Both in the evaluation of investment ideas and in the design of buildings, not to mention their operation, artificial intelligence holds distinct promise. But is AI the “new ESG”, an acronym thrown around so much that it loses its power, and does not fulfil its promise? It is highly unlikely. Like it or not, AI is already changing our lives, and will surely shape the future of most property professions. Beyond all the novelty and scaremongering, there is more to “like” about AI than “not”. Big data and proprietary models already inform the investment decisions at firms such as CenterSquare Investment Management. That’s true both for the company’s investment into property stocks as well as its direct investment into real estate itself. The equity analysis looks at how property stocks perform in different market conditions, while asset-specific data helps inform the identification of properties that stand to out­perform in terms of returns. There’s still a human piloting the decision. But CenterSquare’s Joachim Kehr, portfolio and regional manager, real estate securities for Asia Pacific, anticipates that our comfort with letting AI take the wheel may chart a similar course to that of self-driving cars. “The complete use of AI in investment decision making will take some time to penetrate the market, but it is not a matter of ‘if’ but ‘when,’” notes Kehr. However, there’s a fiduciary duty for an investment manager that is always going to necessitate human knowledge built up over years of investment experience. “That knowledge and experience are going to take some time for machines to replicate and prove out,” adds Kehr. AI elements are far likelier to “invade” aspects of data gathering and processing long before a computer system is making totally automated investment decisions. Cutting out the overdesign On the operational side, the uses of AI are far more obvious. In fact, the “sustainability” sliver of ESG stands to benefit greatly from the use of AI, combining last year’s issue of the day with the current meta. Alex Bent, managing partner at Undivided Ventures, looks to invest in companies that use tech and innovation that are “solving for a sustainability pain point in the real estate or infrastructure sectors,” he explains. While excited about the potential of AI, he’s aware it may render some companies or innovations irrelevant as long-term investment targets. Ultimately, he’s confident most property professionals can grow with and adapt to AI. “There is a lot of waste in terms of time and resources in the real estate industry, with day-to-day operating decisions, construction decisions and financial decisions, and these conclusions can be more effectively generated with AI,” says Bent. His company has also recently invested into StructurePal, a company that uses AI to cut down on the over­design excess built into the building process, cutting back both costs and carbon emissions. That should help improve the ratings buildings achieve, certifying their green or tech credentials. The company WiredScore offers both its eponymous WiredScore certification rating a building’s digital connectivity and its SmartScore certification reflecting the smart functionality of buildings including how well they can capture and analyse data. Both certifications are intended to inform the decisions of occupiers, operators, owners and investors. “We are in an age of implementation with AI solutions,” says Thomasin Crowley, WiredScore’s global director for the Asia Pacific region, something she sees growing as systems become more cost effective and accurate. “Landlords are becoming increasingly aware that they need to future-proof their assets to keep up with the evolving needs of occupiers,” she continues. “All in all, the core aim of AI application is to minimise manual input, processing time, and amplify human productivity. And I think we’re unanimously striving for this across the real estate sector.” The proliferation of digital twins In other words, it’s an exciting rather than scary time. Kehr is particularly excited about the use of “digital twins”, the creation of virtual 3D models of real-world buildings. Those “will be incredibly influential in how we plan, build, transact and operate real estate in the future,” says Kehr. Jonathan Hannam, managing partner at Taronga Ventures, agrees. His company has invested into Open­Space, a San Francisco–based software company that has created technology to project a real-world “digital reality” of a construction site or building onto building plans and models. This can help detect deviation away from the plan and red flag construction errors in real time, before they get out of hand. OpenSpace is also building its data pool of construction work, with 17 billion images of construction on file, and growing. That is “teaching” the OpenSpace system how to identify the kind of costly mistakes that drive up construction costs in real estate and extend completion times. As it stands, all too often, the slight error isn’t detected until well after it is built into the site, necessitating expensive reworks. For “pen and paper” property construction, about as traditional a business as you can get, “the digitisation of plans has been a key enabler step for the industry,” says Hannam. And it sets up the implementation of new tech not only in construction, but also building management and financial management. “Let’s say you are acquiring an asset in the future,” notes Hannam. The purchase will also include the digital twin, showing a rendition, original plans and how the building was really built. “This will also be valuable for insurers, who are likely to offer lower premiums based on the higher level of knowledge, and to bankers providing mortgages.” Digital real estate becomes a real asset. On the marketing side, it is easier and quicker to “tour” a virtual building in 3D, and explore it from wherever you happen to be. For operations, digital twins can also demonstrate the back-of-house systems running the building, to help demonstrate potential savings from — or upgrades to — water, heating and waste-management systems, adds Kehr. You’re already using AI Crowley notes that AI is already all around in the real estate industry. Sustainability has been the first step, relying on carbon-emissions reporting tools to improve operations, but she expects computer automation will extend throughout the early planning stages through to times when a property nears its end of use. For institutional investors, the bottom line will be return on investment. While investors realise the relevance of high-quality digital infrastructure and smart technology in their properties, they are still harnessing the data output. With a large portfolio of buildings, “evaluating the performance of each building can be resource intensive,” Crowley points out. “The development of AI will make it much more efficient for them to make a persuasive argument on ROI.” With its full implementation, investors may be able to identify current problems and anticipate future issues within an operating property portfolio, just as building managers can use AI for predictive maintenance. AI has the potential to “readily support property-management roles by streamlining tedious administrative tasks across functions and reducing human error,” adds Crowley. “We’re also predicting significant adoption of generative AI by architects to quickly generate drawings and by construction and city planning to conduct site feasibility studies.” Taronga Ventures invests into emerging tech companies that affect the real estate sector. As such, “you can imagine that the potential impact of AI is dominating many of our discussions with investors in our funds, and within the emerging tech companies in which we invest,” says Hannam. “Yet, AI technology has been a part of the real estate sector for many years, often embedded in the technologies digitising the operations of this sector.” Point vs generative AI The systems already running in the property sector typically involve “point” AI, notes Hannam, serving a specific location. Those uses include facial-recognition software providing security at entry points, enhancing safety at construction sites, and detecting water leaks within a building. Taronga Ventures has invested in companies such as Presien, WINT and OpenSpace serving those functions. What has captured the public imagination in recent months is the possibilities behind “generative” AI, so these uses are normally what springs to mind when we say “AI”. Those models have also existed for some time, says Hannam, but have recently improved and become cheaper to use. That makes them more useful and more accessible. “These new models now far outperform older machine-learning models on most tasks, even simple engagements such as suggesting the next words in an email,” states Hannam. “This is driven by the fact that these are larger models with more parameters and that they are more flexible,” able to be fine-tuned quite easily and switched from one task to another. Those generative uses are likely to have applications for property professionals, aiding in property design and architecture on the creative side, as well as in marketing properties for lease or sale. And it looks like those uses will be both meaningful and here to stay. “My sense is that, unlike crypto and blockchain, which have struggled to gain widespread acceptance, the use of AI is going to be much more impactful,” Hannam posits. He cites companies in a range of fields that Taronga Ventures has reviewed for potential investment: improving property listings by boosting results with machine-learning capability; generating images or copy for marketing purposes based on existing models; lease management by using machine learning to review and “manage” large numbers of leases; mortgage and insurance provision through personalised underwriting and claims management; construction bidding by using AI models to generate costings and even base-level designs; boosting buildings management by making it more efficient; improving energy use and sustainability with predictive office-use modelling; and customer service, by improving the tenant experience, which could potentially lead to higher rents. Healthcare aid Specific property sectors may benefit from purpose-built tech. For instance, the use of AI in hospitals and the health-related aspects of aged care will likely improve care and boost both the health and experience of residents and patients. Taronga Ventures has invested in two companies in that field. Xandar Kardian uses radar technology to detect vibrations from human bodies. This can already deliver the breathing rate and resting heart rate of a person, in real time. The company’s medical device has been approved by the US Federal Drug Administration, and it will be using both AI and machine-learning algorithms to detect patterns not only alerting staff to current health issues but also learning about human conditions to, ultimately, deliver predictive diagnoses that could stave off health problems before symptoms become apparent. The Xandar Kardian sensor can also be used to analyse sleep patterns. Participants must, of course, agree to the monitoring, which is privacy-protected. Calumino has created a smart thermal sensor that can be scaled up to care for large-scale facilities. Its AI analytics study movement patterns that can be used to detect incidents such as falls in real time that is particularly valuable for aged-care facilities. The reason AI is taking hold in the property sector and for other applications is that the user experience has dramatically improved. “In all these cases, the creation of an application interface or dashboard that effectively allows users to engage with these new models and formats has been a key turning point for adoption,” says Hannam. Creating full-service “solutions” that improve the user experience for the staff operating the systems and for the customers or patients “is likely to be a part of the next horizon for the industry,” he concludes. The resistance to overcome What are the challenges or problems to overcome, as AI embeds itself into the property industry? Perhaps the first to note is that the real estate industry can be rather hidebound, and resist change. There’s a temptation to stick with existing methods of operating a business — “if it ain’t broke, don’t fix it”. The most immediate effect from AI may therefore come in more “revolutionary” uses, particularly if they perform a task from start to finish. It will be harder for property companies to adopt AI if it is offering incremental improvements on an existing process or system. What is more, it will take time and effort, not to mention cost, to implement new models or systems in house. “On the flip side, property professionals and businesses need to be open to change,” suggests Hannam. “Professionals, both new and particularly older, will need to be open to experimenting with new tooling to enable their jobs.” Bent seconds the notion that the resistance to overcome will be human, not technological. “I don’t think [the issue] is so much where AI still needs to deliver but whether the real estate industry is willing to collaborate on research and data across large portfolios to come to effective conclusions,” he states. To look at the issue in a positive light, the built environment has both a great responsibility, as an industry that is one of the largest greenhouse-gas emitters, to address the issue, as well as great potential to succeed. “This is only realistic if real estate firms, construction companies and regulatory bodies come together around problem statements and data to help solve for these issues,” continues Bent. “Once that accurate data is available, AI can be very effective in making sense of these gigantic problems, and coming to conclusions around new building materials, effective heating and cooling, energy systems, building facades, and insulation, and so on.” When does AI cross the line? Are the machines taking over the world or, at a more-pragmatic level, taking our jobs? There is plenty of high-level anxiety over how AI will be used and how it will be regulated, as well as concerns about plagiarism and intellectual-property theft, if an AI system is used to draft an architecture plan “in the style of” a certain starchitect. “Data ethics and data privacy are undoubtedly two of the most important concerns,” says Crowley. “Being accountable and transparent about the data collected is crucial here. Further, once a set of data has been processed, the AI is unable to ‘unlearn’ it or not be biased by it, much like humans,” she says. “As such, it is imperative for businesses to develop guidelines and policies that ensure their organisations are using AI responsibly and keeping private internal data out of publicly accessible AI platforms.” Surveillance technology raises a separate set of issues. Video analytics and facial recognition present potential conflict between public safety and privacy rights. Cybersecurity risks necessitate air-tight data systems. Only as good as the data In fact, of course, AI already exists in our pocket with Apple’s virtual assistant Siri, in our workplace, in our car and our home. The means of communication are already being enhanced even in simple ways, such as with predictive word choice and spellcheck. Property professionals will certainly need to adopt AI-augmented tech to improve property management, customer service and building design. “The challenge really is that AI is only as good as the data it is being given,” says Kehr. Equally, the AI system is only as strong as its computer coders and creators, embedding any biases they may not consider. “AI is still in its infancy. And like any infant, that means that it’s really about helping AI learn and understand the environment it is exposed to,” he adds. “Once AI has been taught — and that only works through copious amounts of data and iterative refinement of its algorithm — AI will be able to contribute materially.” The process, he feels, will likely take years. Bent raises the issue of the data “moat”. Big Tech has already demonstrated how valuable relatively simple information tracking consumers can be. The creation of data banks and the ease of access to data “will become more valuable and important,” he says. Then again, many property companies already generate reams of data, in building management on the operations side and in portfolio management on the investment side. This represents the largest opportunity for big developers, asset managers and funds to “make sense of the vast amounts of data that their portfolio is generating, and come to conclusions around what they need to do from a sustainability and climate perspective to future-proof those portfolios,” says Bent. In practical terms, one sliver of the property market that will see large-scale demand due to the use of AI is the data centre specialty segment. Huge amounts of data will need to be sorted and stored across the world, and particularly in decentralised locations, nearer to users of the information. “The emergence of AI goes hand-in-hand with the infrastructure needed to support this intelligence,” states Hannam. Data-capacity demand will spike, and data centres will need to adapt quickly to new uses. “Just as technology such as graphic processors and hardware chips have had to evolve for the AI revolution, so too will the data centre need to be able to support the extent and use cases that emerge from the widespread adoption of AI.” The AI promise There’s no doubt the adoption moves far beyond tinkering with ChatGPT to see if it churns out a few funny answers. AI is not coming for you specifically, or your job, but it is coming, and we in the property world would do well to offer a welcome mat more than a cold shoulder. When asked what changes he anticipates near term, Kehr expects generative AI to take some time to develop, and above all to consume the copious amounts of data necessary to posit what solutions are possible and feasible. Looking further beyond, though, Kehr envisions a world where AI will not only be a co-pilot but may assume much day-to-day management of the planet: addressing issues such as overcrowding, affordability and environmental decay. Kehr also recalls a comment from Microsoft pioneer Bill Gates, “We always overestimate the change that will occur in the next two years, and underestimate the change that will occur in the next 10. Don’t let yourself be lulled into inaction.” It’s a fitting final word to a story written not yet by AI but a breathing human tapping a keyboard. Content is easy to create, but insight is hard to attain. Alex Frew McMillan is a freelance writer based in Hong Kong.

The recent volatility in global stock and bond portfolios has fundamentally reinforced the need for the secondaries market. Elevated public-market valuations in 2021, superseded by a stark correction in 2022, have driven the denominator effect — the relative weight of private holdings now represents an enlarged slice of limited partners’ (LPs) overall portfolio. A perfect storm of rising rates, intensifying inflationary pressures and the precipitous cliff of unrealised value across mature private market fund vintages has exacerbated this effect, amplifying investors’ need for strategic portfolio rebalancing. In Asia, this has triggered a trend of “West selling East”. Amidst such heightened market uncertainty, North American and European investors have been actively seeking near-term liquidity for their noncore Asian private market exposure to pivot back to their home markets. As a result, many of these LPs have turned towards the secondaries market to obtain liquidity for a traditionally illiquid asset class. In addition, a challenging exit environment has also pushed general partners (GPs) to utilise secondaries solutions as an alternative way to generate portfolio liquidity for their investors, mainly in the form of GP-led transactions such as continuation funds, strip sales and tender offers. Despite the evident demand for secondaries liquidity and unprecedented growth in Asia’s private capital markets assets under management, the region’s complex and disintermediated operating environment has proven a formidable barrier to entry for the fresh supply of secondaries liquidity, resulting in a fundamental lack of dedicated secondaries capital in Asia. In consequence, the region’s growing imbalance in demand and supply for secondaries liquidity presents an exceptional opportunity for secondaries investors with strong on-the-ground coverage, dedicated capital and local expertise to benefit from. Superior risk-adjusted returns of secondaries Secondaries have historically delivered superior risk-adjusted returns to investors relative to other private-market strategies. The secondaries market offers buyers the opportunity to access diversified pools of real estate assets later in a fund’s lifecycle at an illiquidity discount, which creates the basis for secondaries to deliver strong returns at reduced risk. Preqin data for private market fund vintages between 2000 to 2020 indicate that secondaries outperformed other asset classes, achieving the highest median internal rate of return of 17.8 percent while maintaining the lowest level of volatility (see chart, page 6). The defensiveness of secondaries is evident in their very low loss rates. According to Preqin, less than 1.5 percent of secondaries funds had a net return of capital below 1.0x, which compares favourably with buyout funds that had a distinctly higher net loss rate of 10.6 percent. This is a testament to the highly diversified nature of secondaries, in terms of assets, strategies, managers and vintages. Furthermore, because secondaries investments generally involve acquiring fund positions at a mature stage of their fund lives, this enables investors to avoid the dreaded J-curve effect experienced in primary investments. These fully prespecified portfolios are often comprised of mature assets with high cashflow and exit visibility, rationalising the underwriting process. Given these funds are usually nearing full liquidation or significantly past their investment periods at the time of a secondaries acquisition, the portfolios should generate a robust cash yield as well as an accelerated return of capital driven by the shorter hold periods. Capitalising on Asia’s disintermediation and market complexity The North American and European secondaries markets have evolved considerably during the past decade to become highly sophisticated and competitive, with typically intermediated processes and large, heavily-standardised transactions where pricing oftentimes serves as the only differentiator. The characteristics of Asia’s secondaries market, however, are fundamentally different. Unlike the more mature markets in the West, the relatively nascent secondaries market in Asia lacks well-established sourcing channels and experiences a significant level of disintermediation. As a result, the majority of secondaries deal flow in Asia is transacted on a proprietary basis not captured by the broader transaction data recorded by brokers, reflecting a tremendous off-market opportunity. From an underwriting perspective, readily available access to information is rare, and therefore unlocking value by leveraging an information advantage becomes a daunting undertaking. Secondaries investors in Asia cannot solely rely on desktop analysis but are required to utilise on-the-ground local coverage and personal relationships with market participants to disentangle the intricacies of the largely unbrokered market to bridge the information gap between buyers and sellers. Given this highly complex and resource-intensive operating environment, most large international secondaries players have yet to build a presence in the region and are instead focused on scaling up the deployment of capital in their domestic markets, where they can benefit from both their local expertise as well as economies of scale. This has resulted in a scarcity of institutional secondaries capital in the region and limited avenues of liquidity for investors in Asian real estate funds. A confluence of disintermediation and market complexity in Asia has resulted in a lack of dedicated institutional secondaries capital relative to more traditional direct strategies. Based on data from Preqin, Asia represented less than 4 percent of the US$265 billion of secondaries capital raised globally during the past five years. This has led to pricing inefficiencies being more profound than global levels, providing further headroom for an increased risk-return profile. Sellers and managers in Asia undertake bilateral negotiations more frequently and require custom-tailored solutions for their complex liquidity needs. Based on research by Greenhill, the average discount to net asset value for global secondaries funds between 2013 to 2022 was 11 percent, while it was considerably higher for Asian secondaries funds, at 20 percent. The opportunity The current timing is unique for the deployment of secondaries capital in Asia. Despite its latent potential, the market remains highly underpenetrated, where numerous sellers with increasing liquidity needs are serviced by a relatively limited pool of institutional buyers. At present, investors’ overallocation to private markets from the denominator effect is a key driver of vast pent-up demand for liquidity solutions. Coupled with the turbulence in the global economy and geopolitical pressures, many institutional investors based in the West have been further compelled to tactically rebalance their portfolios by trimming noncore Asian exposure and pivoting back to more familiar domestic markets. This trend of “West selling East” has aggravated the near-term demand-supply imbalance and serves as a potent catalyst for further liquidity needs with LP portfolios. Meanwhile, a wall of impending real estate fund maturities across Asia is incentivising fund managers to tap the secondaries market for portfolio liquidity. Based on data from Preqin, there is currently US$124 billion of unrealised value outstanding for pre-2013 vintage Asian funds. The disruption created by the global pandemic has further challenged the original exit plans of many Asian GPs, necessitating fund managers to seek alternative sources of exit liquidity. Tapping the secondaries market for such liquidity through GP-led offerings not only satisfies the existing LP base but also allows these GPs to hold winners longer and maximise value creation. Working through this immense cliff of unrealised private market value will fuel the Asian secondaries market for the foreseeable future. Asia’s favourable secondaries market dynamics present a generational opportunity for investors with deep-rooted local relationships, dedicated capital and on-the-ground coverage to capitalise on the urgent, unmet demand for secondaries liquidity in the region. In exchange for solving the illiquidity challenges of Asia’s complex secondaries ecosystem, investors gain access to diversified pools of quality real estate at higher cashflow visibility and lower risk. The positive structural forces that have emerged in the Asian secondaries landscape are likely to persist into the long term, enabling transaction volumes to remain robust and sustaining it as a buyer’s market for institutional secondaries capital for many years to come. Bastian Wolff is a founding partner at Aquilius Investment Partners, an Asia-focused secondaries fund manager, based in Singapore. Shawn Chow is a senior associate in the firm’s investment division.

The current macroeconomic climate is uncertain, but a closer look at the Asia Pacific region reveals numerous opportunities for investors with varying risk appetites and investment horizons. Amidst market risks and the consequences of the denominator effect, now is an opportune time for investors to review and diversify their real assets portfolios strategically. This article explores how investors can benefit from ongoing cyclical and secular trends in Asia Pacific, highlighting potential investment strategies (see chart, page 23) and the region’s appeal in a cross-regional investment approach. As long as interest rates are high and return enhancers are low, investors should take a renewed focus on market fundamentals and assess prospective investments through a clear lens of opportunity:

1. Core/core-plus: Incremental access to prime-grade commercial assets with solid attributes in strategic locations, and asset classes that thrive given current demographic trends and that are well-positioned to deliver stable and risk-adjusted returns.
2. Value-added/tactical: The evolution of workspace strategies has paved the way for lower-grade assets that are ripe for repurposing or upgrade as part of a lease-up strategy; incorporating energy efficiency and sustainability practices presents another layer of opportunity within data centre developments and enhancements.
3. Opportunistic/contrarian: Elevated repricing and refinancing risks are creating more viable opportunities for distressed investments, credit strategies, and potentially attractive entry points for core office and retail assets.