Diversification  Global investment is back in favour, but are investors achieving effective diversification? Peter Hobbs looks at the complex behaviour of global cities

As real estate markets have recovered from the global financial crisis and confidence has returned to the investment market, there has been increasing interest in building ‘global’ real estate exposure. The crisis was associated with a return to domestic investing, but the uneven recovery since the crisis, coupled with a new wave of investors, has brought rising appetite for global investment.  

Running alongside this globalisation of real estate is increased awareness of differential performance across and within countries. Over the past year, for instance, there was a massive 2,100bps (21%) separating the best performing city, Calgary (19%), from the worst performing – Barcelona (-2%).

Over the longer term, there are also significant variations in the behaviour of the more volatile financial cities of Hong Kong, London or New York, the volatile tech and commodity cities of Calgary, Perth and San Francisco, and the more stable cities of Munich, Paris, Seoul and Tokyo. These variations are important when constructing and managing individual property portfolios, given the potential diversification benefits when investing across and within countries.

It is within this context that IPD’s recently released Global Cities report provides valuable insights into variations across global markets. In the US, for example, the city with the best performance (San Diego) was separated from the worst (Washington DC) by 1159bps, an even wider range than the 928bps separating the best and worst US cities a year ago.
Smaller countries also showed significant variations across their domestic markets, as illustrated by the UK and Canada where around 800bps separated the best and worst cities: London (8.2%) and Birmingham (0.2%); Calgary (19%) and Vancouver (11.2%).
There were also marked variations of nearly 500bps between cities in Germany and Australia, and over 300bps for Belgium, France, Portugal, and New Zealand.

The analysis also confirms the general slowdown in global performance during 2012, although the momentum varied significantly across cities. Cities decelerating the fastest tended to be those that peaked early in the recovery, especially cities in the south of England (London, Birmingham), those along the I-95 corridor of the US Mid-Atlantic coast (Washington DC, Philadelphia, New York) as well as the Asian cities of Hong Kong and Singapore. North America’s Pacific Northwest (Seattle, Portland, Vancouver) also lost a moderate degree of momentum over the course of the year, as did Central and Eastern European cities (Warsaw, Prague, Budapest).

A handful of cities broke with the global trend and provided moderately higher returns in 2012 than a year ago. Houston, Dublin, Lisbon, Cape Town, Johannesburg, Yokohama, Toulouse, and Wellington all fit this pattern. Of the three European cities with returns improving by at least 100bps in 2012, two of them – Dublin and Lisbon – had experienced deep troughs in this cycle. The other, Toulouse, is a city whose economy is closely tied to the fortunes of a single firm, EADS, the parent company of Airbus.

Although there has been a slowing of performance across most markets, there are also clear patterns in behaviour between similar types of city. These patterns are often captured by comparing the volatility of performance but can also be shown by exploring trends in value growth since the trough of 2009. It is clear that the recovery has benefited three groups of cities: those driven by commodities; financial and capital cities; those dominated by tech industries. It has tended to be these cities with globally relevant functions that have recovered most strongly since the crisis, certainly compared with the secondary cities across most countries.

Just as city-level returns within individual countries mask significant differences in performance, so property sectors mask performance within cities too. At least six cities – Calgary, Perth, Manchester, Brussels, Rotterdam, and Barcelona – had more than 1,000bps separating their respective top and bottom performing property sectors in 2012.
Now compare this with the US, where the best performing sector (office) of the best performing city (San Diego) outpaced the worst performing sector (also office) of the worst performing city (Washington, DC) by just over 1400bps. Barcelona alone had more variation than that among its own property sectors in 2012.

Close examination of city and sector data can also inform our broader perspectives of national and global performance. Although the office sector was a drag on performance in many cities in 2012, it gave a boost to the top three city markets – Calgary, San Diego, and Houston – clearly differentiating all-property performance in these cities from their peers. And in Europe, the retail and apartment sectors were often outperformers among otherwise sluggish all-property performance at the city level. Also, the apartment sector, which had outperformed in US cities in recent years, looked less enticing in 2012 than it did just a year ago.

These marked variations are also apparent at the sub-market and asset-specific levels. This is clearly shown for the European market as a whole and for London in particular.
Across Europe, the greatest sub-national variations are in the UK and Germany, but they also exist in Belgium, France, and even in Italy where national office returns were actually better than in the two major cities in 2012.

Within London there are also significant variations, with the West End office market turning in the strongest returns of over 10%, stronger than the City (5-10%) and far superior to the negative returns for outer London. Ten-year annualised results show a similar hierarchy, proving that 2012 was not necessarily an atypical year. The analysis can be taken a step further to explore asset-level performance series for London. The graph shows that most assets perform close to the median but the tails to the distribution are particularly significant, with close to 20% of assets in London experiencing declining returns and a similar proportion achieving over 10%. The rich variation in city-level property performance underscores the importance of a well-structured portfolio in managing risks and exploiting opportunities. Geography still matters, right down to the asset address.

Cities in the multi-asset class portfolio
At the other extreme, there are important insights from comparing cities with the performance of other asset classes. In 2012, performance of the major asset classes correlated strongly to volatility. Equities, especially property equities, tended to provide high returns but were also associated with higher risk. Bonds and direct property offered lower returns but with a historical record of more stability. The variations in the performance and behaviour of cities also varied significantly across markets, some of which approached both the return and volatility of global equities and others behaving more like bonds. If exposure to real estate determines the asset class’s role in the overall portfolio, then there are some city property markets that can provide great diversification benefits for multi-asset investors. Some city markets, on the other hand, may simply correlate with other asset classes, thus providing an equivalent amount of return per unit of risk.

Beyond exploring the variations in recent and longer-term behaviour of cities, the report also provides valuable insights into outlook for performance, with three specific areas for concern as we move through 2013. The first is cyclical performance, as shown in figure 4 that has office sector returns in 2012 on the vertical scale against the annualised returns for the previous three years. The result is a rough approximation of cyclical movement during the recovery, with the top right quadrant representing cities that have outperformed the global average over the past three years and continued to do so in 2012 (‘peaking’ cities). Moving clockwise, the bottom-right quadrant shows above-average performance on a three-year annualised basis, but sub-par performance in 2012 (‘slipping’ cities). The bottom-left quadrant shows cities that have lagged global performance consistently in both time periods (‘sluggish’ cities). Finally, the top-left quadrant shows cities that lagged the global average on an annualised three-year basis, but were relative outperformers in 2012 (‘improving’ cities).

Among these office markets, the three major regions of EMEA, Asia Pacific, and North America sprawl across most of the cyclical categories. EMEA cities were concentrated most heavily among the sluggish markets, but modest relative improvements could be discerned in office markets of Antwerp and Geneva. The South African cities performed relatively well in 2012, as did the London and Paris office markets, while Oslo slipped just below the global level. The Asia Pacific office markets are split mostly between peaking Australian cities and sluggish Japanese markets. Seoul, Auckland, and Wellington find their office markets somewhere in the middle. North American cities are tightly packed among the peaking cities, despite having lost momentum in 2012. The mid-Atlantic markets (New York, Washington DC, and Philadelphia) are shown here as they transition from peaking to slipping. Atlanta, the region’s late bloomer in this cycle is an isolated example of a marginally improved North American office market. It is in these peaking cities that there might be most concern over the outlook, particularly for those that have performed exceptionally, such as the commodity and tech cities of Calgary, San Diego, Perth and San Francisco.  

The second area of concern relates to the low level of income returns across most markets. There are exceptions to this trend, such as for Dublin where its extraordinary income return of 10.1% in 2011 edged up again in 2012 to 10.3%, setting another record in this city’s recent history. Relatively high income returns of 7-9% are also found in Australia, New Zealand, and South Africa, where long-term national bond yields have historically been higher than many of the other countries and have had structural impacts on the income return over time. Beyond these cities, the income returns for most of the cities in Europe and North America were at, or generally near, historic lows. Zurich and Geneva were hovering between 4% and 4.5% in 2012, the lowest of any of the cities in IPD’s coverage.

The third area of concern relates to the spreads between property and bond yields. On the face of it, these spreads were, at the end of 2012, remarkably attractive, at around 400bps in London, Tokyo, Berlin and Zurich, and over 250bps in Paris and New York. These wide spreads have attracted capital into the real estate market, and remain an attractive feature of the asset class. But this was the situation at historically low levels of bond yields and as interest rates start to climb, the gap will narrow. Since the end of 2012, bond yields have risen by around 90bps in the US and the UK, significantly narrowing spreads with real estate yields, and raising concerns over the implications for real estate pricing.  

This brief summary provides a range of insights into city performance across markets, teasing out some of the longer-term behaviours from the short-term pricing and prospects.  These behaviour differences are critically important in constructing individual property portfolios that are often structured with a mix of stable and volatile performing markets. They also have important implications for multi-asset portfolios, with exposure to different cities influencing the role of real estate in the overall multi-asset portfolio.

It is possible, for instance, to build a real estate portfolio with exposure to cities that have equity-market characteristics that offer little diversification to an equity-dominated portfolio, or to build one with more bond-type characteristics that provide strong diversification benefits. As real estate becomes increasingly important in multi-asset portfolios, it is these types of granular insight that will help navigate through the opportunities and risks as investors build their global real estate exposure.

Peter Hobbs is head of research at IPD