Contact

Menu

Article

How the AI revolution boosts real estate asset management performance

Data science, the study of data using statistics, artificial intelligence, machine learning and computer engineering to analyze large amounts of data, is a tool used in Global Real Estate for extracting meaningful insights to help make business decisions and seize investment opportunities. Data science is often used for predictive and prescriptive data analysis.

November 2, 2023

Gerald Kremer

Chief digital officer, Global Real Estate, Credit Suisse Asset Management (Switzerland) Ltd., and designated crew product lead real estate, UBS Asset Management

Key takeaways

Predictive analysis uses historical data to make accurate forecasts about data patterns that may occur in the future. Using a combination of historical, external and alternative data with machine learning, forecasting and predictive modeling, it is possible to identify causality connections in the data.

In this issue, we present a number of data science applications used to support our real estate operations, which we leverage to make strategic investment decision with regard to our assets and funds.

We will then explain how the use of the emerging LLMs further boosts the added value provided by data science applications.

Data Science use cases in CSAM GRE

Spatial Real Estate Risk

Credit Suisse's Spatial Real Estate Risk report uses a collection of 11 million property advertisements in Switzerland containing sales and rental offers from the past 12 years, augmented with additional external and internal data sets, to predict the rental yield spatially.

Low-yield/high-yield property density
To estimate if there is a property bubble building in the market, the density of underperforming or low-yielding properties is used as an indicator.
Source: Credit Suisse AG. For illustrative purposes only

Commercial Tenant Churn

Credit Suisse's Commercial Tenant Churn report uses historical commercial lease contracts to derive a labeled dataset. This report is supplemented with regional information (e.g., bankruptcies, new competitors) to estimate the likelihood that a tenant will terminate their contract within 6 to 12 months, thus enabling asset managers to anticipate vacancies and plan the optimal use of space well in advance.

Market AI
Data from advertisements over the last seven years is updated and used to build a machine learning model that predicts the rent and gives indications as to which attributes have contributed to increasing (or decreasing) rents.
Source: Credit Suisse AG. For illustrative purposes only

Mass Market Rent Prediction

Based on advertisements, a predictive model is used to anticipate what the market is willing to pay for the units in our portfolio. The amount that each attribute contributes to the rent (or by which the rent is reduced) is provided for each unit. This allows asset managers to optimize the performance of our properties, achieving the best balance between high rental yields and low vacancy rates.

Vacancy risk
Commercial tenants are moving around based on changing market conditions. A model that learns how historical contracts have reacted to changing conditions in the market, such as delinquencies in the same sector or new space requirements, is used to anticipate changes and gives recommendations about what to adjust to help the tenant.
Source: Credit Suisse AG. For illustrative purposes only

This use of data science in our daily business enables more accurate investment forecasts and helps us to mitigate risks and make better decisions. It can identify new opportunities and enables us to be more proactive.

Innovations in artificial intelligence

As mentioned at the beginning of this article, there have been some recent exciting developments in artificial intelligence and machine learning. The use of LLMs, a type of artificial intelligence algorithm that applies deep-learning techniques and extremely large data sets to understand, summarize, generate and predict new content, has been made accessible to everyone. LLMs are excellent at classifying and categorizing data, performing sentiment analysis (understanding the intent of a piece of content) and implementing conversational AI or chatbots (enabling users to interact with the AI in a more natural conversant way). 

Concrete ways in which we can apply LLMs to our existing reports include:

Spatial Real Estate Risk

By adding the ability to understand text from advertisements and news articles, a sentiment analysis could front-run market movements that the current model with its historical view couldn’t do. Additionally, an LLM can also directly query the results, provide a market analysis for specific segments and prepare a report.

Commercial Tenant Churn

Connecting the model-driven insights with an LLM that directly prepares communication with the tenants to alleviate the issues the business faces will improve our relationship with our tenants. Adding previous communications will also give our client advisers the ability to communicate extremely accurately with tenants, which in turn will provide tenants with the assurance that we understand them and have their backs.

Mass Market Rent Prediction

Current models use quantitative attributes and only a limited number of qualitative assessments of the properties. With the help of an LLM, this gap can be filled and a more precise model built by adding the information contained in the description of advertisements.

These recent innovations will enable us to be even more efficient, accurate and innovative in our business decisions, strategic planning and investment opportunities.

Investment possibilities

Find investment products that suit your personal needs. Choose from our extensive range of investment solutions across all major asset classes, and access all product-related information.

Get in touch with Asset Management

Contact us to learn about exciting investment opportunities. We are here to help you achieve your investment goals.

Source: Credit Suisse, unless otherwise specified.
Unless noted otherwise, all illustrations in this document were produced by Credit Suisse AG and/or its affiliates with the greatest of care and to the best of its knowledge and belief.

This material constitutes marketing material of Credit Suisse AG and/or its affiliates (hereafter "CS"). This material does not constitute or form part of an offer or invitation to issue or sell, or of a solicitation of an offer to subscribe or buy, any securities or other financial instruments, or enter into any other financial transaction, nor does it constitute an inducement or incitement to participate in any product, offering or investment. This marketing material is not a contractually binding document or an information document required by any legislative provision. Nothing in this material constitutes investment research or investment advice and may not be relied upon. It is not tailored to your individual circumstances, or otherwise constitutes a personal recommendation, and is not sufficient to take an investment decision. The information and views expressed herein are those of CS at the time of writing and are subject to change at any time without notice. They are derived from sources believed to be reliable. CS provides no guarantee with regard to the content and completeness of the information and where legally possible does not accept any liability for losses that might arise from making use of the information. If nothing is indicated to the contrary, all figures are unaudited. The information provided herein is for the exclusive use of the recipient. The information provided in this material may change after the date of this material without notice and CS has no obligation to update the information. This material may contain information that is licensed and/or protected under intellectual property rights of the licensors and property right holders. Nothing in this material shall be construed to impose any liability on the licensors or property right holders. Unauthorised copying of the information of the licensors or property right holders is strictly prohibited. This material may not be forwarded or distributed to any other person and may not be reproduced. Any forwarding, distribution or reproduction is unauthorized and may result in a violation of the U.S. Securities Act of 1933, as amended (the “Securities Act”). In addition, there may be conflicts of interest with regard to the investment. In connection with the provision of services, Credit Suisse AG and/or its affiliates may pay third parties or receive from third parties, as part of their fee or otherwise, a one-time or recurring fee (e.g., issuing commissions, placement commissions or trailer fees). Prospective investors should independently and carefully assess (with their tax, legal and financial advisers) the specific risks described in available materials, and applicable legal, regulatory, credit, tax and accounting consequences prior to making any investment decision. 
Copyright © 2023 CREDIT SUISSE. All rights reserved.

Distributor: Credit Suisse (Switzerland) Ltd., Paradeplatz 8, CH-8001 Zurich I Distributor: Credit Suisse AG, Paradeplatz 8, CH-8001 Zurich I Distributor: Credit Suisse Asset Management (Switzerland) Ltd., Kalandergasse 4, CH-8045 Zurich I Supervisor (Entity of Registration): Swiss Financial Market Supervisory Authority (FINMA)