How can AI bridge the gap for real estate investment managers?

Traditional artificial intelligence (AI), incorporating expert systems, natural language processing (NLP) and decision trees, is proving to be a helpful tool for use in business, having been established already in some industries. Meanwhile, generative AI creates something new; in this case, most industries globally are still exploring its capabilities and applications.

This article investigates the extent to which investment managers trading different asset classes can take advantage of both traditional and generative forms of AI. There are clear differences between the finance and real estate sectors, but there could be potential for AI implementation to bridge the gap. 

Use of AI in financial investment management 

Because of high data transparency and market liquidity, traditional AI has been a resource in the financial sector for many years. Investment managers and investors have used it alongside other technologies to provide quick and meaningful support to clients looking to invest in financial assets such as equities and bonds.

Investment managers have also been using technological tools and algorithms through machine learning to automate the trading of assets. Financial assets can be transacted instantly under pre-determined conditions, therefore reducing latency, minimising human error, and enabling the implementation of arbitrage and hedging strategies. While also having the potential to lower transaction fees. 

Additionally, there are a number of traditional AI tools that the finance sector uses already, one of which is Robo-Advisors. A Robo-Advisor is a digital wealth management tool used as an alternative to financial advice that can analyse financial data and provide bespoke advice to clients based on investment objectives and risk tolerance. Robo-Advisor is an example of an expert system, that is able to perform human-like tasks through their use of AI. Expert systems are computer programmes designed to mimic the expertise and judgement of humans in a particular field. 

The advantages of Robo-Advisors are that emotional biases often associated with financial advice can be mitigated; fees for clients can be reduced; data driven insights and advice can be provided to large numbers of clients; and financial advice can be made more accessible, particularly to people with lower financial literacy aiding customer relationship management. 

Another tool is natural language processing (NLP) which is used to analyse and interpret human language in spoken or written form. Siri is one example of an NLP system, as it listens to voice commands and analyses them using NLP. NLP can be used in financial analysis to identify underlying sentiments from company and industry reports or announcements, in order to identify trends provide insights for future direction.

AI in commercial real estate

Implementing AI could improve efficiencies for the commercial real estate industry across a range of functions and real estate investment management could be included. But so far, the utilisation of AI within commercial real estate investment has been slow. Why hasn’t traditional AI technology been implemented here like it has when investing across other asset classes?

Unlike equities and bonds, the real estate asset class is indivisible, and every asset differs with individual features, meaning the asset class cannot be homogenised. Additionally, real estate assets are often larger, less liquid, require significant due diligence and, unlike equities and bonds, lack a public marketplace and therefore, transaction costs can be high. 

These characteristics coupled with quality data scarcity and a lack of pricing transparency (in a less liquid market), particularly in private equity and private debt investment in real estate, are the main limitations to using the trading-related technologies that are implemented in finance. Performance data often isn’t freely available and can be expensive to access from third-party data providers. Market information can also be segmented across regions, whereas stock and bond price data can be accessed globally. These factors make it impractical to develop pre-set conditions that execute transactions like algorithmic trading tools do. 

Similar to their use in the finance sector, Robo-Advisors could provide investment advice on market trends and performance attributes across real estate sectors and geographies, provided performance data is available. But an obvious contrast between the two sectors is the availability of data, given that AI models can only be as powerful as the underlying data which feeds into them. Moreover, commercial real estate often requires high amounts of capital to access, and therefore it is difficult for investors whom Robo-Advisors are targeted towards to execute direct real estate investment strategies.

Although the commercial real estate sector does have vast datasets spanning decades, the unstructured nature, nuances and limitations to the data have likely contributed to a lag in technological adoption compared with companies focused on other asset classes.

How can generative AI help real estate investment managers?

Generative AI could provide a new opportunity for real estate investment managers to bridge the technology gap that currently exists between them and those investing in traditional asset classes such as equities and bonds.

One of the key issues faced by real estate companies is the unstructured labyrinth of real estate data. Often, within larger real estate firms, each team will collect, maintain and store their own data in different files and formats. Generative AI has the capability to collate and interrogate large vacuums of unstructured data which could help real estate professionals and the industry to use it for broader use and with greater impact. Generative AI can also search and process documents such as leases for crucial datapoints, values or dates which could improve, or at least maximise the use of the data available for businesses, without requiring time-consuming manual data collection. Better data visibility and quality enabled by these methods should help to optimise and speed up positioning analyses for real estate portfolios. 

AI applications can also enhance performance reporting. Firstly, real estate investment managers can use generative AI to speed up the often time-intensive production of regular performance reports for clients’ portfolios. And in conjunction with platforms like Tableau, it can quickly visualise data analysis through dashboards that are crucial for client or stakeholder interactions to make informed decisions.

Finally, generative AI platforms have the ability to instantaneously use and interpret collated performance data analytics to calculate strategic portfolio allocations to real estate sectors and geographies. With this tool, investment managers could quickly generate bespoke portfolio allocations for clients based on risk and return objectives. This capability of Generative AI mimics the function that Robo-Advisors can perform for retail investors, but for institutional investors on a larger scale given data availability. Additionally, a private investor unable to access private commercial real estate investments, could use AI to calculate portfolio models on publicly available REIT price data and create an indirect real estate portfolio given their risk and return objectives, without having to pay fees and use a Robo-Advisor. 

To summarise, generative AI has the capability to address key challenges faced by real estate investment managers and provide technological solutions in investment practices. Improving data quality and speeding up the creation of bespoke client solutions would lead to enhanced transparency and knowledge of commercial real estate as an asset class. This could encourage investors of traditional assets to actively consider the sector when formulating their portfolios.