Leveraging data science in real estate

a.s.r. real estate is presenting a new report today that explains the value of data science in real estate investment management. Data models are helpful for decision-making purposes and can enhance our understanding of the complex relationship between macroeconomics, demographics, and the real estate sector. a.s.r. real estate has developed models for appraising the value of agricultural land, as a predictor of the housing market, and for determining the recovery capacity of the European retail market.

Data science models support investment decisions

Predictor of levels of growth on Dutch housing market

The housing forecast tool makes it easier to predict and see the level of growth on the Dutch housing investment market. The model devised by a.s.r. real estate contains an abundance of macroeconomic, financial, and property-related indicators. It uses a combination of publicly accessible and purchased data, which enables a.s.r. real estate to make forecasts about the direction of travel of the market. In the first quarter of 2022, the model predicted four successive quarters of reduced rates of increase in values, which were also visible on the Dutch rental housing market.

Capacity of the European retail market to recover

With the help of a cluster analysis, a.s.r. real estate has identified the capacity of the European retails markets to recover. For example, some markets are seriously affected by a crisis, but recover quickly, as in the case of inner London after the pandemic. Other retail markets take much longer to recover, even though they had initially been less seriously affected. The model showed this to be the case in Milan, for example. The model gives a.s.r. real assets investment partners a clearer picture that enables it to further refine its spread and timing policies with regard to investments.

Agricultural land valuation challenger

The agricultural land valuation challenger is the third model to have been created by a.s.r. real estate. The valuation challenger uses various factors, such as the prices of surrounding plots, physical characteristics like soil quality, and the percentage of peat soil, to gain a clear picture of the value of agricultural land. This gives a.s.r. real estate a better understanding of appraisals and to plan acquisitions and disposals more effectively.

Good collaboration essential

Data science will become increasingly important for real estate companies in the next few years. The combination of data science and economics will make forecasts more reliable, thereby helping policymakers make the right decisions. Geographic allocations are already largely based on data, for example, while investment decisions will be increasingly supported by insights from data science models.

Vinoo Khandekar, head of Research and Intelligence at a.s.r. real estate, says: “Data science offers many opportunities, but its success ultimately depends on effective collaboration between data scientists, domain experts, and asset managers. In recent years, we have invested a great deal in expanding our data science expertise, in improving our data infrastructure, and in the automation of our research tools. This is helping us to enlarge our capacity for analysis, so that we can apply machine learning techniques and predictive analytics accurately. That will lead to better decisions relating to real estate investments and will help us closely monitor and proactively anticipate developments.”