. This creates volatility (risk) before and after the lease event, compounding the direct impact of the event.  

The second largest source of high specific risk in UK commercial properties is associated with asset management. This involves a variety of activities, from the defensive refurbishment of a property in order to make it more lettable to more opportunistic activities such as reconfiguring space to make it more efficient, engineering a change of use through the local government planning process, and the re-gearing of leases. 

In the same way that volatile sentiment accentuates the direct impact of lease events, refurbishment activity tends to generate big downside swings in performance when being undertaken but significant upside once successfully completed.  

While the timing of these trains of events varies from property to property (being dependent on their own lease cycles), meaning that they represent a form of specific risk, the impact seems to be accentuated by the market cycle. Those properties most exposed to high refurbishment expenditure and to the letting market tend to have relatively high market sensitivities (betas).

Properties with high capital values relative to their market segment and with relatively small numbers of tenants, high yields and short unexpired lease terms also tend to have above average market sensitivities.

In other asset classes, the realisation that there are systematic drivers other than and independent of the market (or simple beta) has led to the development of ‘factor-based’ investment strategies – for example, a bias towards or focus on ‘value’ rather than ‘growth’ which reflects the belief that stocks with either of these characteristics behave differently over the cycle to the average stock.  

The IPF research, however, suggests that few commercial properties are systematically affected by factors such as yield, size and quality. The reason appears to be that individual properties have more significant idiosyncrasies that overwhelm these relatively small factor influences. Lots of properties are required to get an exposure to such factors, making implementation difficult. 

This also applies to the influence of town, which is rarely a significant influence on a property’s performance. Picking a favourably performing town is no guarantee that the property will perform well; large numbers of assets are required to achieve such performance.

What are the implications of this research for investors? First, risk in individual properties needs to be priced more systematically and with a greater range than in the past. Those types of property with high beta (for example, properties frequently exposed to the letting market and with high refurbishment needs) on average have historically delivered relatively poor risk-adjusted returns (negative alpha), while large properties and those infrequently exposed to the letting market have tended to experience significant alpha.

There are also lessons for portfolio structuring and risk control. Over the 10 years to 2013, diversification of portfolios made up of ‘plain vanilla’ properties – those with relatively low exposure to capital spending and the letting market and with slightly better than average characteristics in terms of lot size and number of tenants – could be achieved with very few properties – as figure 2 shows (portfolio risk is also low).

By contrast, properties frequently exposed to the letting market and to relatively high capital expenditure – labelled in the report as ‘asset management intensive’ – have relatively high levels of portfolio risk regardless of the number of properties. Those investing in such properties need to diversify to keep portfolio risk down but, because of the high betas in the underlying properties, can never avoid relatively high levels of risk.

The third lesson is that the factor-based investment strategies that have become popular in the equity market are unlikely to take off in real estate. The exception might be those based on unique property characteristics, such as asset management intensity and its converse.

Finally, investors should be concerned more with structuring property portfolios on the basis of exposure to lease events and the need for intensive active management than on the traditional basis of geography, because these issues make a bigger difference to a property’s performance and risk profile.

Paul Mitchell is a consultant at Paul Mitchell Real Estate Consultancy