Asia Mega Cities 2030

SECTOR BRIEFING 67 DBS Asian Insights DBS Group Research Aug 2018 number Asia Mega Cities 2030

02 Asia Mega Cities 2030 Carol WU Head of Research DBS HK carol_wu@dbs.com Danielle Wang CFA Senior Research Director Group Research danielle_wang@dbs.com Ken HE CFA Research Director Group Research ken_he@dbs.com Jason LAM Research Analyst Group Research jasonlamch@dbs.com Jeff Yau Senior Research Director Group Research Jeff_yau@dbs.com Ian Chui Senior Research Analyst Group Research ianchui@dbs.com Derek TAN Senior Research Director Group Research Derektan@dbs.com Rachel Lih Rui Tan Assistant Research Director Group Research racheltanlr@dbs.com QUAH He Wei Senior Analyst AllianceDBS Research hewei@alliancedbs.com Maynard Priajaya Arif Indonesia Head of Research Group Research maynardpriajaya@dbs.com Victor Stefano Research Analyst Group Research victor.stefano@id.dbsvickers.com Chanpen Sirithanarattanakul Head of Research DBS Vickers Securities (Thailand) Co., Ltd. Chanpens@th.dbs.com Special thanks to Kruti Shah from AVP- Economist, Emkay Global Financial Services Ltd. Produced by: Asian Insights Office DBS Group Research go.dbs.com/research @dbsinsights asianinsights@dbs.com Goh Chien Yen Martin Tacchi Editor-in-Chief Art Director

03 04 06 Executive Summary Asia s emerging mega cities in 2030 Key Findings Cities to watch 14 17 22 26 Cities with best property price potential Cities with best office investment value Cities with best retail rent growth Cities with best outlook on hotel assets

04 Executive Summary Mega cities, mega trends The bedrock of Asia s emerging mega cities is undergoing a sea change. In the last few decades, the key factors that drove the growth of Asia s mega cities were investment in infrastructure, foreign direct investment and job creation. In the next few years, this growth will need to be supplemented by other factors. By 2030, wealth, innovation and liveability will shape the new mega cities. These new mega cities will provide high-quality living environments to attract affluent populations and have innovative strategies to recruit and retain top talent. More cities will shift their focus from growth to quality, from building infrastructure to innovation and efficiency, and from job creation to fostering a good living environment. To identify Asia s new mega cities in 2030, we looked at cities that can create wealth via innovation and attract top talent by offering a high-quality living environment. We collected data from 302 cities, including Singapore, Hong Kong, Taipei, 291 cities in China, and eight cities in India. For the in-depth analysis, we selected 55 cities, each with a population bigger than 5mn and an annual GDP per capita of over US$8,000. We then ranked the competitiveness of these cities based on three categories: (1) wealth, (2) innovation, and (3) liveability. Indicator categories and measurements Category Indicator Measurement Wealth & growth Population GDP Income Bank deposits Resident for six months or longer GDP per capita GDP growth Average wages of employed staff Value of bank deposits per capita Innovation High value-added sector % of employees in tertiary sector Intellectual property Number of patents applied per 1,000 people Government spending on R&D as % of GDP Government spending on education per capita Universities Number of universities Entrepreneurship % of employees in private sector Liveability International travel connections Number of airline travellers Air quality Monthly average PM2.5 scores Source: DBS Bank

05 China, land of the rising stars Asia s top mega cities in 2030 City Wealth & Growth Innovation Liveability Overall Shenzhen 1 3 4 1 Hong Kong 9 1 1 2 Guangzhou 2 8 3 3 Shanghai 5 5 4 4 Singapore 16 2 2 5 Nanjing 8 4 10 6 Beijing 4 6 13 7 Hangzhou 2 10 16 8 Qingdao 18 12 9 9 Wuhan 10 9 21 10 Taipei 20 13 8 11 Changsha 7 13 22 12 Tianjin 12 7 24 13 Mumbai 23 19 20 20 Bangalore 29 25 29 27 Bangkok 46 23 22 31 Delhi 35 32 33 35 Jakarta 47 32 39 42 Source: DBS Bank The ranking shows that Chinese cities have great potential in terms of innovation and quality growth. By 2030, China will house eight of our top 10 Asian mega cities, with Shenzhen topping the list. Singapore (No. 2) and Hong Kong (No. 5) will remain Asia s wealthiest cities while Shenzhen, Shanghai, and Hangzhou will have GDPs as large or larger than Hong Kong s now. Hong Kong and Singapore ranked highly on the list due to their good living environments and higher government spending on research & development (R&D) and education, but these cities will see slower growth in both population and GDP. Taipei, despite lower population growth, also ranks high on the list, supported by its good living environment. The key Indian cities of Mumbai, Bangalore, and Delhi ranked towards the bottom mainly due to lower scores on a per capita basis. By 2030, Wuhan, Guangzhou will emerge as new mega cities with the highest growth potential in almost all asset classes. Hong Kong, Singapore will keep their leads but real estate growth likely to be smaller than for new Chinese mega cities By 2030, our analysis suggests that Shanghai and Shenzhen will overtake Hong Kong to become the largest Grade A office markets. Beijing, Shanghai and Guangzhou will see stronger retail rental growth than Hong Kong as their retail rents are low compared to their growth potential in terms of retail sales per capita. Inter-regional connectivity will support property markets in Hong Kong and Singapore, but growth rates will decelerate. Guangzhou is expected to register the biggest improvement in office rental rates, while Hangzhou and Wuhan may see more rental upside, if supply is well controlled. In the hotel sector, we believe that there is great potential for hotel revenue per available room (RevPAR) in Asia-Pacific cities to catch up with that in the rest of the world. In the Chinese cities of Shenzhen, Shanghai, and Beijing, hotel RevPars could approach international levels by 2030.

06 Asia s emerging mega cities in 2030 The bedrock of Asia s emerging mega cities is undergoing a sea change. In the last few decades, economic growth, foreign direct investment (FDI) and job creation were the key factors driving a city s competitiveness and attractiveness. This has started to change as quality of growth, innovation, education and air quality come into play. 1. From scale to quality In the past several decades, urbanisation drove growth in city economies. Massive infrastructure developments facilitated the urbanisation process in most Asian cities. Yet, the quality of life has also been affected by overloaded facilities, overcrowded city centres, and pollution. Asia s mega cities of the future are likely to focus on the quality of growth, for example, growing a green economy with minimal or no pollution. 2. From infrastructure investment to innovation investment Asian cities have gone through a decade of swift development and infrastructure upgrades including in transportation and technology. Lifestyles are changing faster than ever, especially in communication and consumption. With smartphone penetration rates shooting up since 2011, technology is reshaping every facet of work and leisure, in communication, transportation and accommodation. In the next decade, we believe that cities with high investment in research & development (R&D), high patent output, and high spending in education will outperform. 3. From job creation to living conditions Over the past decade, we have seen greater attention paid to the effects of air pollution. As media reports on pollution grew, expatriates started to leave highly polluted cities in greater numbers, and city authorities began to work harder to target pollution. In a globally mobile world, top talent increasingly place a premium on an excellent living environment, and this also includes education, healthcare, and travel connectivity. So, by 2030, wealth, innovation and liveability will increasingly shape Asia s mega cities. To identify Asia s mega cities in 2030, we factored in these new influences. We believe Asia s mega cities of the future will house big populations of affluence, provide a foundation for innovation and creativity, and foster excellent living environments.

07 Methodology We collected data on 302 cities in Asia, including Singapore, Hong Kong, Taipei, 291 cities in China, and eight cities in India. Of these, 55 were selected for ranking based on having a population bigger than 5mn and an annual gross domestic product (GDP) per capita of over US$8,000. We ranked these cities by three categories (1) wealth creation, (2) innovation, and (3) liveability. Indicator categories and measurements Category Indicator Measurement Wealth & growth Population GDP Income Bank deposits Resident for six months or longer GDP per capita GDP growth Average wages of employed staff Value of bank deposits per capita Innovation High value-added sector % of employees in tertiary sector Intellectual property Number of patents applied per 1,000 people Government spending on R&D as % of GDP Government spending on education per capita Universities Number of universities Entrepreneurship % of employees in private sector Liveability International travel connections Number of airline travellers Air quality Monthly average PM2.5 scores Asia s top mega cities in 2030 City Wealth & Growth Innovation Liveability Overall Shenzhen 1 3 4 1 Hong Kong 9 1 1 2 Guangzhou 2 8 3 3 Shanghai 5 5 4 4 Singapore 16 2 2 5 Nanjing 8 4 10 6 Beijing 4 6 13 7 Hangzhou 2 10 16 8 Qingdao 18 12 9 9 Wuhan 10 9 21 10 Taipei 20 13 8 11 Changsha 7 13 22 12 Tianjin 12 7 24 13 Mumbai 23 19 20 20 Bangalore 29 25 29 27 Bangkok 46 23 22 31 Delhi 35 32 33 35 Jakarta 47 32 39 42 Sources: DBS Bank

08 Key Findings By 2030, eight of Asia s top 10 mega cities will be in China. Shenzhen tops our list of top Asian mega cities in 2030, with Guangzhou and Shanghai coming in third and fourth, respectively. Our ranking shows that Chinese cities have great potential in terms of innovation and quality of growth. Hong Kong and Singapore ranked second and fifth respectively. Both cities continue to offer good living environments and higher government spending on research & development (R&D) and education, but their growth in both population and GDP is projected to moderate over the next decade. Taipei, despite lower population growth, also ranks high on the list, supported by its good living environment. India s key cities Mumbai, Bangalore, and Delhi came in towards the bottom mainly due to lower scores on a per capita basis. China s rising stars By 2030, China s top eight mega cities will have bigger GDPs than Hong Kong or Singapore now, assuming their GDPs slow by 0.1-0.2% pa over the next decade. By 2030, Shenzhen, Guangzhou and Tianjin will have bigger GDPs than Hong Kong or Singapore now, and narrow the gap with Beijing. Shanghai will maintain its top position in terms of GDP. By 2030, Shenzhen s GDP per capita will surpass Hong Kong s now and move closer to Singapore s current level. Hangzhou s GDP per capita will grow closer to Hong Kong s current level. In 2030, Singapore will continue to have the highest GDP per capita of all the cities on our list. We also assessed the real-estate potential of these mega cities against their economic potential. We found that Wuhan and Guangzhou have the best real-estate investment potential in almost all asset classes. In terms of property prices, Changsha, Wuhan and Guangzhou have the best growth potential while Hong Kong, Singapore, Shanghai and Beijing will witness slower growth. GDP per capita has been and will continue to be the key long-term driver for residential property prices. Using property price-to-gdp per capita ratio as a key indicator, we attempt to predict cities residential price upside potential, assuming zero inflation. Singapore s current price-to-gdp per capita ratio is 24%, which we consider a comfortable level. Applying this level to GDP per capita in 2030, we find that Changsha, Wuhan, and Guangzhou will have more upside potential in their property prices. Residential property price growth in Hong Kong, Singapore, Shanghai, and Beijing have less upside potential. Land shortages to disappear Land shortages faced by Shenzhen, Hong Kong now are unlikely to continue into 2030. We analysed the built-up area as a percentage of total land area for the top 12 Asian mega cities on our list. Shenzhen is the most highly built up at 45%. Hong Kong and Wuhan are around 25%, while Beijing has the lowest ratio mainly due to its larger total land area. In our view, supply factors affect property prices only over a shorter time horizon. Transportation connections, increase in density or plot ratio, and even expansion of the administration areas of cities can break supply constraints in the longer term.

09 By 2030, we expect Hong Kong s pole position as the largest Grade A office market to be overtaken by Shanghai and Shenzhen, driven by higher growth in tertiary sector GDP. Our analysis shows a strong correlation between historical nominal GDP in the service industry, and occupied Grade A office space in the past 12-17 years. We believe this trend will continue. Based on the projected 2030 tertiary GDPs for these mega cities, our regression model shows that by 2030, occupied office space in Shanghai and Shenzhen will grow to 22.7mn sqm and 18.4mn sqm respectively (up 135% and 251% respectively from 2017), exceeding that in Hong Kong or Singapore. Rental growth potential In terms of office rental rates, Guangzhou will register the biggest improvement. Over 2018-2030, Grade A office rental rates are projected to jump by a CAGR of 5.0% for Guangzhou, followed by 4.2% for Shanghai. Tertiary GDP per square metre of office space is the key ratio to examine in rental growth potential. To this end, we used the office space occupancy cost ratio (Grade A office rents over tertiary GDP per sqm). Among the four Chinese tier 1 cities, Guangzhou has the highest office rental growth potential by 2030, while Hong Kong and Singapore will still see 2% and 4% CAGRs respectively over 2018-2030. We also compare the rental costs per square metre of Grade A office space versus tertiary GDP per occupied Grade A office space, and conclude that Hangzhou and Wuhan may offer more rental upside as occupancy costs are low compared to current tertiary GDPs. Wuhan, Shenzhen, Changsha, and Nanjing are likely to see stronger retail sales growth, driven by bigger GDPs and populations. We expect local retail sales per capita to rise in tandem with GDP expansion. This, coupled with growing populations, should continue to drive retail sales growth into 2030. We based our 2030 retail sales per capita on 2017 retail sales per capita multiplied by our 2017-2030 GDP projection. This 2030 retail sales per capita multiplied by our 2030 forecast population gives us our projection for total retail sales for each city in 2030. Based on the above calculation, Wuhan tops the list in terms of retail sales growth, followed by Shenzhen, Changsha, and Nanjing. Beijing, Shanghai, and Guangzhou will offer a greater retail rental upside, whereas Hong Kong is likely to see a slowdown. Comparing ground-floor rents with retail sales per capita, we believe that Beijing, Shanghai, and Guangzhou may have more rental upside as their ground-floor rents are low compared to retail sales per capita. In Hong Kong, we expect upcoming bridges and high-speed trains to bring additional shopper traffic to retail malls. However, given its already higher-than-peer rental rate, we expect rental growth in Hong Kong to slow down. Singapore s retail sales resumed growth in 2017, after declining for the three years of 2014-2016. Retail sales are expected to grow at a low single digit percentage, in tandem with income. Still, Singapore remains a gateway to Southeast Asia and will benefit from the rising affluence in this region. Hotel RevPars to rise In the hotel sector, data shows that the average daily rate (ADR) in Asia-Pacific cities is significantly lower than in other regions despite fast growing demand. However, we

010 believe that there is great potential for the hotel revenue per available room (RevPAR) in Asia-Pacific cities to rise, given that the oversupply is gradually being digested by the fast growing number of tourists, that old hotels are increasingly being converted into other uses, that the pipeline of new room supply is well-controlled, and that the hotel management incentive structure is being refined. In particular, China hotel RevPars can catch up with levels across cities in the United States, Europe and the Middle East by 2030. Historical hotel trends in Shanghai, Hong Kong and New York indicate a strong negative relationship between available room night per international traveller and hotel RevPAR. Current cross-city comparisons also indicate a similar relationship. Given that, we see hotel room rates in Shenzhen and Wuhan as likely to have more upside, while Nanjing s hotels may have to rely more on domestic tourist trade. Hotel RevPars in Shenzhen, Shanghai, and Beijing have every chance of approaching levels in the best cities internationally. Cities to watch 2030 Mega City No. 1 Shenzhen Shenzhen spends more on R&D as a percentage of its GDP than any other city on our list. It also stands out for its spending on education per capita, GDP per capita, air quality, and international connection indicators. Shenzhen, which houses Tencent s headquarters, aims to become one of the world s most sustainable and innovative city by 2030. With total GDP exceeding Hong Kong s in 2017, it plans to grow its population by 2.4% per annum to 2030, the highest growth rate among tier 1 Chinese cities and faster than most cities in the region. By 2030, the mega city of Shenzhen is expected to have a GDP per capita comparable to that of Singapore s or Hong Kong s in 2017. 2030 Mega City No. 2 Hong Kong Hong Kong is a developed economy, with slower GDP growth than most cities and a smaller population. But it ranks high on our list for its higher spending in R&D and education, and for its well-developed services sector, good air quality, and international travel connections. In 2018, it rolled out its plan to attract talents in science and technology. Given its intention to embrace the innovation-driven era, Hong Kong is likely to maintain its leading status in the region. 2030 Mega City No. 3 Guangzhou Guangzhou is the all-rounder city, with balanced high scores in all the three categories we studied. It enjoys high rankings in both GDP and deposit per capita. With moderate growth in GDP, Guangzhou ranks No. 2 in the wealth and growth category. Good air quality, international travel connections and a large number of universities enable it to be ranked among the top 8 and top 3 in the innovation and liveability categories, respectively. We believe the integrated development in the Greater Bay area and reasonable local property prices will help Guangzhou city to continue to stand out in the next decade.

011 2030 Mega City No. 4 Shanghai Shanghai is big, wealthy and innovative. It has two of the busiest airports in China and the second largest city population (Chongqing has the biggest). Shanghai has made great strides in R&D and education, and ranked highly on our list in terms of GDP and bank deposits per capita. Shanghai, as the faucet city in the Yangtze River Delta region, leads the wealthy cities of Zhejiang and Jiangsu provinces, and aims to be an international economic, finance, trade, and shipping centre. Shanghai intends to curb its population growth in the coming decade, with planned annual growth the lowest among key cities at 0.23% CAGR up to 2030. 2030 Mega City No. 5 Singapore Singapore is a developed economy with slower GDP growth and a smaller population than most cities. It earns its high ranking from its higher spending on R&D and education, its developed services sector, its good air quality, and its international connections. Singapore aims to remain a leader in all aspects of sustainability, from technological advancement, to green investments, policies and lifestyle. These initiatives will be co-created between the public and private sectors. Driving these targets will be the Smart Nation initiative, where people are empowered by technology to lead more meaningful and fulfilled lives. As Singapore is compact with high population density, the government will continue to invest in public transit infrastructure so people can travel seamlessly in a car-lite environment. The city state aims to actively reduce its carbon footprint and be a zero-waste nation by 2030. 2030 Mega City No. 6 Nanjing Nanjing, China s top tier-2 city, made our list because of its wealth on a GDP per capita basis, many universities, high ratio of patent retention, and strong inter-city connections in the Yangtze River Delta region. Nanjing aims to be a city of innovation and has rolled out a talent attraction plan to allow young graduates to obtain local residency, and housing purchase and rental subsidies. It also provides subsidies and other assistance to encourage start-ups. 2030 Mega City No. 7 Beijing Beijing, despite being in the more polluted north of China, made our top 10 due to its sizeable population, high average wages, its position as China s gateway city connecting the world, and its high government spending on R&D and education. Although Beijing is keeping its annual population growth below 0.7%, it aims to attract worldwide talents in the areas of start-ups, technology, and culture and creative industries by granting local residency status and start-up subsidies. 2030 Mega City No. 8 Hangzhou Home to the world s most successful e-commerce company Alibaba, Hangzhou enjoys one of the highest GDP growth rates among Asian cities, and has high bank deposits and many patents per capita. Hangzhou will curb population growth in the next decade, selectively attracting young educated graduates in the areas of computer science, environmental science, telecommunication, biotech, new materials, architecture, marketing and logistics, to support its economic growth.

12 2030 Mega City No. 9 Qingdao Located in the wealthy Chinese province of Shandong, Qingdao is the only northern city other than Beijing to make the top 10 on our list. It stands out for its good air quality and international travel connections. Qingdao s government also spent more on education than any other tier 2 Chinese city. 2030 Mega City No. 10 Wuhan Wuhan, home to some of the most reputable universities in the country, squeaks into the top 10 with its efforts to become China s Silicon Valley. Wuhan spends a large proportion of its GDP on R&D and has invested handsomely in its R&D centre and business park, enabling it to benefit in this digital age. The ninth largest city by population in China, Wuhan plans to grow its population by 3.4% per annum in the coming decade, which is the fastest of all the cities on our list. 2030 Mega City No. 11 Taipei Taipei scored well in terms of liveability and bank deposits per capita but fell outside the top 10 due to slower GDP and population growth. Government spending on education per capita and number of top 200 universities were also high but R&D spending as a percentage of GDP and number of patents received were lower than for other top-ranked cities. 2030 Mega City No. 12 Changsha As a core city in central China, Changsha aims to be an intelligent manufacturing centre, an innovative and creative centre, and a transportation and logistics centre. Changsha s GDP per capita ranks among the top 10, as does its number of patents and universities. Changsha is aggressive in attracting talents by providing housing purchase subsidies to young graduates and fast approval for local hukou (or residency) application. Changsha s population growth plan for the next decade is also one of the highest. 2030 Mega City No. 13 Tianjin Located in northern China next to Beijing, Tianjin houses a sizeable population and makes solid efforts in R&D and education. To support environmental improvement, Tianjin s GDP growth has slowed significantly to 3.6% in 2017 with the shuttering of old, highly pollutive industrial plants. To revitalise the economy and realise a speedy upgrade, Tianjin has been aggressive in accepting talent inflow and fostering new industries.

13 Asia s mega cities in 2030 Beijing: 7 Tianjin: 13 Delhi: 35 Mumbai: 20 Changsha: 12 Nanjing: 6 Wuhan: 10 Guangzhou: 3 Shenzhen: 1 Qingdao: 9 Hangzhou: 8 Taipei: 11 Hong Kong: 2 Shanghai: 4 Bengaluru: 27 Bangkok: 31 Singapore: 5 Jakarta: 42 Source: DBS HK

14 Cities with best property price potential Rapid urbanisation in recent years has left many cities facing challenges in transportation, healthcare, and education. As a result, many Chinese cities now have comprehensive population growth plans to curb and control resident numbers. This has a direct effect on demand for property. Wuhan and Shenzhen top the planned population growth plans by forecasting 3.4% and 2.4% CAGR up to 2030, while Shanghai and Hangzhou will focus on population curbs with a low 0.23-0.24% CAGR. Still, Shanghai is expected to continue to have China s largest population for a city. Population by 2030 30 mn Population 20 Present 2030E 10 - Shanghai Beijing Guangzhou Shenzhen Hangzhou Nanjing Tianjin Wuhan Changsha Qingdao Singapore Hong Kong Source: NDRC On the supply side, Shenzhen is the most built-up city out of the top 12 on our list. We analysed the built-up area as a percentage of total land area and found that Shenzhen is the most built up with 45%, followed by Hong Kong and Wuhan with around 25%. Beijing has the lowest ratio mainly due to its larger total land area. However, in our view, these near-term supply shortages are unlike to continue into 2030. Supply may affect property prices only in the shorter time horizon. Transportation connections, increase in density or plot ratio, and even expansions in the administration areas of cities can break supply constraints in the longer term.

15 In the short term, residential property prices are affected by multiple factors such as policies, supply, credit environment, and price expectations. In the long term, we believe they are driven by the wealth of a city s population, in other words, GDP per capita. GDP per capita by 2030 Rmb k 500 450 Present 2030E 400 350 300 250 200 150 100 50 0 Shanghai Beijing Guangzhou Shenzhen Hangzhou Nanjing Tianjin Wuhan Changsha Qingdao Singapore Hong Kong Source: NDRC By 2030, the top eight Chinese mega cities will have GDPs bigger than Singapore or Hong Kong now. GDP growth in Chinese cities ranged between 3.6% and 9% in 2017, mostly slowing down. Singapore and Hong Kong, however, improved from 2.4% and 2.3% in 2016 and to 3.6% and 3.8% in 2017 respectively. Assuming a 0.2-ppt per annum slowdown in GDP growth in Chinese cities and 0.1ppt slowdown for Singapore and Hong Kong, we expect every highranking Chinese city on our list to grow its GDP in 2030 to beyond Hong Kong or Singapore s current GDP. Shenzhen s GDP in 2030 will exceed Hong Kong or Singapore s current GDP and catch up to Beijing s. Shanghai will still top the list in terms of GDP by 2030. Singapore will remain the wealthiest city in the region in 2030. Shenzhen s GDP per capita by 2030 will exceed Singapore and Hong Kong s current levels, while Hangzhou s GDP per capita is approaching the current level in Hong Kong. Using property price-to-gdp per capita ratio as a key indicator, we predicted cities residential price upside potential, assuming zero inflation. We use Singapore s property price-to-gdp per capita ratio, which is 24%, as a reasonable benchmark for our forecast.

16 Key findings Applying this ratio to all cities GDPs per capita by 2030, without considering inflation, we find that Changsha, Wuhan, and Guangzhou will have the greatest upside in their property prices. Hong Kong, Singapore, Shanghai, and Beijing have less upside in residential property prices based on local economic fundamentals. However, Singapore property prices might still exceed Hong Kong s by 2030. Shenzhen s average property prices, although the highest among mainland cities, will still be lower than for Singapore and Hong Kong. Singapore s residential property priceto-gdp per capital ratio is about the city average in our study. Singapore s property prices are likely to be driven by GDP per capita growth in the next decade. Its property price CAGR is likely to be around 2%, compared to the relatively flat trend in Hong Kong due to the gradual normalisation in the price-to-gdp per capita ratio. Price upside Rmb/sm 120,000 100,000 80,000 60,000 40,000 20,000 0 Shanghai Present 2030 CAGR Beijing Guangzhou Shenzhen Hangzhou Nanjing Tianjin Wuhan Changsha Qingdao Singapore Hong Kong 14% 12% 10% 8% 6% 4% 2% 0% Source: DBS Changsha, Wuhan, and Guangzhou will have the greatest upside in their property prices

17 Cities with best office investment value The rapid expansion of China s tertiary (or service) sector has driven up demand for office space and we expect this to continue. In many key Chinese cities, the service industry s contribution to local GDP is still low compared to in key economic areas/cities globally and there is much room for growth. We expect the rapid urbanisation of the past few decades and rising levels of higher tertiary education to continue to drive industry upgrades, alongside regional collaborations such as in the Jing-Jin-Ji, the Yangtze River Delta and the Greater Bay Area. We also expect the high-tech and fin-tech sectors to become the key drivers for Grade A office space. In particular, Asia has seen a flourishing of unicorns, or privately-held start-up companies valued above US$1bn. According to Deloitte, there are 115 unicorns in Asia, accounting for 46% of the global total. China has 98, India 10, Korea three, Singapore two, and Indonesia one. China s unicorn distribution, by valuation China s unicorn distribution, by number of companies Greater Bay 11% Other regions 4% Jing-Jin- Ji 44% Greater Bay 14% Other regions 5% Jing-Jin-Ji 44% Yangtze River Delta 41% Yangtze River Delta 37% Sources: Ministry of Science and Technology of PRC, DBS

18 Watch the unicorns We expect these unicorns to be the next-generation corporations, which will be big GDP contributors and Grade-A office occupiers. Cities to watch are Beijing, Hangzhou, and Shanghai, which house the biggest unicorn companies in terms of valuation, according to China s Ministry of Science and Technology. Shanghai and Shenzhen each have their ambitions as global financial centres. Demand for office space in these cities can only increase, with the liberalisation of the financial sector, growing adoption of fintech, and introduction of ever more financial products (upcoming C-REITs) and market connections (SH/SZ-HK stock connect, bond connect and potential SH- London stock connect, etc). Across ASEAN, intra-region capital flows are expected to expand as more financial initiatives (such as REITs) are introduced. India is likely to kick off its first REIT in 2018, having established its REIT code in 2014. Indonesia and Philippines are also working on potential tax reforms to foster or accelerate local REIT development. Singapore, as the largest REIT market in Asia ex- Japan, will continue to attract global assets to list REIT products. Methodology To establish the correlation between tertiary sector GDP and office space, we ran linear regressions for historical nominal GDPs for the service industry and occupied Grade A office spaces for each tier 1 city in China for the past 12-17 years. We found strong correlation coefficients of between 98% and 99%, and believe this will continue. Shanghai: Linear regression between tertiary GDP and occupied Grade A office space Beijing: Linear regression between tertiary GDP and occupied Grade A office space Sources: JLL, DBS

19 In key Chinese cities, tertiary industry as a percentage of local GDP has risen by 0.8-2.5ppts each year over the last several years. We use this trend as a key assumption to forecast the service industry s local GDP contribution in 2030, subject to some adjustments based on regional financial and high-technology potential. Hong Kong s economy is closely tied to China s growth. Hong Kong s tertiary industry contributed 93.0% of its GDP in 2017 the top rate among Asian cities and has kept the rate within the range of 92.2-93.1% over the past decade. We expect it to remain at the level for the next decade. Singapore s tertiary industry contributed 73.8% of its GDP in 2016, the fourth highest rate in Asia. Unlike Hong Kong, Singapore s growth is correlated with the ASEAN economy, given its location and role as a gateway to the region. Southeast Asia has a young population with an average age of 30 years, and the demographic dividend will continue to drive economic growth in the region. In addition, China s belt-and-road initiatives have enhanced intra-asia cooperation and capital flow. Southeast Asia s economic growth will continue to underpin demand for office space, especially in Singapore, which houses the regional headquarters of many multinational corporations. Singapore s tertiary industry contribution to GDP has ranged from 70.6-75.2% over the past decade and we expect it to edge up to 75.0% in 2030. In 2017, Beijing and Shanghai outpaced Hong Kong in terms of tertiary GDP, while Guangzhou s tertiary GDP also surpassed Singapore s. We expect Beijing to keep its lead among Chinese cities in 2030, in terms of service industry s contribution to GDP, followed by Shanghai, Tianjin, Guangzhou, Hangzhou, and Shenzhen. In mature markets like Hong Kong and Singapore, we expect the tertiary industry s contribution to local GDP to stay relatively steady into 2030. As a result, we expect the four tier-1 cities in China to outpace Hong Kong in terms of tertiary GDP and the key tier-2 cities to surpass Singapore. Tertiary GDP and as % of GDP, 2017 Tertiary GDP Tertiary as % of total RMB bn 2,500 100% 2,000 80% 1,500 60% 1,000 40% 500 20% 0 0% Beijing Shanghai Hong Kong Guangzhou Singapore Shenzhen Tianjin Hangzhou Wuhan Nanjing Qingdao Changsha Tertiary GDP and as % of GDP, 2030 Tertiary GDP RMB bn 6,000 5,000 4,000 3,000 2,000 1,000 0 Shanghai Beijing Shenzhen Guangzhou Tianjin Hong Kong Hangzhou Tertiary as % of total Wuhan Nanjing Singapore Changsha Qingdao 100% 80% 60% 40% 20% 0% Sources: CEIC, DBS

20 We estimated the 2030 Grade-A office space data in the four tier-1 cities in China, based on the regression relationship between historical occupied Grade-A office space and local tertiary GDP, as well as our forecasted 2030 tertiary GDP. For mature markets like Hong Kong and Singapore, we used their occupied Grade-A office space per tertiary GDP in 2017, multiplied by our estimated 2030 tertiary GDP to forecast their demand for office space in 2030. Occupied office stock space, 2030 sqm 25,000,000 2017 2030 20,000,000 15,000,000 10,000,000 5,000,000 0 Findings Shanghai Shenzhen Hong Kong Beijing Guangzhou Singapore Source: JLL, DBS In 2030, we expect Shanghai and Shenzhen to replace Hong Kong as the largest providers of Grade-A office space. In Shanghai, we expect the tertiary industry to occupy 22.7mn sqm in 2030, up from 135% from 2017. For Shenzhen, the tertiary sector will take up 18.4mn sqm in 2030, up 251% from 2017. Across the four tier-1 cities of Shanghai, Shenzhen, Beijing and Guangzhou, we expect land supply to be tightly controlled given the limited acreage, with occupancy rates in 2030 ranging between 85% and 98%. Rental costs and rates To gauge the occupancy cost for tertiary industries across the cities, we use total office rental costs (occupied Grade-A office space multiplied by annual rents) divided by tertiary GDP as a proxy. The ratio is 1.0% for Singapore and 5.6% for Hong Kong. For the four tier-1 Chinese cities of Shanghai, Shenzhen, Beijing and Guangzhou, this ratio ranged from 0.6-1.5%, growing over the past decade at an average of 0.07-0.08ppt per annum. By 2030, assuming this trend continues, we expect occupancy costs to go up to 1.7-2.3% for tertiary industries across the four cities. Then, we derived the total Grade-A office rental costs in 2030 for each city based on our forecasted 2030 tertiary industry s occupancy cost multiplied by forecasted 2030 tertiary GDP. After that, we use total rental costs divided by 2030 forecasted occupied Grade-A office to calculate our forecasted rental rate for 2030. For tier 1 cities, we expect Guangzhou to register the highest CAGR of 5% for Grade-A office rental rate in 2017-2030, followed by Shanghai s 4.2%.

21 Hong Kong s rental rate in the central area has spiked on strong demand from China companies. But we believe the upcoming infrastructure projects (Hong Kong-Zhuhai- Macau Bridge, Shenzhen-Zhongshan Bridge, Hong Kong portion of high-speed train) will reduce travel time and costs in the intra-greater Bay Area, moving office demand instead to Guangzhou, Shenzhen and the decentralised areas in Hong Kong, where rents are now at a 76%, 64% and 67% discount respectively to CBD rents in Hong Kong. We therefore expect a 2.0% CAGR for Hong Kong CBD s rents over 2017-2030. Singapore office rents have recovered from 2H17, and we expect further rental growth, given limited supply ahead (0.63 mn sqf per annum scheduled in 2018-2023 versus 1.38 mn sqf per annum in 2010-2017). Additional supply is likely to be carefully calibrated to support business needs and keep business costs affordable. We therefore expect a 3-5% CAGR for Singapore office rents in 2018-2030. 2030 rental forecast and CAGR growth (2017-2030) USD/sm/year 3,000 2017 rental 2030 rental CAGR growth 2,500 2,000 1,500 1,000 500 6% 5% 4% 3% 2% 1% 0 0% Beijing Shanghai Shenzhen Guangzhou HongKong Singapore Source: JLL, DBS Upsides for Hangzhou, Wuhan We also compare the rental costs per Grade-A office space with tertiary GDP per occupied Grade-A office space, and observed that Hangzhou and Wuhan may have more rental upside if supply is well controlled, as the occupancy costs are low compared to tertiary GDP generated. Co-working spaces are rising in popularity but we do not expect these to have a significant impact on Grade-A office space. According to JLL, co-working spaces recorded a 35.7% CAGR in the Asia Pacific over 2014-2017, as compared to 25.7% in the US and 21.6% in Europe. So far, co-working space penetration (as a percentage of total office space) has ranged from 1.2-3.7% in regional gateway cities in the Asia Pacific. However, around 80% of those co-working space were from the conversion of Grade-B offices, and the upper floors in retail malls and hotels. Therefore, we expect co-working spaces to have a limited impact on Grade-A office space supplydemand dynamics.

22 Cities with best retail rent growth Prime retail space is generally scarce in Asian cities and we expect pressure on supply to continue, even with the healthy pipeline of retail mall projects. In addition, while shopping online is very popular, shopping habits are increasingly leaning back towards brick-and-mortar stores. Millennials are leading the way back to physical shops in their search for greater convenience, product quality and authenticity, and individualisation. This, coupled with growing spending power as incomes rise, is safeguarding demand for physical retailers. Even e-retailers are sitting up and opening brick-and-mortar stores, especially in China, where 35% of the population in 2030 will comprise high-income and upper-middle income groups, up from 10% in 2015, according to the Economist Intelligence Unit. In China, there is low mall penetration in most cities. Across key Asian cities, the average retail space per capita is around 0.45 sqm, versus 1.3 sqm in the New York Bay area and 2.5 sqm in the San Francisco Bay area. China remains the key area with a pipeline of retail mall projects. Emerging markets, especially in Asia, also remain highly active in terms of delivery of retail mall space. This has raised oversupply concerns for some but we believe the supply risk is overestimated. This is because (i) a large percentage of pipeline malls has been delayed or cancelled, and we expect this to continue; (ii) the supply risk for well-managed malls could be overstated as non-performing malls might not be direct competitors; and because of (iii) the latest trend to convert mall space into co-working space and China s new policy to allow the conversion of commercial land into long-term rental housing usage.

23 Retail space per capita comparison (2017) sm 3.0 2.5 2.0 1.5 1.0 0.5 0.0 San Francisco Bay New York Bay Wuhan Nanjing Singapore Shanghai Changsha HongKong Hangzhou Greater Tokyo Bay Qingdao Shenzhen Beijing Tianjin Guangzhou Source: CBRE, JLL, CEIC, DBS Retail sales stars in 2030 The Chinese city of Wuhan tops our list of retail sales stars in 2030. Wuhan is expected to register the biggest growth in retail sales per capita growth, followed by Shenzhen, Changsha, and Nanjing. We expect local retail sales per capita to rise in tandem with GDP. This, coupled with growing populations, should continue to drive retail sales growth into 2030. Therefore, we base our forecast for 2030 retail sales per capita on 2017 retail sales per capita multiplied by our 2017-2030 GDP growth projection. The 2030 retail sales per capita multiplied by our 2030 forecasted population results in our projection for total retail sales for each city in 2030. Retail sales per capita projections Retail sales projections Rmb 200,000 160,000 120,000 80,000 40,000 0 2017 retail sales per capita 2030 retail sales per capita Growth 250% 200% 150% 100% 50% 0% Rmb mn 3,000,000 2,500,000 2,000,000 1,500,000 1,000,000 500,000 0 2017 retail sales 2030 retail sales Growth 400% 320% 240% 160% 80% 0% Nanjing Guangzhou Shenzhen Hangzhou Changsha Wuhan Beijing Shanghai Qingdao HongKong Tianjin Singapore Beijing Shanghai Guangzhou Shenzhen Wuhan Nanjing Tianjin Hangzhou Changsha Qingdao HongKong Singapore Source: CEIC, DBS

24 The sophisticated cities of Beijing and Shanghai are still attracting both shoppers and brand tenants nationally and internationally. Demand in both cities remains strong and we expect existing luxury malls in core areas and suburban malls with strong population influx to keep outperforming. Beijing, however, can expect to see more rental upside, as ground-floor rents in the Chinese capital are low compared to retail sales per capita. By the same measure, we expect Guangzhou to see more rental upside. Nanjing and Hangzhou may also see more rental upside if supply is well controlled. Ground-floor rents vs retail sales per capita (by end-2017) Ground - 50,000 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 rents (Rmb/sm/year) Singapore Shenzhen Changsha Wuhan Tianjin Qingdao HongKong Shanghai Beijing Hangzhou Guangzhou Nanjing 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 Retail sales per capita (Rmb/year) Source: JLL, DBS For Hong Kong, we expect upcoming bridges and high-speed trains to bring additional shopper traffic to retail malls. However, given its already higher-than-peer rental rate, we expect rental growth in Hong Kong to slow down. Singapore s retail sales regained growth in 2017, after three years of decline in 2014-2016. Retail sales are expected to grow at a low single-digit percentage, in tandem with income. On the other hand, the supply of new retail malls in the next five years is low and will be in suburban towns supporting the needs of catchment populations. Still, Singapore remains a gateway to Southeast Asia and will benefit from the region s rising affluence, so we expect its retail scene to remain healthy and landlords to continue to enjoy strong portfolio occupancy rates across market cycles. Beijing and Guangzhou also have a limited supply pipeline. Given the limited land supply in Beijing and Shanghai, we also see a rising trend in the conversion of retail space into office or co-working spaces, which could cap future supply.

25 New supply in next four years (five years for Beijing, Shanghai) sm 14,000,000 12,000,000 10,000,000 8,000,000 6,000,000 4,000,000 2,000,000 0 Changsha Existing stock New supply in the next 4 years Shenzhen Hangzhou Shanghai Wuhan Nanjing Guangzhou New supply as % of existing stock 60% 50% 40% 30% 20% 10% 0% Tianjin Qingdao Beijing Source: JLL, DBS Existing retail space and occupancy (by end-2017) sm 14,000,000 Existing stock Occupancy 98% 12,000,000 96% 10,000,000 94% 8,000,000 92% 6,000,000 90% 4,000,000 88% 2,000,000 86% 0 84% Shanghai Beijing Wuhan Nanjing Shenzhen Hangzhou Tianjin Changsha Qingdao Guangzhou Source: JLL, DBS

26 Cities with best outlook on hotel assets A ccording to STR Global, demand for luxury hotels across the Asia Pacific has grown at a blistering CAGR of 6.9% over the past three years, heaps ahead of other regions CAGRs of 1.3-3.0%. However, APAC hotels suffer from lower average daily room rates (ADRs). In our view, this is due to (i) a temporary oversupply which has lowered occupancy rates, especially in China, and (ii) the outsourcing of luxury hotel management to international operators by APAC developers or landlords, who do not usually have this capability. Generally speaking, these hotel managers are responsible for daily operations and marketing with a term of 15-30 years. Also, the hotel managers provide global market and advertising, centralised reservations, sales services and other hotel-specific services. In return, the hotel managers charge basic fees (normally a certain percentage of hotel revenue) and incentive fees (a certain percentage of hotel gross operating profit) depending on hotel operating results (criteria vary with some purely based on occupancy rate while others are based on gross operating profit). In the earlier years, landlords had less experience in negotiating terms with hotel managers and some contracts were based purely on incentive fees linked to hotel occupancy rates. These left hotel managers more willing to keep ADRs low to secure occupancy rates to be eligible for incentive fees. As a result, we observed an interesting finding: current hotel ADRs are highly correlated to local labour costs. Payroll usually accounts for around half of operating costs for luxury hotels. Our analysis reveals that the correlation between star-rated hotel room rates and local wages is as high as 0.93. In our view, ADR increases in the past have mainly been pushed up by rising labour costs.

27 RevPAR vs wages: An 11-city comparison Wage: Rmb per annual 230,000 210,000 190,000 170,000 150,000 Correlation = 0.93 Singapore Hong Kong 130,000 110,000 Tianjin Nanjing Beijing Shanghai 90,000 HangzhouShenzhen Guangzhou Changsha 70,000 Wuhan 50,000 200 300 400 500 600 700 800 900 1,000 RevPar:R mb/night Source: DBS Luxury hotel ADR comparison (US$/room/night) Luxury hotel occupancy comparison US$/room/night 400 350 300 250 200 150 100 50 0 2016 2017 2018 Amercias Europe APAC Middle-East Asia 72 70 68 66 64 62 60 58 56 % 2016 2017 2018 Amercias Europe APAC Middle-East Asia Source: STR Global, Hotel News Now, DBS Bullish on Asia Still, we are bullish on hotels in Asia for various reasons. (i) Demand in APAC to remain robust</bold> As previously noted, demand for luxury hotels in APAC registered a CAGR of 6.9% over the past three years and we expect such a trend to continue. China s belt-and-road initiatives will continue to build up intra-regional cooperation and connections, driving up demand for hotels. In the key gateway cities of China, demand is mainly driven by domestic travellers and consumption upgrades.

28 Meanwhile, Chinese outbound travellers are one of the key factors driving the regional hotel recovery. The number of Chinese outbound travellers posted a 23% CAGR during the past five years. We expect the current consumption upgrade trend to continue. (ii) More hotels being converted into offices In key gateway cities with limited land supply, the desultory returns on commercial properties is driving the conversion of hotels into offices. Based on our calculation, luxury hotels EBITDA in Beijing or Shanghai is only one-third of Grade-A offices EBITDA. According to CBRE, there were five hotel/mall transactions (total consideration of RMB13bn) in Shanghai and Beijing during 2014-2015 involving conversion into office space. The number of such transactions increased to 16 during 2016-2017, with a total consideration of RMB23bn. We expect such a trend to continue, with more hotels being converted into office space, or co-working space and long-term rental apartments in China, given the government s policy inclination towards the rental housing business. Aged hotels with good location and convenient transportation links but weak financial performances are the major targets for conversion into long-term rental housing. In Hong Kong, given the buoyant commercial property valuations, we believe an increasing number of well-located three -to four-star hotels will be redeveloped into office or commercial buildings. If the Excelsior, J Plus, and Crowne Plaza Hong Kong Causeway Bay are redeveloped into commercial properties, hotel inventory in the area would be cut by c.1,200 rooms or 12%. This could moderate the net growth in hotel-room supply, which should in turn push up the hotel occupancy as well as RevPAR growth. Hotel rooms per million local population 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0 New York HongKong Beijing Singapore Shanghai Tokyo Guangzhou Chengdu Source: STR Global, Hotel News Now, CEIC, DBS

29 (iii) Supply to remain tight We expect supply to stay tight as both government and developers rationalise hotel investments. Following the value-added tax reform in China which lowered taxes generated from hotel assets, local governments have less incentive to issue hotel land for sale. Previously, the business tax was 5.6% on hotel revenue, but is now 6% of gross profit. In addition, income tax generated from the hotel sector is limited, given hotels inferior financial performance. Shanghai, Beijing and Guangzhou are expected to see the lowest supply coming onstream ahead as a percentage of existing supply. We also see no oversupply hangover on the horizon. Beijing s 2008 Olympics and Shanghai s 2010 World Expo caused a significant increase in star-rated hotels to support large tourist inflows during the events. Despite strong hotel performances in the event years, the local hotel market experienced several years of downturn afterwards, as the number of tourists dwindled. We are unaware of any events in the next decade capable of triggering a significant spike in hotel supply and therefore, an oversupply hangover. Existing hotel rooms and future supply Rooms in construction 300,000 Existing hotel rooms 250,000 200,000 150,000 100,000 50,000 0 Frankfurt Nashville New York Bali Dallas London Houston Future supply as % of existing stock Istanbul Los Angeles Shanghai Beijing 25% 20% 15% 10% 5% 0% Source: STR Global, Hotel News Now, CEIC, DBS

30 World-class events in the next decade Football World Cup (in various cities) 2022 Qatar 2026 Canada/Mexico/United States 2030 TBD Rugby World Cup (in various cities) 2019 Japan 2023 France Summer Olympics 2020 Tokyo 2024 Paris 2028 Los Angeles Winter Olympics 2022 Beijing 2026 TBD 2030 TBD World Expo 2020 Dubai 2023 Buenos Aires 2025 TBD 2028 TBD Source: DBS (iv) Hotel owners regaining control As earlier mentioned, landlords had less experience in negotiating contract terms with hotel managers in the early years. However, given the gradual conclusion of contract periods for the first batch of luxury hotels, we believe hotel owners will regain their bargaining power over hotel managers, as there is limited future supply and hotel managers will have to accept less favourable terms to maintain their presence in key cities. For example, when the contract expired for the Shanghai Westin hotel in Jingan district, the landlord took back the hotel and rebranded it as Kunlun Hotel, leaving Westin with no presence in the core districts of Shanghai. Going forward, we believe the interests of hotel managers will be more closely aligned with hotel owners, and both parties will work together to improve ADRs and hotel operations. So far, luxury hotel room rates in major Asian cities (except for Hong Kong, Singapore and Tokyo) are still far below those in global gateway cities. However, we see much potential for Asian city room rates to catch up with their global peers in the next decade.

31 Luxury hotels ADR and occupancy comparison US$/room/night 500 ADR Occupancy 90% 450 85% 400 80% 350 75% 300 70% 250 65% 200 60% 150 55% 100 50% Beijing Shanghai Shenzhen Guangzhou Bangkok HongKong Singapore Tokyo Amsterdam London New York Source: STR Global, Hotel News Now, CEIC, DBS Asian hotel growth drivers Stronger intra-regional connection, consumers trading up and a well-controlled supply will drive up long-term growth for the hotel sector. In addition, the rising popularity of REIT products in the region will also speed up consolidation in the hotel industry, as evidenced by the experience of hotels in the US. In the early 1990s, the US hotel industry also suffered from oversupply. After the supply of new assets slowed down, there was a gradual recovery in hotel ADR and RevPAR. Coupled with the emergence of modern equity REITs, this resulted in mounting acquisition activity by REITs and a larger number of hotel rooms being controlled by fewer players. This has also led to a continuous increase in RevPAR, except during the two Black Swan periods of 9/11 and the housing/banking crisis. Asian hotel stars in 2030 Shenzhen and Wuhan are among the Asian hotel sector stars in 2030, based on their potential for growth in hotel room rates. Historical trends for hotels across Shanghai, Hong Kong and New York all indicate a strong negative relationship between available room night per international traveller and hotel RevPAR, as does current cross-city comparisons. Based on this, Shenzhen and Wuhan s hotel room rates are likely to have more upside, while Beijing and Nanjing s might need to rely more on domestic tourists to take up available rooms. In terms of revenue growth, the cities to watch are Shanghai, Hangzhou and Shenzhen.

32 As noted, we found a strong correlation between hotel RevPAR and local wages and expect this to continue. We project that local wages will grow in tandem with GDP per capita growth and made a forecast for 2030 RevPAR based on 2017 RevPAR multiplied by our 2017-2030 GDP per capita growth projection. Based on the above calculation, Shanghai, Hangzhou and Shenzhen are likely to enjoy higher RevPAR growth. Hong Kong and Singapore will see their CAGRs slow but continue to enjoy top RevPAR in 2030. RevPAR vs available room per international tourist Hotel RevPAR 2017-2030 Source: DBS Rmb/room/night 1,400 2017 RevPAR 2030 RevPAR CAGR growth 1,200 1,000 800 600 400 200 0 Shanghai Beijing Guangzhou Shenzhen Hangzhou Nanjing Tianjin Wuhan Changsha Singapore Hong Kong 6% 5% 4% 3% 2% 1% 0% Source: DBS

33