Finding real estate's relative value
With yields at historic lows, real estate is priced attractively versus other asset classes. But, ask Hans Vrensen and Dennis Schoenmaker, how can investors best assess its value?
Unprecedented quantitative easing (QE) policies and unconventionally low (nominal) policy rates created an influx of a ‘wall of money’ into financial markets. Owing to this, commercial real estate has matured into a widely accepted financial asset type in investment portfolios.
As a consequence (figure 1), prime European commercial real estate yields have tightened to historic lows. Average European yields across our five selected core markets (France, Germany, the UK, Belgium and the Netherlands) compressed from about 6.5% in 2009 to a record low of just above 4%. Also, we note that UK prime yields have stabilised since the Brexit vote, while the rest of Europe has continued to tighten.
A natural question arises: is there still good value in European direct real estate? The simple answer is that real estate is expensive in both absolute and historical terms. However, in an increasingly interconnected world in which multi-asset investors are becoming more dominant, all valuations are relative. Therefore, multi-asset investors compare direct real estate with other asset classes to determine its relative value. So, the more appropriate approach is to compare direct real estate with other asset classes, including bonds, equities and REITs.
This requires us to answer the next question: how does one determine relative value? To do this, we construct a direct commercial real estate relative value index (RVI). This RVI compares direct real estate with the aforementioned asset classes, following three steps: it compares prime property yields with 10-year government bond yields; it compares net operating income (NOI) multiples with stock market earnings multiples; and it determines premium or discount of REIT prices to their underlying net asset value.
To quantify each step, we implement the Z-score methodology – that is, we normalise and standardise the data and implicitly assume a reversion to the historical mean. Doing this we can easily combine the three comparisons into the RVI.
We stick with our selected core markets of France, Germany, the UK, Belgium and the Netherlands to construct the average European RVI. Ideally, one would use corporate bond yields. But we use government bonds owing to the lack of sufficient issuance in some key European markets. Finally, we only have consistent historical data for all key global markets back to 2006.
Government bond yields have – mainly due to QE – fallen since 2009 (figure 2). In Germany, bond yields even traded in negative territory briefly. In the past year, bond yields have started to return to normal as some central banks, such as the Bank of England, have increased their base policy rates, and while the ECB started to reduce QE. In this respect, it is noteworthy that UK government-bond yields are about 100bps higher than in Germany.
In essence, since real estate yields have tightened less than government bonds, the spread between real estate and bond spreads is about 300bps, well above the historical average. However, the current spread is below the historical peak of over 400bps in 2015 and 2016. All in all, the spread nowadays remains attractive and is the strongest signal of relative good value for prime direct real estate.
Price-earnings (P/E) multiples are one of the most popular methods equity investors use to determine valuations. P/E multiples are quoted as the price per share, divided by its previous 12-months’ earnings. If a P/E ratio is 20, then investors are willing to pay 20 times the earnings per share.
To construct the European P/E aggregate, we use P/E ratios from the FTSE250, DAX30, CAC40, BEL20 and AEX25 indices. For comparison with direct real estate we use the NOI multiple as estimated by the inverse of the prime property yield.
As figure 3 shows, the direct real estate NOI multiple is high, just below 25, whereas the stock markets’ P/E ratio is just below 20. This indicates that investors are willing to pay a higher price per euro income for property than for stocks overall.
The ratio between the NOI to P/E-multiple ratio has fluctuated between 0.75 and 1.75 since 2006. A reading above 1.0 indicates that investors are paying more for a unit of income in real estate. In the aftermath of the financial crisis, the ratio moved in a much tighter range, between 1.05 and 1.35, meaning real estate is relatively less attractive than equities. Some regional differences are noteworthy – the ratio between the NOI to P/E multiple is highest in Germany, whereas in the UK it is lowest. This means that UK direct real estate is less attractive relative to overall stocks when compared with German direct real estate.
The third component in our RVI is to determine where direct property is priced relative to REITs (figure 4). On average REITs listed in our five core countries traded at a 1% premium to their EPRA NAV at the end of 2017.
UK REITs traded at a discount to reported EPRA NAVs as of year-end 2017, while German REITs traded at a premium. Of course, REIT pricing can change significantly in the short term.
Also, constituent companies for specific countries – especially Germany – and balance sheet leverage have also changed significantly over the past 12 years. For most direct real-estate investors, the longer the timeframe the stronger the analysis. Despite having a longer series of data for the US and European markets, we arrived at the 2006-17 period to allow for the consistent comparison across all three regions.
The interesting angle here is to look at the historical perspective of this NAV premium/discount. This is also where the Z-score methodology can be explained in greater detail. As of the end of 2017, UK REITs traded at a discount of 12% to their EPRA NAV, implying that direct real estate was relatively more expensive at that point.
However, this 12% discount of UK REITs was just below its 12-year average of 13%. Therefore, based on the 12-year period under review, direct UK real estate was still attractively priced as of year-end 2017.
In contrast, German REITs traded at a 12-year average discount of 11%, while at year-end 2017 they traded at a 9% premium to EPRA NAV, making direct real estate much more attractive. Combined, the REITs active in the five European countries in our analysis were priced at a 0% discount to EPRA NAV at the end of 2017. When compared with their average 9% discount over the past 12 years, direct real estate remains more attractive.
Despite the mismatch of property type exposure between different country’s REIT constituents, the within-country NAV discount comparison does measure relative pricing as both REIT pricing and the NAVs are based on each REIT’s underlying properties. Given these data, coverage and analytical limitations to our comparison, European direct real estate looks attractive from the end of 2017 relative to REIT pricing over the 12-year period reviewed.
If we combine all three comparisons based on the aforementioned Z-scores methodology, based on year-end 2017 pricing of European prime direct real estate offers good relative value when compared to bonds, equities and REITs (figure 5). Theoretically, the further below the zero line the RVI scores, the more attractive a market is.
The historical trends show that differences between the main markets are small but can show divergence. The UK is a case in point, as it is on a different trend than Germany and France.
As a consequence of property yield stabilisation over the past two years in the UK, the market is now marginally more attractive than France and Germany. However, the European aggregate is currently below the main markets, as the Netherlands and Belgium have RVI scores below (better than) the European average.
In addition, the current relative value view takes into account year-end 2017 asset pricings. This is likely to change as bond yields normalise (or diverge, as has been seen recently with Italian bonds). Exactly, where commercial real estate relative value ends up will depend on not just bond yields but on REITs and equity pricing.
In addition to the RVI, we see less risk on new supply, rising interest rates and leverage compared with previous European real estate cycles.
The search for relative value
Before we place European direct real estate in a global perspective, we take a closer look at the Asia-Pacific and US situations. We apply the exact same methodology, even though we have more, and better, data for the US.
In the case of Asia-Pacific, we note that markets have become more inter-related over time (figure 6). There is a similar trend in Europe. Despite looking like relatively good value from 2012, Australia has re-priced to such an extent it is in line with the markets in figure 6.
Looking at the US – both the largest single-country market globally and also the birthplace of the AEW direct relative value index – we break out the results by property types. It is clear that residential/multifamily and industrial/logistics offer attractive relative value across US markets (figure 7). However, retail looks relatively unattractive based on year-end 2017 pricing. In our final, and global, perspective, we compare the relative value indices from all three regions. Owing to our methodology, there are some limitations to this comparison.
Each region’s relative-value index assumes that investors can invest in other assets available in that region. It does not assume investors are freely able to invest across all global asset types. However, given the lingering home bias among many small and mid-size investors, this is perhaps not such a restriction.
Nevertheless, we recommend investors interpret the precise RVI score changes carefully, as only at the extreme ends of the standard deviations are they likely to provide strong-enough signals to adjust strategic allocations.
Based on figure 8, it is clear that all three regional indices offer relative value. While our RVI does not directly address relative value across regions, our results do show that European real estate is attractive on a relative basis for multi-asset investors.
It is perhaps useful to point out that for European-based investors attracted to other regions’ better growth rates, much of these higher growth rates should already be reflected in prices.
Finally, our RVI comparison does not address any potential diversification benefits of cross-regional investing, or the associated costs of currency hedging for that matter.
Hans Vrensen is managing director and head of European research and strategy. Dennis Schoenmaker is an associate in the research and strategy team
*In constructing this European aggregate we weight each of these three measures using each country’s GDP in Purchasing Power Parity (PPP).
Acknowledgements: The authors would like to thank Mike Acton and Russ Devlin of AEW’s Boston office and Glyn Nelson of AEW’s Singapore office