Diversification failed to bring the expected benefits during the downturn, but it is too early to reject it. Pacific Star Europe and the Henley Business School have investigated some of the effects on portfolios

Global portfolio optimisation looks attractive for a number of investors from different countries. As opposed to property portfolios concentrated on the home market, global diversification enables a remarkable reduction of risk (eg for UK investors) or increase in return (eg for German and Swiss investors). Correlations between property markets have risen substantially over the last two years, but diversification still works. Considering currency effects without hedging will lead to an overweight of the home currency market, but hedging currency risk may improve the risk-return profile significantly. Euro investors are recommended to start a global expansion with investments in Asia Pacific, as diversification benefits have been consistently higher than those of the Americas.

In the course of the financial and economic crisis since 2008 many real estate investors have reconsidered their strategies. Obviously, the enthusiasm about cross-border investments has lessened. Based on disappointment with the performance in some overseas markets, on a reduced risk appetite and an increased need for control and transparency, some investors are now even focusing purely on their home market. Although such reaction on the crisis is comprehensible, it might be over-hasty to renounce international diversification.

What are the risk-return expectations of major real estate markets over the next five years? Is it true that cross-border diversification benefits have become negligible? Does a global real estate exposure make sense any more? And what is the impact of currencies on the asset allocation? Those were central questions of a recent study performed by Pacific Star Europe in cooperation with the School of Real Estate and Planning at Henley Business School/University of Reading.

In a first step, the project team developed a forecast of the property total return for some 50 countries over the period 2010-14. This was based on an analysis of historic property total return indices, other property market data and macroeconomic figures from different sources. According to the projections, the UK, France and Sweden will belong to the top performers in Western Europe with total returns of at least 8% pa, whereas countries like Austria, Ireland and Spain will mark the lower end with just 3% pa.

On average, returns are expected to be higher in the Americas than in Europe and even more elevated in Asia Pacific. The projected return for the US is 7.1% pa at the lower end for the Americas, while Colombia, Peru and Mexico will more likely produce some 15% pa. Indonesia, Vietnam and India are likely top-performers in Asia Pacific with expected total returns of 18%, 15.6% and 14% respectively. On the other hand, Japan has the most conservative forecast in this region with 6.9% pa.

Special attention has been paid to quantify the investment risk for the above period. For this purpose a new figure called total risk has been introduced. The first component of the total risk is the volatility of the country's yearly property total return, whereas the second component is a newly defined country score. The country score reflects the likelihood that the country-specific environment for property investments - its political, business, financial and legal framework - adversely affects property returns.

Following such comprehensive understanding of risk, a market like the UK - despite of its high maturity and transparency - has to be classified as ‘medium risk' rather than ‘low risk'. The reason for this is the significant volatility of property returns in the UK. On the other hand, an emerging market with a less developed business environment might well achieve a medium total risk status if the property performance is fairly stable, as in Poland, for example. The project team is convinced that this new, broader, balanced concept of total risk is more appropriate as a base for asset allocation than the classical approaches.

Many investors have complained about rising correlations between the returns of different real estate markets which would make geographical diversification less effective. In fact, the team from the University of Reading and Pacific Star has found that the average volatility of total returns, as well as the correlations, have increased significantly. For example, the correlation between the UK and France for a period from 1994 to 2007 compared with the period 1994 to 2009 has increased from 0.1 to 0.5.

The research team, however, expects that the differences in the growth rate of economies will widen and that the correlations between real estate returns will decrease again in the mid-term. Furthermore, as the results of the study show, even in the current scenario for some investors it might be beneficial to improve their risk-return profile by diversifying within their region rather than to keep single home country exposure, and they might profit even more from a global diversification. For other investors it might be more attractive to go global directly.

The optimised global minimum risk-portfolio contains a 40% allocation to Europe and a 30% allocation to Asia Pacific and Americas each. The expected total return of this portfolio is 8% with a total risk of 8.7%. Taking this as a base, the main advantage of global diversification for UK investors will be de-risking. A core investor with an UK-only exposure could - by giving up less than 90 bps of his return - decrease the total risk by 450 bps, if he invests in the optimised global property portfolio instead. From the perspective of a German investor, return enhancement will be the dominant effect. By investing globally instead of purely in Germany, the investor could more than double his returns by increasing total risk by just 100 bps.

While the above optimisation results were calculated in local currencies, for many investors the results from the perspective of their home currency might be more important. Therefore the optimisation was also run taking into account currency effects. The first scenario assumed currency movements that could be observed during the last 10 years without any hedging, whereas the second scenario assumed a theoretical total hedging based on current market conditions.

In the first - unhedged - scenario, for each of the currencies euro, sterling and the US dollar the resulting global minimum-risk portfolio looked different. In the case of the euro and sterling, the Europe allocation reached the allowed maximum of 42%. Instead, the optimal portfolio for dollar-based investors includes 39% Americas, whereas the Europe allocation falls to the minimum of 28%. These results speak in favour of the common strategy to overweight the home currency markets.

Interestingly, in the second - hedged - scenario, the global minimum-risk portfolio looks identical for euro, sterling and dollar-based investors: 42% Europe, 30% Americas and 28% Asia. This can be explained by the fact that there is no additional, diverging currency volatility and that the relative attractiveness of countries, depending on the level of interest rates, and consequently on the level of hedging costs or hedging profits, is in principle the same from the perspective of each home currency.

An important implication of the comparison hedged versus unhedged results is that the risk-return profile might be improved significantly via hedging. If the euro-based investor hedges the currency risk completely, in our example the return will be 90 bps higher and the total risk 270 bps lower than without hedging. Generally speaking, investors who expect a strong home currency over the holding period will consider hedging the exposure at least to the main foreign currencies in their property portfolio. Following such strategy, an overweight to the home currency markets may not be required.

The risk-return matrix - hedged euro shows the total return of relevant markets in relation to the total risk. Total risk consists of return volatility and the country risk score, whereas total return comprises income return and value growth in local currencies adjusted with hedging cost into euros. Therefore the matrix displays the real performance for a euro investor, for example a pension scheme or insurer with liabilities mainly in euros. Consequently the different hedging situation and costs, for example for sterling and for euro investors will lead to unequal efficient portfolios.

Finally, it was analysed which way a euro-based investor with core investment style and present exposure to European markets only should start global diversification: in the classical way by "going West", starting in the Americas and particularly in the US, or "going East" with a focus on Asia Pacific. The results clearly favour the second way with a focus on Asia. The main reason is that correlations between property returns in the Americas and Europe have been consistently higher than between Asia Pacific and Europe for the different periods under review, indicating significantly better diversification benefits between the latter. If the global portfolio optimisation for the euro-based investor is run without any limits for the three regions, the Americas will not be considered at all. Instead, the minimum risk portfolio consists of two-thirds Europe and one-third Asia.

Obviously, the Americas share in the previous hedged calculation has been replaced by Europe here, while Asia Pacific remains stable. Given the described correlations between the regions, this is comprehensible. The Americas share in the previous calculation is only a result of the restriction that the optimised portfolio may not differ too strongly from the neutral allocation, ie the global market share. However, the research team of Pacific Star Europe and University of Reading would recommend including the Americas in the long term, with the strategic target allocation for euro-based investors being around 40% for Europe and 30% each for the Americas and Asia Pacific.