AI that Actually Works for Real Estate
At a glance
- Conventional AI, especially current machine learning algorithms, are failing to meet the demands of real estate investors. They can’t function with the small datasets that are ubiquitous in real estate, or leverage the deep domain expertise of investors.
- Causal AI is the only enterprise AI that works for real estate investment. Models built with Causal AI seamlessly integrate domain knowledge and produce results with small data.
- Pre-eminent real estate investors leverage Causal AI today to solve their most pressing and demanding challenges.
2021: what is top of mind for our clients
Macroeconomic and socioeconomic factors are shifting, leading to high uncertainty in global real estate markets. How will diverse inflationary and recovery scenarios impact real estate? What’s the effect of increased online shopping and flexible working?
Opportunities are emerging in the aftermath of the crisis. Which real estate sectors and cities are on the rise, and which will experience outflows? At a more opportunistic level, what are the drivers of premium for specific properties?
With ESG at the top of the agenda for investors, how can climate risk and sustainability be measured and anticipated? These are some of the challenging questions that are currently top of mind for our real estate investment clients.
Causal AI versus conventional AI in real estate investment
Real estate investors have been slow to adopt AI relative to other asset managers, and understandably so.
Real estate data is characteristically small; however standard AI algorithms need big datasets to operate. These standard AI algorithms also lack explainability and interactivity, and so they can’t harness the deep domain expertise of real estate investors. Businesses are finding that current algorithms are largely failing to deliver meaningful financial returns.
“Transparency and explainability of AI models requires an understanding of causality — an inherent advantage of the causaLens platform.” ~ Wendy Harrington, Head of Nuveen Labs
However, Causal AI has the potential to transform real estate investment. It can help investors see past their biases, integrate alternative datasets with their models, zero in on important factors that human intuition misses, and dynamically optimize their portfolios.
Causal AI is the only enterprise AI that performs in small data environments. It is transparent, provides insightful explanations, and is easy for domain experts to share knowledge with.
We highlight three use cases — among many potential applications — that illustrate how Causal AI is transforming the real estate investment landscape today.
We deployed Causal AI to find the macroeconomic and socioeconomic indicators that are causally driving city-level revenue growth. Our automated modelling pipeline is able to eliminate noisy, misleading correlations even with sparse, quarterly data. This allows us to build highly accurate causal models of revenue growth for many cities, that make better predictions ahead of other market participants.
Figure 1. Weighted causal drivers of multi-family residential rental growth in San Francisco. Drivers that positively impact rent growth are shaded blue, negative drivers are red. Causal discovery algorithms autonomously identify high-tech job growth as a driver of prices. This research was based on client work for one of the top-five largest global real estate investors, and became an integral part of the client’s fundraising efforts.
Causal AI adapts to disruptions and recent trends — such as the decline of offices and rise of flexible working — for which relevant historical data is in short supply. Our platform can build models of natural vacancy rates and prime rent outlook in different cities. These models are automatically integrated with ongoing macroeconomic and demographic trends that may impact the future of offices.
Figure 2. The causaLens dashboard shows the results of our forecasts, making them directly accessible to business stakeholders with tailored information. Here we show prime rent outlook forecasts for London. Forecasts can be used to inform our state-of-the-art portfolio optimizer, specifically designed for real estate markets.
Causal models are uniquely able to simulate how real estate will be impacted by big changes to the market. This includes planning for major macroeconomic shifts, such as rising inflation.
Analysts can use our platform to answer “What If?” questions, investigating how real estate sectors and geographies will perform as an inflation hedge. They can also seamlessly share their expertise with the AI.
Figure 3. The causaLens dashboard houses an “understanding” environment, which enables business stakeholders to ask and answer “What If” hypothetical questions. This is a signature capability of Causal AI, which requires a causal model of the market. Here we illustrate the results of a “What If” analysis of prime rent in central London.
“Causal AI empowers our strategists and portfolio managers to generate alpha by identifying new causal relationships in economic, financial and alternative data, with sophisticated, adaptive and explainable models that don’t suffer from overfitting.” ~ Michael Grady, Head of Investment Strategy and Chief Economist, Aviva Investors
The world’s leading real estate investors are already gaining a competitive edge with Causal AI. Find out how we can accelerate your AI journey.