After historically low risk premia in UK real estate, it is now at its highest level since the early 2000s. But is this overly cautious, asks Paul Mitchell
It goes without saying that the poor real estate performances in most European countries over the past one to two years predominantly reflect large increases in yields of around 300bps in the UK and around 200bps on the continent. Of more fundamental interest are the influences behind these increases in yields and the large falls that preceded them in the mid-2000s. Was real estate in a bubble where investors lost all sight of risk, or is the downturn an unlucky consequence of an unpredictable, unprecedented economic recession?
Real estate's investment value can be seen simply as its expected future income discounted by the ‘hurdle rate' required of investors. In turn this hurdle rate is the combination of a base, ‘risk-free' interest rate (such as bonds) and the risk premium which investors require over this risk-free rate. What the typical level of this risk premium should be is subject to endless debate in the investment and fund management industries.
My view for the UK is derived from ‘backing out' what has been priced, ex-ante, over the last 30 years and also from face-to-face interviews with major institutional investors and investment consultants, undertaken for the Investment Property Forum's (IPF) Multi-Asset Allocation in the Modern World research. Both these indicate a long-term risk premium of around 2%. While difficult to quantify, I would suggest a similar forward-looking estimate for continental Europe, albeit with some variation around countries.
When risk appetite is high, this premium will be relatively low (and vice versa). Of course, a low risk premium could also be expected over time to rise back to normal levels and in doing so undermine performance as yields rise; however, the willingness to accept a lower return in the face of such a strong risk appetite and the possibility that market conditions could turn out better than expected might make such low expected performance tolerable.
By drawing on the capital values captured by Investment Property Databank (IPD) and the rental growth forecasts from the IPF UK Consensus Forecast (which I assume to be representative of the overall market), figure 1 illustrates the risk premia over bonds implicit in the market's pricing over the last five years. The calculations are on a similar basis to those I undertook in some new research for the IPF.*
It can be seen that UK real estate's risk premium over bonds fell sharply between the end of 2005 and the middle of 2007; this was the principal source of strong returns over this period. It is not clear if the investment market expected risk to remain priced at this low level.
Having fallen to exceptionally low levels, figure 1 also illustrates the sharp rise in the risk premium over the past two years. It has risen from about 1% over UK bonds at the peak of the market in mid-2007 to almost 3.75% now, the highest level since the early 2000s and well above the historical average. This increase has had a substantial impact on returns over every period from the 3Q 2007 up to the 2Q 2009. In particular, it accounted for most of the poor UK performance in the latter part of 2007 and early 2008.
From the 2Q 2008, however, the rental growth outlook was seen to be deteriorating rapidly and this became the most influential factor behind declines in UK capital values and worsening total returns. Figure 2 illustrates this downgrading in the rental growth outlook.
It shows that in the April 2007 IPF survey, the consensus expected rents to increase at a steady rate. In May 2008, the consensus had downgraded modestly, anticipating stability in 2008 and 2009 and a return to growth thereafter. This opinion had dramatically changed by the middle of 2009, such that rents by the end of 2011 were predicted to be 25% lower than anticipated a year earlier and 28% lower than was being predicted in April 2007. Such expectations would have been reflected in the prices investors would have been willing to pay and hence capital values.
This sharp deterioration in rental growth prospects, of course, coincided with intensifying concerns over the UK's economic outlook from the middle of 2008. It is also interesting to observe that the risk premium priced into UK commercial real estate was back at long-term levels by the middle of 2008; since then, the risk premium has increased in line with growing uncertainty over the UK economy.
Given the similarity in the UK economic circumstances, it is interesting to compare the risk premia investors are now demanding from UK real estate with those during the early 1990s.
Real estate forecasting was in its infancy at this time and there was no IPF Consensus Forecast capturing the rental expectations of the investment community. Instead, I have drawn on the limited information available from that time, including my own rental forecasts made in the early 1990s. These allow an approximation of how UK real estate was priced in the first half of the 1990s.
UK real estate yields were a lot higher in the early 1990s than now and, corresponding to this, prospective returns were higher, even after accounting for the poor rental growth outlook of the time. One factor behind this was that ‘real' interest rates and bond yields were also much higher, and as a consequence real estate yields had to be higher to ensure competitive returns. Even so, Figure 3 indicates that UK real estate's risk premium over bonds between 1991 and 1993 was high, at around 4%. This is comparable to the current level.
With hindsight, UK economic recovery was underway by 1993 and the high levels of risk premia priced in UK real estate proved overly cautious, providing a platform for strong returns in the rest of the 1990s.
Because of continuing uncertainty over the depth of the economic downturn and the pace of subsequent recovery, it is too early to say whether or not the UK economy and real estate market is now at the same juncture as in 1993.
However, a very significant difference this time is that nominal and real bond yields look to be at unsustainably low levels. Any increases in bond yields back to longer-term norms will eat into the pricing of real estate, limiting the scope for the strong recovery in returns that followed the early 1990s downturn. Even so, UK real estate pricing is capturing the risks a lot more than at the end of 2008.
In conclusion, excessive risk appetite appears to have been the initial basis for the downturn in UK real estate prices but since the middle of 2008 an unexpectedly sharp reversal in economic - and hence rental - prospects has been much more influential. Although now high, it is too early to say if the risk being priced in UK real estate is as excessive as it proved to be in 1992 and 1993.
*Paul Mitchell and Dr Shaun Bond, The IPF UK Consensus Forecast and the Returns implied by Property Derivative Pricing: Evolution, Record and Influences, IPF, 2009
Paul Mitchell is founding director of the Paul Mitchell Real Estate Consultancy