Can AI boost liquidity in real estate markets?

A significant difficulty for investors in real estate is the illiquidity of property markets relative to other asset types. In times of economic downturn and the unforeseen global events that often drive them, pricing uncertainty can be significant and it can take a long time for markets to re-calibrate. In any market there will always be a period of misalignment between buyer’s and seller’s intentions, but this is typically resolved more quickly within financial markets, due to the higher frequency of trading and fast, transparent reporting. In contrast, real estate assets are typically traded less often, and some even off-market, resulting in a smaller pool of transactional evidence to assess pricing.

Moreover, as well as the economic outlook there are a range of factors which influence investor choice. Other considerations include; use type, purpose, location, structure and the general appeal of the building. With both logical and emotional reasons at play, it can be hard in some cases to pinpoint the reasons for some prices paid, or apply the same numbers to a ‘comparable’ building.

Still, improved data quality and visibility, and capital markets research have helped transparency in the real estate industry.

Despite the illiquidity in property markets, the industry holds significant datasets on historic transactions and activity, and real estate advisors are privy to proprietary data and local market knowledge. This makes the investment world ripe for AI intervention, with opportunities for evidence-based decision making and a platform for AI to improve on existing data capabilities. There can be gaps and inconsistencies in the investment data more so than in other tradable sectors. However, as highlighted in another of our articles, AI has the capability to combine and interpret large datasets effectively. This could enable us to better harness the power of data on a large scale and ensure its optimum completeness – through AI’s ability to identify and fill gaps, where data is available, much quicker than a human could.

The capability to bring together and analyse big data offers a significant advantage to those operating within real estate, particularly organisations of scale with access to large quantities of real-time data. Transactional investment teams can greatly benefit from using this emerging diagnostic capability to help their clients to evaluate market opportunities. At CBRE we are leveraging this ability through our tool, Capital AI, which transforms the potential of our robust proprietary dataset. Each metric – including transactional evidence, investor profiles, bid intel, pricing, and more – offers significant insights independently. However, the ability of AI to combine and analyse this data as a collective amplifies its impact and enhances the accessibility of information and speed to market.

Brokers, acting on behalf of buyers and sellers, can even submit strategy information relating to risk appetite, sector types, or locational preferences into the tool to derive insights such as location and asset scoring. Additionally, they can use the technology to rank potential buyers through analysis of recent bids. For optimal results, the analysis would go further to consider how geographies, markets, and sectors will change over time, and what the future demand for a particular asset might be. This can be done through combining real estate insights alongside economic and demographic data; past and forecast.

Through identifying new or untapped sources of capital and buyers adapting their strategies, this technology in use alongside CBRE’s proprietary data could expand bidder pools by up to 20%. The value of a tool like this is apparent already, but it will grow exponentially as the data set scales. The more information that is fed into it, the more it can learn and enhance its capabilities for future interpretations of market dynamics.

The entire process is much faster than it would be for a human; particularly useful for answering multi-faceted questions which wouldn’t have been feasible previously. The technology fundamentally expands what is possible and should aid more precise investment decisions. Importantly it also maximises the utilisation of real-time data, instead of relying on traditional set reporting periods; potentially leading to a better-informed market. The accuracy of analysis and datasets supported by AI in other ways will contribute to this (for example, we expect AI to upgrade some aspects of real estate forecasting). While trading in real estate may always take more time compared to other asset types, a lower level of uncertainty could encourage deal flow and liquidity.

There are still concerns around AI with regards to biased training data, cybersecurity threats, and hallucinations, so precautions should be in place in any instance where it is used. Advisors, investors, and sellers using this type of technology need to be aware of the potential risks and of the regulatory landscape around AI which is still developing. As with all current use cases for AI, verification by humans at every stage is fundamental, and this is likely to continue as a best practice approach to ensure the most precise outcomes and holistic service provision. But conceptually AI is demonstrating significant advantages in terms of speed, efficiency, and broader capabilities for professionals in real estate investment. It has the potential to bolster real-time data, thereby supporting more informed investment decisions.