As inflation creeps back into the system the need for an effective hedging strategy becomes clear. Property in general and real estate securities in particular can play a valuable role, as Robin Goodchild reports
The credit crunch is causing considerable uncertainty across all areas of investment. At times like these, investors hope they have assets in their portfolio that have not fallen in price and can actually be sold at their current value. Few assets have met that test during the last 12 months. Government bonds have probably performed best but, even there, the increased volatility has made trading more difficult than usual, and asset allocators know that a government bond portfolio will not produce sufficient returns to meet liabilities or minimum profit levels over the medium term.
Investors reviewing their investment strategy at this time should not place too much emphasis on current events, except perhaps as a reminder that there are no free lunches. New investment ideas that offer returns, which appear too good to be true, invariably have a catch - as investors in CDOs, SIVs and structured products have recently discovered. Property is a tried and tested investment class and offers plenty of benefits in a multi-asset portfolio.
Real estate investing began for many pension funds in the 1970s - the decade when inflation first really created a challenge for institutional investors. During Mervyn King's ‘NICE' (non-inflationary consistent expansion) decade, assets that provide a hedge against inflation have not been a pressing requirement for asset allocators and, in any event, many governments and some corporates now issue inflation-indexed bonds (known as TIPs in the USA and index linked gilts in the UK) so there is a ready made hedge. So isn't viewing property as an inflation hedge within a multi-asset portfolio highly dated thinking?
Inflation-indexed bonds have become expensive so that very few funds can meet their liabilities from investing in such bonds alone - alternatively it can be strongly argued that there are too few inflation linked bonds and, if governments issued more, pricing might become more attractive;
The risk of unexpected inflation is always a major threat for pension funds (and other investors) seeking to meet their liabilities;
Price inflation has recently spiked up as a result of the huge increase in oil and food prices.
My answer is a resounding ‘no', for three reasons:
Even though today's elevated inflation levels may well be temporary, the effect of high inflation on the returns from, for example fixed-income bonds, is catastrophic. Prudent trustees and their advisors must place material weight on any scenario that has a disproportionate downside, even if its probability of occurring is expected to be low.
After the events that precipitated last year's credit crunch and the subsequent ripples, everyone should be aware that investors operate in a world of ‘uncertainty' not ‘risk'. While analysts can calculate the probability of an event based on past performance, no one knows the reliability of that estimate because the future may well be very different from the past - and often is.Why else do we have so many statistical ‘black swans'?
Pension fund sponsors are keen to minimise the likelihood of unexpected cash calls to maintain solvency measures. Thus liability driven investment (LDI) strategies have become an important topic in the pension fund industry and inflation-hedging a key component of those products. Most LDI portfolios comprise a mix of bonds (both fixed and index linked) that will partially ‘match' a fund's liabilities in terms of duration, with the balance in risk-seeking assets that are expected to produce higher returns so that, overall, there is sufficient to pay expected pensions, on a variety of scenarios.
The move to LDI has not had too much effect on real estate allocations because they are generally low (5-10% of the total portfolio in many cases) and this allocation is often retained among the risk-seeking assets. Recent research by EDHEC, the increasingly influential French financial research institution, on the appropriate asset mix for an LDI, suggests this may change. Where the investment time scale exceeds 10 years (as it should for all but the most mature pension funds), EDHEC find that real estate has a closer correlation with the liabilities than any other asset class, including an equal weighting of the investments analysed (see figure 1).
EDHEC suggests that the equally weighted portfolio is a viable solution and describes its approach as ‘an imperfect hedging portfolio'. At the moment, their research is still a ‘work-in-progress' but it could have very significant ramifications both for LDI strategies in general and for property investment in particular. If EDHEC concludes that an asset mix other than equal proportions is optimal, it is highly likely that property will be one of the assets that warrants an over-weight.
As everyone should know, property is a good match for pension fund liabilities because investment returns are derived from a mix of income and capital growth. The income return is normally at a level greater than a government 10-year bond and the appreciation element is driven both by economic growth and inflation. Property is sometimes likened to a convertible bond because of its fixed income and equity components. Moreover there is some similarity with index-linked bonds, albeit with significant tracking error. This can be reduced by focusing on properties with rents linked to CPI, whether through standard lease indexation, as in most continental European markets, or by substituting RPI increases for the usual upward-only rent review clause in a UK lease.
It is not difficult to see, therefore, why property can be a good match for liabilities.
Figure 2 shows the correlation numbers between the three main asset classes (using global data in dollars). The low correlation over the last 20 years shows the merits of a diversified portfolio with the expected hierarchy of returns. Moreover, the greater diversity of national real estate returns reduces the volatility of a global portfolio more than for the other asset classes; as a result, the variance is the same as for global bonds (albeit without adjusting for appraisal smoothing).
More interestingly though is how the correlation of bond and real estate returns has changed over time (see figure 3). In the early 1990s, when inflationary expectations were still elevated, the correlation is negligible or negative but during the current decade it has been close to 1. In other words there has been no diversification benefit.
These data clearly demonstrate the role that real estate can play in an LDI strategy, both as a bond substitute and as an inflation hedge, the convertible referred to above. Moreover, real estate acts as a bond substitute but at a higher level of expected return so making the whole strategy more robust. It will be interesting to see whether EDHEC's research endorses this opinion.
Global real estate investing is still comparatively new and few institutions run a truly worldwide portfolio. The IPD Global Index, issued officially in June this year for the first time, points the way. More investors will have global property portfolios in the future but what will they look like?
There is a handful of very large sovereign wealth and pension funds that have truly global direct real estate portfolios. However, such portfolios are practicable only for the super funds with total assets of over €100bn. Some investors are constructing global portfolios from private indirect assets and this product is a major growth area for the leading real estate fund managers. But it still requires significant scale to achieve sound diversification though fund of fund products will reduce the minimum investment in time.
Real estate securities are a much more accessible way for most investors to access the global market place. With the growth of tax transparent REIT structures around the world, it is now possible to invest in the five largest national markets (USA, Japan, UK, Germany and France) without suffering double taxation, as well as other important markets such as Australia and Hong Kong.
In a portfolio context, global REITs have important advantages because there is a well recognised index that is investible, for example EPRA/NAREIT. Thus, institutions can invest passively in global REITs - an option that is not available in the private market (apart from through the embryonic derivatives market). This is an important attribute as more pension funds hold ‘core' and ‘satellite' portfolios to differentiate where they are seeking ‘alpha' and ‘beta'. In fact, Steven Graham of Watson Wyatt advocated the use of global REITs in a ‘core' portfolio in this column a few months ago (see IPE RE Jan/Feb), in conjunction with a domestic real estate portfolio.
LaSalle's own analysis shows that global REITs consistently enhance the risk/return characteristics of a multi-asset portfolio over the inclusion of domestic REITs. In fact the global REIT market appears to be one of the few public markets where it is still possible to get diversification benefits across geographies (see figure 4). Most other public markets, for example the S&P 500, FTSE 100 and Nikki, all show high levels of cross-correlation. Global REITs lend themselves to actively-managed strategies, as well as passive, because the underlying national markets are driven by local factors.
The REIT market is still developing globally. Germany, for example, needs to make its structure less rigid so that more G REITs are created. The UK may also need to adjust its REIT rules although they could be workable if the operating environment were more favourable. But the advantages of daily trading and pricing over the private markets are massive in a world where asset allocators want to switch exposures quickly, while defined contribution plans need instant access and daily pricing too. Given that DC plans are likely to grow much faster than DB schemes, the future is obvious.
Pension funds, particularly in the UK, have long regarded publicly-listed property stocks as a different asset type from private property, either as a part of the wider stock market or distinct in its own right. This is based on correlations and variances that take no account of the artificial smoothing of private property returns as a result of the appraisal process. When de-smoothed returns are used, the correlations are higher and the difference in variance less marked. In the medium term, say five years, there is little difference between the returns from public and private, so the significant difference in liquidity strongly tilts the scales in favour of the public markets. Property professionals need to be alive to the opportunities even if this is not a fashionable conclusion.
Dr Robin Goodchild is Head of European Strategy at LaSalle Investment Management