During 2000 and 2001, property has provided a timely reminder for institutional investors of its value in a mixed asset portfolio. Not only has property out-performed bonds and equities during this period, that out-performance is now sustained over both three and five years (see table 1).
But property’s place in a multi-asset portfolio is not solely based on return levels. In fact, property is not expected to out-perform equities, only bonds, because of its lower volatility. Its most important benefit in the multi-asset portfolio is diversification. Property returns have a very low correlation with all the other main asset classes as has been demonstrated over the last few years (see table 2). Institutions with high relative property weights have seen significant benefits at the portfolio level.
These features of property are well known but some investors find it hard to access the benefits because of their size. At LaSalle, we regard £50m (e80m) as the minimum portfolio size for a UK institutional segregated account of direct assets. In continental Europe, the threshold is higher at e100m because costs of structuring and managing cross-border holdings are higher. Why are these minima so high? Because any portfolio should comprise at least 10 individual properties to diversify away specific asset risk and have a prospect of consistently matching the appropriate IPD benchmark.
Investors not able to commit £50m for a direct portfolio of hard assets have a number of options. They can either invest in the growing number of private vehicles or go to the public markets. The latter option is usually dismissed as providing returns much closer to common stocks than the property market, as well as incurring tax penalties. However, this conclusion is so important it is worth re-examining. The analysis is carried out solely with UK data because it has both the largest public property company market capitalisation in Europe and the best private market data so all the relevant comparisons can be made. In some ways the UK is not the best place to test this issue as British property companies pay taxes which cannot be recovered by gross investors, unlike notably the Netherlands, where the Dutch Beleggings Instelling structure enables investors to receive net rental income tax-free.1
Returns from property companies must meet three tests to be worth holding as direct property substitutes. These are:
q a low correlation with the rest of the stock market;
q a high correlation with direct market returns;
q negligible effect on the main stock market indices from excluding the real estate sector.
This third test is necessary because, if the other two tests are met, property stocks should be excluded from a general equities portfolio and it is important this does not effect the expected pattern and level of returns from the asset class.

Correlation with rest of stock market
Property’s low correlation with stocks is one of its key characteristics. Our analysis indicates that the correlation of property stocks with the rest of the market has fallen steadily over the last five years, and averages 0.44 over this period (see figure 1). While this is higher than for direct property, it is sufficiently low to constitute diversification. It is, for example, markedly lower than the correlation between equities and bonds (see table 2). In any event, the low correlation between direct property and stocks is partly due to the different ways in which their respective market places operate. (Direct property returns are based on appraisals not daily traded prices.) The measurement of property stock returns is entirely comparable with the rest of the equity market, so inevitably the average correlation is higher. The recent increase in correlation is due to the property sector being affected by the events surrounding 11 September, in line with the rest of the market.

Relationship with direct property
The public markets must produce returns that resemble those from direct property to be an appropriate substitute. It is apparent from figure 2 that the same cyclical patterns are observed in both markets, but with a lag. This lag is explicable when the nature of the two series is understood. Public markets prices move on sentiment. In contrast, the private market capital values are based on valuers’ appraisals of portfolios of properties, not on transactions. Appraisals tend to be ‘backward looking’ in that professional valuers like to base their opinions on evidence of completed transactions.
The transmission mechanism works as follows: changes in sentiment lead to different prices being paid for properties; valuers change their appraisals once these new transactions have occurred. It thus takes about six months for market sentiment to be reflected in appraisals. If the data is analysed with the public markets lagged six months, then the correlation averages 0.73 – demonstrating that investors are getting the same type of returns.
This conclusion is emphasised in table 3 where investors obtain similar returns from the public and private markets, except in the short run when the lag effect can dominate. It is not surprising that returns from public markets are lower slightly than in the private, because dividend yields are lower than property yields in the UK. However, the public market investor has greater liquidity even if market capitalisation is not sufficient to permit instant trading.

Effect on rest of stock market mandates
If it is sensible to invest in property company stocks to obtain a real estate exposure, then real estate stocks should be excluded from other stock mandates. If not, there is a danger that the equity and real estate managers for a fund will both have a position in the same company. The best solution therefore would be to exclude real estate stocks from equity mandates where the property manager is authorised to use the public markets.
But does this change have any impact on the equity market exposure in terms of the level and pattern of returns? Property’s weight in the FT All-Share is only 1.4%. It is even lower in the FTSE 100. The only major index where it could be regarded as significant is the FT Small Cap. where the weighting is 6.5%. In all cases the correlation of the index with its ‘ex. property’ equivalent is over 0.99. Furthermore, the average returns are virtually unchanged too. There is therefore no adverse effect on the residual mandate and there may be benefits. Anecdotally, equity fund managers often find the real estate stocks difficult to call and so may actually welcome their exclusion.
Real estate stocks are a sensible means for obtaining property exposure where portfolio size makes a direct portfolio sub-optimal. They are widely used in the US, the Netherlands and Australia. Increasingly investors are seeking a global real estate exposure and, for all but the very largest investors, this cannot be achieved without using the public markets. There is also a case for including real estate stocks within a property portfolio comprising mainly direct assets, but that may need another article!
Robin Goodchild is European director and Chris Saunders is investment analyst at LaSalle Investment Management, London
1 It is interesting to compare the dividend yield of Haslemere BV, a Dutch BI which invests solely in UK property and has a dual quote in both Amsterdam and London, with the FT Real Estate sector average. Haslemere pays about 9% while the average is around 2.75%