[2Win5] Poster session 2

[2Win5-89] Home improvement stores sales forecast using SAR satellite images

〇Yoshihiko Ichikawa¹, Keiichi Suga¹, Kaito Takano², Kei Nakagawa^2,3 (1.Insight Edge, Inc., 2.Nomura Asset Management Co., Ltd., 3.Osaka Metropolitan University)

Keywords:Alternative Data, Satellite Imagery, Sales forecast, Google Earth Engine, SAR

We examine a method to estimate the sales of home improvement stores using synthetic aperture radar~(SAR) satellite images. Sales of home improvement stores and other retail businesses are important indicators for evaluating corporate performance and consumer demand, especially for investors and analysts.
Real-time information is required for financial data such as sales, which is directly linked to profit forecasts and investment decisions, but the data is not publicly available and difficult to collect.
Therefore in this study we hypothesize that the number of cars in a home improvement store's parking lot is proportional to sales, and uses the high resolution and all-weather characteristics of SAR satellite images to verify this hypothesis.
To verify the hypothesis, we analyzed the parking lot on the premises of Joyful Honda Co., Ltd. as a case study and estimated sales. The results suggest that sales can be estimated by using SAR satellite images.