ASIAN COMMERCIAL REAL ESTATE
AS AN INFLATION HEDGE:
A MULTI COUNTRY MULTI ASSET CLASS
ANALYSIS
Patrick Lecomte
March 2012
ACKNOWLEDGMENT AND DISCLAIMER
This report is based on research conducted by Patrick Lecomte (ESSEC Business School, Singapore) and Yun W. Park (Chung-Ang University, Korea) and presented at APREA Property Leaders Forum in Beijing (April 2011). The research team would like to thank Rico Kanthatham (Cornerstone) for his advice as well as Andrew Ness and Alan Dalgleish (CBRE) for communicating the data used in this project. This project is part of APREA’s research program conducted under the stewardship of APREA’s Research Committee and in cooperation with academics and professional members of APREA.
This document is for information purposes only. The information herein is believed to be correct, but cannot be guaranteed. The opinions expressed in this document constitute the authors' judgment only and are subject to change.
Reliance should not be placed on the information and opinions set out herein for the purposes of any particular transaction or advice. APREA does not accept any responsibility or liability arising from any use of this document. Content of this report should not be reproduced, in whole or in part, without the prior written permission of APREA.
APREA is a non-profit industry association that represents and promotes real estate as an asset class in Asia Pacific through the provision of better information to investors, improving the general operating environment, encouraging best practices and unifying and strengthening the industry. It covers the four quadrants.
Its membership comprises developers and real estate operating companies, listed real estate trusts, unlisted property funds, investment managers, financial institutions, property securities fund managers, pension funds, sovereign wealth funds and other institutional investors, real estate consultants, corporate advisors, stockbrokers, investment advisors and universities.
The APREA Institute, APREA's education and training arm, provides the foundation for raising standards in the industry through the provision of practical and applied training programs. Its Certificate of Real Estate Investment Finance program is the only course of its kind in the region to be both developed and delivered by industry practitioners.
For more information on APREA, please visit www.aprea.asia
CONTENTS
1.0 Executive Summary 2.0 Introduction
3.0 Scope and Research Questions 4.0 Context: Inflation in Asia
5.0 Methodology and Main Findings
6.0 Commercial Real Estate Markets and Monetary Policies in Asia
References
Appendix
1.0 EXECUTIVE SUMMARY
Inflation is a common risk across Asia. However, the intensity of inflation risk varies extensively from country to country. Noticeably, unexpected inflation is the largest in China, especially in the tier 1 cities – Beijing and Shanghai, where the unexpected inflation rates have historically fluctuated widely on a quarterly basis.
While Asian property delivers positive real returns on average, the short run hedging effectiveness of real estate investments both in securitized and unsecuritized Asian real estate markets is neither strong nor consistent in the eight countries/cities under study (i.e. China: Beijing, Shanghai, Guangzhou, Hong Kong; Japan, South Korea, Singapore, and Thailand).
This might be due to the fact that real estate assets are only remotely linked to inflation in the short run (e.g. no CPI based rental indexation in the countries under study, except South Korea) and the role played by market authorities in controlling real estate markets in Asia.
However, our findings show that inflation rates and real estate returns are co-integrated, or correlated in the long run, indicating that Asian real estate assets do hedge inflation over long periods of time (i.e. over 10 years).
Rental indices (e.g. Beijing and Shanghai retail markets) do exhibit co-integration with inflation rates in the long run, albeit with a lag, as opposed to Capital Value indices which tend to be more responsive.
Correlations between real estate returns and real GDP growth are relatively strong whereas correlations between real estate returns and real domestic stock market returns are consistently weak (with the exception of Hong Kong Office market and the Japanese securitized market), suggesting that Asian commercial real estate markets (both securitized and unsecuritized) are distinct asset classes from the public equity capital.
Further research on the office market indicates that in the absence of official inflation targeting in most Asian countries (with the exception of South Korea and Thailand), real estate markets are influenced by expansionary monetary policies epitomized by strong monetary growth and large Taylor ratios. This is particularly the case in Beijing’s and Shanghai’s office markets where real estate tends to act as a vector between money supply and inflation.
2.0 INTRODUCTION
The search for an inflation hedge has long been on the radar of long term investors. In the absence of mature inflation-linked securities (e.g. inflation-linked bonds, inflation-indexed derivatives), analysts customarily claim that Asian properties provide efficient hedging against inflation. Whilst the topic has already been extensively covered in academic literature, this claim is somewhat arbitrary as past research shows that findings are methodology and period dependent.
Nonetheless, the issue of real estate’s inflation hedging benefits has taken a new relevance in the aftermath of the subprime crisis. Massive government stimulus packages have helped financial markets to recover and led investors to worry about inflation.
The situation facing investors nowadays is nonetheless very different from that of the past (1970s/ 1980s) when inflation was a major concern of investors in long term assets. Most Asia centric inflation studies were carried out at a time when the investment universe was much smaller than the wide scope of investment options currently available to real estate investors. In recent years, Asian countries have developed sophisticated property vehicles (e.g. REITs) while witnessing an exponential growth in their direct markets owing to unprecedented economic development. This has considerably increased the pool of potential hedging tools available to investors.
Second, over the last 20 years, the economy has become global in a so-called ‘flat world’. By being at the epicenter of intense transnational flows, Asia is positioned as the manufacturing hub of this new global economy. The issue of Asian properties’ inflation hedging ability is therefore enmeshed into the broader and complex issue of imported inflation between Asia and the West, in particular America.
Finally, in most Asian countries, the real estate sector plays an overwhelming role in the economy. As a result, commercial real estate markets are often stigmatized by market authorities for playing an active role in complex inflationary processes.
3.0 SCOPE AND RESEARCH QUESTIONS
This research project analyses Asian real estate’s inflation hedging abilities. Due to their diversity and idiosyncrasies, Asian real estate markets present an array of challenges in terms of data collection that the study addresses.
The research covers both securitized and non-securitized property markets in six Asian countries: China, Hong Kong, Japan, Singapore, South Korea and Thailand.
In the case of some unsecuritized real estate markets, the analysis is conducted at the city level rather than the national level. Hence, in China, we look at various
property types for three tier 1 cities: Beijing, Shanghai and Guangzhou, while in South Korea and Japan, we focus on Seoul and Tokyo respectively. Whenever possible, the research provides detail by major commercial property types. The scope of this project and the choices made in terms of level of analysis were constrained by data availability in Asian real estate markets. In the absence of established indices on private commercial real estate assets in Asia, it is still a challenge to conduct in-depth research on Asian direct commercial real estate markets.
The coverage includes 8 cities for 3 property types (office, industrial, retail):
China (Beijing, Shanghai, Guangzhou), Hong Kong, Japan (Tokyo), South Korea (Seoul), Singapore, Thailand (Bangkok). Overall, 22 submarkets are analysed as summarized in table 1 (4 securitized markets) and table 2 (18 non securitized markets).
Data sources used in this research are:
for securitized real estate: FTSE EPRA Total Return, FTSE Xinhua 600 Real Estate Total Return (China)
for unsecuritized real estate: CBRE Capital Value and Rental indices.
With respect to macroeconomic data, consumer price indices (CPI) and gross domestic products (GDP) figures are from Datastream. For the three Chinese cities, Hong Kong, Singapore, and Tokyo, the relevant CPI and GDP were collected at the city level.
Index Data source Time Period Frequency
China FTSE XINHUA REAL ESTATE 600 TR Index Datastream Q3 2001 - Q2 2010 Quarterly Hong kong FTSE EPRA NAREIT Hong Kong TR EPRA Q1 1990 - Q2 2010 Quarterly Japan FTSE EPRA/NAREIT Japan TR Index EPRA Q1 1990 - Q2 2010 Quarterly Singapore FTSE EPRA/NAREIT Singapore TR Index EPRA Q1 1990 - Q2 2010 Quarterly
S ECURITIZED INDICES
Table 1: Indices on securitized real estate used in the research project
Index Data source Time Period Frequency OFFICE
Beijing Capital Value CBRE Q2 1998 - Q2 2010 Quarterly
Guangzhou Capital Value CBRE Q2 2000 - Q2 2010 Quarterly
Shanghai Capital Value CBRE Q2 1996 - Q2 2010 Quarterly
Hong Kong Capital Value CBRE Q2 1993 - Q2 2010 Quarterly
Tokyo Capital Value CBRE Q2 1998 - Q2 2010 Quarterly
Seoul Capital Value CBRE Q2 2002 - Q2 2010 Quarterly
Singapore Capital Value CBRE Q2 1993 - Q2 2010 Quarterly
Bangkok Capital Value CBRE Q2 1998 - Q2 2010 Quarterly
RETAIL
Beijing Rental Index CBRE Q2 2003 - Q2 2010 Quarterly
Guangzhou Rental Index CBRE Q2 2003 - Q2 2010 Quarterly
Shanghai Rental Index CBRE Q2 1998 - Q2 2010 Quarterly
Hong Kong Capital Value CBRE Q2 2003- Q2 2010 Quarterly
Singapore Capital Value CBRE Q2 1998 - Q2 2010 Quarterly
INDUSTRIAL
Beijing Land Price - Capital Value CBRE Q1 2003 - Q2 2010 Quarterly Guangzhou Land Price - Capital Value CBRE Q1 2000 - Q2 2010 Quarterly Shanghai Land Price - Capital Value CBRE Q1 2000 - Q2 2010 Quarterly Hong Kong Warehouse- Capital Value CBRE Q2 2000 - Q2 2010 Quarterly Singapore Warehouse- Capital Value CBRE Q1 1997 - Q2 2010 Quarterly
UNS ECURITIZED INDICES
Table 2: Indices on unsecuritized real estate used in the research project
This study aims to answer three main research questions:
• Are Asian securitized and unsecuritized property markets providing efficient hedges against inflation in the short run /long run?
• What is the impact of expected inflation vs. unexpected inflation?
• What is the positioning of commercial property markets in the context of expansionary monetary policies in Asia?
4.0 CONTEXT: INFLATION IN ASIA
A COMMON RISK BUT VARIOUS DEGREES
Inflation is a common risk in the region but at various degrees as illustrated in table 3 below.
1993- 2010
YoY% China (Urban areas) Beijing Shanghai Hong Kong Japan Korea Singapore Thailand
Average -0.17 -0.27 -0.31 2.14 0.05 3.74 1.56 3.43
Maximum 12.37 17.76 13.95 10.38 2.57 9.55 7.55 10.58
Minimum -12.50 -12.01 -11.16 -6.16 -2.53 0.12 -1.47 -4.38
Standard Deviation
4.74 4.92 4.99 4.34 0.95 1.68 1.76 2.59
Table 3: Inflation Risk: Fluctuation Ranges and Standard Deviations (1993-2010)
Over the last 20 years, cities like Beijing and Shanghai have witnessed extreme fluctuations in their inflation rates even though their yearly inflation averages are deceivingly close to zero. Unsurprisingly, these two cities are also more at risk than less developed areas in China. Historically, Hong Kong has also recorded wide fluctuations in its inflation rate with a significantly larger yearly average than mainland China. In addition to Hong Kong, the two countries where inflation is on average high are South Korea and Thailand.
The graphs below illustrate the contrasting situations between China and other Asian countries less prone to wide fluctuations in yearly CPI, such as Singapore.
-15.0%
-10.0%
-5.0%
0.0%
5.0%
10.0%
15.0%
janv.-93 oct.-93 juil.-94 avr.-95 janv.-96 oct.-96 juil.-97 avr.-98 janv.-99 oct.-99 juil.-00 avr.-01 janv.-02 oct.-02 juil.-03 avr.-04 janv.-05 oct.-05 juil.-06 avr.-07 janv.-08 oct.-08 juil.-09 avr.-10
CHINA- URBAN AREAS
YoY % CPI China- Urban Areas Average 1993-2010
Source: Datastream
-15.0%
-10.0%
-5.0%
0.0%
5.0%
10.0%
15.0%
janv.-93 nov.-93 sept.-94 juil.-95 mai-96 mars-97 janv.-98 nov.-98 sept.-99 juil.-00 mai-01 mars-02 janv.-03 nov.-03 sept.-04 juil.-05 mai-06 mars-07 janv.-08 nov.-08 sept.-09 juil.-10
SINGAPORE
YoY % CPI Singapore Average 1993-2010
Figure 1: Chinese Urban Areas vs. Singapore Inflation Rates (1993-2010)
Source: Datastream
DEMAND PULL vs. COST PUSH
Following the 2007 subprime crisis, policy makers in Asia reacted with massive inflows of liquidity in their economies. Meanwhile, interest rates were maintained at artificially low levels (according to the Taylor rule) while due to demand pressures and external shocks, commodity and food prices rose out of control.
Unsurprisingly, the immediate result was increased CPI inflation and spiralling inflation expectations. Commodities and foods were identified as the main culprits for these sudden inflationary pressures.
Notwithstanding this situation, inflation in Asia is not traditionally cost-push. The Asian Development Bank explains that excess aggregate demand and inflationary expectations jointly account for 60% of CPI inflation for the region as a whole. In China and Singapore for instance, food price inflation shocks explain about 5-6%
of CPI inflation at most1. Hence, Asian inflation is primarily demand-pull and monetary variables such as money supply and interest rates play an important role in triggering and controlling inflation.
INFLATION TARGETING
As a consequence of the Asian crisis of the late 1990s, several countries included in this study adopted the inflation targeting model, most notably South Korea in 1998 and Thailand in 2000. This model was at the time strongly encouraged by the International Monetary Fund2.
Inflation targeting is an economic policy whereby each year, central banks publicly set a target for the CPI inflation rate. Most of the time, this target is actually a range. Table 4 below summarizes the inflation targeting regimes in the countries included in this study.
1 In Inflation in Developing Asia: Demand-Pull or Cost-Push? Asian Development Bank, September 2008.
2 In Monetary Policy and Financial Stability: Is Inflation Targeting Passé? Asian Development Bank.
July 2010.
Economy Exchange Rate
Regime Central Bank Inflation
Targeting Target Variable Target (Range) CHINA (PR) Fixed (de facto) People’s Bank of
China No
JAPAN Floating Bank of Japan No
HONG KONG Fixed (currency
board) HK Monetary
Authority No
KOREA Floating Bank of Korea Yes CPI Inflation rate For 2004-2009,
3% ± 0.5%
For 2010-2012, 3% ± 1%
SINGAPORE Floating (basket) Monetary Authority of Singapore
No
THAILAND Floating Bank of Thailand Yes Core CPI inflation For 2000-2009, 0-3.5%
From 2009 onwards, 0.5-3%
Table 4: Inflation Targeting Regimes in Asia
Inflation targeting gives reassurance to investors that the central bank is dedicated to keeping inflation under control. Noticeably, the regime was instated in the two countries, South Korea and Thailand, where inflation has historically been an issue (with an average annual inflation rate over the last twenty years of 3.74%
and 3.43% respectively).
RENTAL INDEXATION
Rent indexation based on CPI is not widespread in Asia. The only country where this is common practice is South Korea. Table 5 below presents the process of rent escalation for office properties in the countries under investigation.
Source: Asian Development Bank, 2010
Country Property Type
Lease Term Rent Adjustment Mechanism
China Office 2-3 years
In Beijing, typically 5-6 years In Shanghai, longer leases possible
Rent review every 3 years at Open Market Rate (OMR) or through arbitrator
Hong Kong Office Standard: 3 years
But most large tenants: 6 years
Rent review every 3 years at Open Market Rent or through independent expert valuer
Japan Office Traditional leases: 2 years perpetually renewable
Fixed term leases: no right to renew
Rent escalation is negotiable. Rents are not indexed.
For traditional leases, rent reviewed every 2 years.
Singapore Office 3 years plus option to renew for 3 years
Rent is normally fixed, not escalated during the term.
Lease renewed at agreed rent or prevailing market rate at the time of renewal.
South Korea
Office 2-3 years. Landlords tend to prefer longer leases.
Rent escalation based on CPI or Market Rent
Thailand Office 3 years with option to renew Rent escalation negotiable. Landlord may agree to capped rate.
Source: CBRE
Table 5: Rent Adjustment Mechanisms in Asia (Office Properties)
5.0 METHODOLOGY AND MAIN FINDINGS
Inflation hedging in the real estate context is actually a misnomer. Contrary to traditional hedging which is akin to a process of negative replication, what investors are looking for is the ability of real estate assets to more than cover the increase in consumer prices.
Methodologies applied in the research are selected to identify the short term correlations and long term co-integrations between real estate returns in Asian property markets and domestic inflation rates proxied by local consumer price indices.
WHAT IS INFLATION HEDGING?
Regular hedging is a process of negative replication. There is a perfect hedge if
Asset Position +1 Hedge -1
Inflation hedging is perfect if:
Inflation +1 Hedge +1
Hence, we are interested in correlations (short run) and co-integrations (long run) between inflation and potential hedging assets.
Methodologies are chosen for their robustness, in particular the Engle Granger co-integration tests which provide more solid results than the usual Ordinary Least Square regressions3.
STEP 1: EXPECTED INFLATION VS. UNEXPECTED INFLATION
Inflation has two components: expected inflation and unexpected inflation.
Expected inflation is the level of inflation anticipated by the market, while unexpected inflation is the difference between actual inflation and expected inflation at the beginning of the period. Unexpected inflation results from new information previously unavailable to the market. It is the true source of inflation risk. Inflation decomposition is based on Autoregressive Integrated Moving Average models (ARIMA).
Figure 2 summarizes the expected mean and standard deviation of unexpected inflation rates in the 8 countries/cities under study. The chart on the top part of the figure indicates the level of inflation risk embedded into each country/city in a classic mean/standard deviation framework. The table on the bottom part lists the fluctuation ranges of quarterly unexpected inflation rates for each city/country from 1993Q1 to 2010Q4.
3 Co-integration tests pioneered by Nobel Laureates Engle and Granger are designed to test whether two or more time series (e.g. real estate returns and inflation rates) share common long term fluctuations
A 3 STEP PROCESS
Step 1: Modeling the inflationary expectation process in each city/country under study.
Step 2: Applying a multi factor regression model to analyze short-term hedging abilities.
Step 3: Running co-integration tests between property returns and inflation rates to assess long-term hedging abilities.
0.00 0.50 1.00 1.50 2.00 2.50
-0.025 -0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010
Standard Deviation %
Mean %
UNEXPECTED INFLATION (QoQ)
Hong Kong
Shanghai Beijing
Japan Guangzhou
Singapore
Korea Thailand
China
1993-2010 China Beijing Shanghai Guangzhou Hong Kong Japan Korea Singapore Thailand
Maximum UIR 3.798 3.971 4.051 4.26 2.022 2.141 3.975 1.437 3.99
Minimum UIR -3.277 -3.101 -4.159 -2.796 -2.315 -1.471 -1.68 -1.445 -4.42
Figure 2: Unexpected Inflation (1993Q1-2010Q4)
Among the cities/countries under study, China and the three Chinese tier 1 cities emerge tops in terms of inflation risk. Although on average, unexpected inflation in Shanghai is nil, the range of fluctuations in this city is the largest in Asia as shown in the table on the bottom part of figure 2 above. Inflation risk is therefore the most significant in China, in particular in Shanghai. Beijing and Guangzhou are markedly different from Shanghai. In these two cities, the average unexpected inflation is negative, indicating that investors tend to overestimate actual inflation rates. Again, the fluctuation range is large, embodying the level of inflation risk involved. At the other end of the spectrum, Singapore appears very stable with one of the lowest unexpected inflation rates in the region and a relatively narrow fluctuation range over the last 20 years.
STEP 2: REGRESSION ANALYSIS
Regression analysis enables us to test the short run hedging effectiveness of real estate against inflation. In order to conduct the analysis, we adopt a multi factor Ordinary Least Square (OLS) regression model where the dependent variable (real estate return) is regressed on expected inflation rate and unexpected inflation rate as well as the real return of the domestic stock market and the real GDP growth rate for the period under study. The model used to conduct this analysis is a simple multi-factor regression line as follows:
although they might not move together in the short run.
t t t
t t
t
c EIR UIR RGRATE RRETURN r
1
2
3
4
where rt is the ex post real rate of return of asset i in period t EIRt is the expected inflation rate in a country at period t UIRt is the unexpected inflation rate in a country at period t RGRATEt is the GDP growth rate in a country during period t
RRETURNt is the rate of return on the domestic stock market during period t.
An asset provides an efficient hedge for anticipated /unexpected inflation when the inflation coefficient β1=1 and the other inflation coefficient β2 =1. When the inflation coefficient is significantly larger than 1, the underlying real estate market provides a super hedge against inflation. Conversely, when the inflation coefficient is significantly lower than 0, the real estate market delivers a perverse hedge.
REGRESSION ANALYSIS
A multifactor model is used to determine the inflation coefficient for expected inflation and unexpected inflation.
The model controls for the impact of domestic stock markets and real GDP growth rates on real estate returns.
An asset is an inflation hedge in the short run when the inflation coefficient is equal or close to 1. The analysis is carried out for both expected and unexpected inflation rates.
Table 6 below summarizes the results for expected inflation.
Market Class Index Estimation Period β1
Singapore Shopping Centres Non-securitized CBRE- Capital Value 1998Q2-2010Q2 6.72
Japan Securitized FTSE EPRA NAREIT Japan Index 1990Q1-2010Q2 4.99
Japan- Tokyo central five wards Office Non-securitized CBRE- Capital Value 1998Q2-2010Q2 3.61
Singapore- Warehouses Non-securitized CBRE- Capital Value 1997Q1-2010Q2 3.32
Hong Kong Central Office Non-securitized CBRE- Capital Value 1993Q2-2010Q2 1.56
China-Beijing Retail Non securitized CBRE- Rental Index 2003Q2-2010Q2 1.5
China- Guangzhou Retail Non securitized CBRE- Rental Index 2003Q2-2010Q2 1.2
China- Beijing Office Non securitized CBRE- Capital Value 1998Q2-2010Q2 0.52
Korea- Seoul Office Non-securitized CBRE- Capital Value 2002Q2-2010Q2 0.36
China- Shanghai Office Non securitized CBRE- Capital Value 1996Q2-2010Q2 0.1
China- Shanghai Retail Non securitized CBRE- Rental Index 1998Q2-2010Q2 0
Hong Kong Warehouses Non-securitized CBRE- Capital Value 2000Q2-2010Q2 -0.21
China- Guangzhou Industrial Non securitized CBRE- Capital Value 2000Q2-2010Q2 -0.45 China- Guangzhou Office Non securitized CBRE- Capital Value 2000Q2-2010Q2 -0.52 China-Beijing Industrial Non securitized CBRE- Capital Value 2003Q1-2010Q2 -1.16 Thailand- Bangkok Office Non-securitized CBRE- Capital Value 1998Q2-2010Q2 -1.83 China- Shanghai industrial Non securitized CBRE- Capital Value 2000Q2-2010Q2 -2.14 Hong Kong Street Shop Retail Non-securitized CBRE- Capital Value 2003Q1-2010Q2 -2.72
China Securitized FTSE Xinhua 600 Real Estate 2001Q3-2010Q2 -5.83
Hong Kong Securitized FTSE EPRA NAREIT HK Index 1990Q1-2010Q2 -6.56
Singapore Office Non-securitized CBRE- Capital Value 1993Q2-2010Q2 -7.8
Singapore Securitized FTSE EPRA NAREIT Singapore Index 1990Q1-2010Q2 -10.7
Table 6: Expected Inflation- Findings (Regression Analysis)
Table 7 below summarizes the results for unexpected inflation.
Market Class Index Estimation Period β2
Japan Securitized FTSE EPRA NAREIT Japan Index 1990Q1-2010Q2 5.87
Hong Kong Securitized FTSE EPRA NAREIT HK Index 1990Q1-2010Q2 3.85
Hong Kong Central Office Non-securitized CBRE- Capital Value 1993Q2-2010Q2 3.83
Korea- Seoul Office Non-securitized CBRE- Capital Value 2002Q2-2010Q2 2.69
Japan- Tokyo central five wards Office Non-securitized CBRE- Capital Value 1998Q2-2010Q2 2.25 Hong Kong Street Shop Retail Non-securitized CBRE- Capital Value 2003Q1-2010Q2 1.88 China- Shanghai industrial Non securitized CBRE- Capital Value 2000Q2-2010Q2 1.3
Hong Kong Warehouses Non-securitized CBRE- Capital Value 2000Q2-2010Q2 0.92
China- Shanghai Retail Non securitized CBRE- Rental Index 1998Q2-2010Q2 0.57
Singapore- Warehouses Non-securitized CBRE- Capital Value 1997Q1-2010Q2 0.57
China-Beijing Retail Non securitized CBRE- Rental Index 2003Q2-2010Q2 0.47
China- Guangzhou Retail Non securitized CBRE- Rental Index 2003Q2-2010Q2 0.4
China- Shanghai Office Non securitized CBRE- Capital Value 1996Q2-2010Q2 0.38 China- Guangzhou Office Non securitized CBRE- Capital Value 2000Q2-2010Q2 0.17 China-Beijing Industrial Non securitized CBRE- Capital Value 2003Q1-2010Q2 0.16 Thailand- Bangkok Office Non-securitized CBRE- Capital Value 1998Q2-2010Q2 0.11
China- Beijing Office Non securitized CBRE- Capital Value 1998Q2-2010Q2 0.1
Singapore Shopping Centres Non-securitized CBRE- Capital Value 1998Q2-2010Q2 0.08
China Securitized FTSE Xinhua 600 Real Estate 2001Q3-2010Q2 0.06
China- Guangzhou Industrial Non securitized CBRE- Capital Value 2000Q2-2010Q2 -0.221
Singapore Office Non-securitized CBRE- Capital Value 1993Q2-2010Q2 -0.54
Singapore Securitized FTSE EPRA NAREIT Singapore Index 1990Q1-2010Q2 -8.19
Table 7: Unexpected Inflation- Findings (Regression Analysis)
Perverse HedgeSuper HedgeSuper Hedge No Hedge
No Hedge
Perverse Hedge
In statistical terms, inflation coefficients β1 and β2 reported in tables 6 and 7 are not significantly different from zero. Hence, we cannot definitely assert that the above-mentioned findings are significant. The evidence of inflation hedging based on OLS multifactor regression models is therefore not conclusive. Step 2 has not enabled us to determine the actual hedging effectiveness of real estate against inflation in the short run. There is a need for a more powerful methodology as exemplified in step 3 thereafter.
STEP 3: CO-INTEGRATION
To assess the ability of real estate markets to provide effective hedging against inflation, we run the Engle-Granger co-integration test. Co-integration tests are designed to test whether two or more time series (e.g. real estate returns and inflation rates) share common long term fluctuations although they might not move together. Co-integration provides assessment of inflation hedging in the long run.
MARKET CLASS INDEX PERIOD YES/NO
Thailand- Bangkok Office Non-securitized CBRE- Capital Value 1998Q2-2010Q2 YES
Hong Kong Securitized FTSE EPRA NAREIT HK Index 1990Q1-2010Q2 YES
Japan Securitized FTSE EPRA NAREIT Japan Index 1990Q1-2010Q2 YES
Singapore Securitized FTSE EPRA NAREIT Singapore Index 1990Q1-2010Q2 YES
China- Shanghai industrial Non securitized CBRE- Capital Value 2000Q2-2010Q2 YES
Korea- Seoul Office Non-securitized CBRE- Capital Value 2002Q2-2010Q2 YES
China- Shanghai Retail Non securitized CBRE- Rental Index 1998Q2-2010Q2 YES
China- Shanghai Office Non securitized CBRE- Capital Value 1996Q2-2010Q2 YES
Hong Kong Central Office Non-securitized CBRE- Capital Value 1993Q2-2010Q2 YES Singapore Shopping Centres Non-securitized CBRE- Capital Value 1998Q2-2010Q2 YES
China- Beijing Office Non securitized CBRE- Capital Value 1998Q2-2010Q2 YES
China- Guangzhou Retail Non securitized CBRE- Rental Index 2003Q2-2010Q2 YES
Singapore- Warehouses Non-securitized CBRE- Capital Value 1997Q1-2010Q2 YES
China-Beijing Retail Non securitized CBRE- Rental Index 2003Q2-2010Q2 YES
Singapore Office Non-securitized CBRE- Capital Value 1993Q2-2010Q2 YES
China- Guangzhou Office Non securitized CBRE- Capital Value 2000Q2-2010Q2 YES
China Securitized FTSE Xinhua 600 Real Estate 2001Q3-2010Q2 YES
China-Beijing Industrial Non securitized CBRE- Capital Value 2003Q1-2010Q2 YES Japan- Tokyo central five wards Office Non-securitized CBRE- Capital Value 1998Q2-2010Q2 YES
Hong Kong Warehouses Non-securitized CBRE- Capital Value 2000Q2-2010Q2 YES
Hong Kong Street Shop Retail Non-securitized CBRE- Capital Value 2003Q1-2010Q2 YES China- Guangzhou Industrial Non securitized CBRE- Capital Value 2000Q2-2010Q2 NO
Table 8: Inflation Rate - Findings (Co-integration)
All submarkets analysed, except for the industrial market in Guangzhou, display co-integration with inflation rates as summarized in table 8 above. Hence, in the long run, Asian real estate markets do provide hedging against inflation.
Inferences are made at the 5% level of significance
MAIN FINDINGS
While Asian property delivers positive real returns on average, the short run hedging effectiveness of real estate investments both in securitized and unsecuritized Asian real estate markets is neither strong nor consistent. This might be due to the fact that real estate assets are only remotely linked to inflation in the short run (e.g. no CPI based rental indexation in the countries under investigation, except South Korea) and the role played by market authorities in controlling real estate markets in Asia.
However, the findings show that inflation rates and real estate returns are co-integrated in the long run, indicating that Asian real estate assets do hedge inflation over long periods of time (i.e. over 10 years).
Correlations between real estate returns and real GDP growth are relatively strong whereas correlations between real estate returns and real domestic stock market returns are consistently weak (with the exception of the Hong Kong Office market and the Japanese securitized market), suggesting that Asian commercial real estate markets (both securitized and unsecuritized) are distinct asset classes from the public equity capital.
6.0 COMMERCIAL REAL ESTATE MARKETS AND MONETARY POLICIES IN ASIA
In the context of expansionary monetary policies in Asia, the impact of money supply on commercial real estate markets is an important question in addition to inflation. As Chicago economist Milton Friedman once wrote: “Inflation is always and everywhere a monetary phenomenon in the sense that it is and can be produced only by a more rapid increase in the quantity of money than in output”4. Therefore, one can legitimately wonder how expansionary monetary policies affect commercial real estate returns. Another important dimension of this question is the positioning of commercial real estate markets in Asian monetary policy frameworks. In other words, are commercial real estate markets inflationary or are they caught in monetary policies that might be too accommodating?
In order to address these issues, we focus on office properties, by analyzing the link between office market returns, inflation rates and money supply (M2) in the three tier 1 Chinese cities, Hong Kong, Seoul, Singapore, and Tokyo. A series of
4 In The Counter-Revolution in Monetary Theory (1970).
3-variable Engle causation tests is conducted. Table 9 summarizes the results.
Each arrow symbolizes a causation link between any two variables.
BEIJING SHANGHAI GUANGZHOU HONG KONG TOKYO SEOUL SINGAPORE
M2 Real Estate
Inflation
Real Estate M2 Inflation
Inflation M2 Real Estate
M2 Inflation Real Estate
Inflation Real Estate
M2
M2 Real Estate
Inflation
Inflation Real Estate
M2
Table 9: Causation Analysis: Are Commercial Real Estate Markets inflationary in Asia? The Case of Office Properties.
In the absence of official inflation targeting in most Asian countries (with the exception of South Korea and Thailand), real estate markets are influenced by expansionary monetary policies epitomized by strong monetary growth and large Taylor ratios. This is particularly the case in the Beijing’s and Shanghai’s office markets where real estate tends to act as a vector between money supply and inflation. From this analysis, we can infer three types of commercial real estate markets in Asia:
Type 1: Money Supply sensitive: Monetary growth and real estate returns are causally linked, positioning commercial real estate as a vector towards inflation.
This is the case of Beijing, Shanghai, and Seoul.
Type 2: Monetary Policy sensitive: Interactions between inflation and money supply ultimately affect real estate returns. This is the case of Guangzhou and Hong Kong.
Type 3: Inflation sensitive: Inflation and real estate returns are causally linked while monetary growth is affected by real estate returns, signaling that commercial real estate markets do play a role in policy makers’ decisions with respect to monetary policy. This is the case of Singapore and Tokyo.
Graphs in the appendix illustrate the link between office market returns and CPI in the 8 cities (1996-2010).
Based on 3 way Engle causality tests for the direct office market
REFERENCES
Asian Development Bank. Inflation in Developing Asia: Demand-Pull or Cost-Push? ERD Working Paper N. 121, September 2008.
Asian Development Bank. Monetary Policy and Financial stability: Is inflation Targeting Passé? ADB Working Paper Series N.206, July 2010.
International Monetary Fund. Inflation Hedging for Long-Term Investors. IMF Working Paper 09/90.
Investment Property Forum. Property and Inflation. Summary Report, November 2010.
APPENDIX
GRAPHS- OFFICE MARKET REAL RETURNS VS. INFLATION IN THE EIGHT ASIAN CITIES UNDER INVESTIGATION (1996- 2010)
Sources: CBRE Capital Value indices, Datastream
