Capital Bay – Revolution of Real Estate Valuation

This project was for summer term 2020, you CAN NOT apply to this project anymore!

Results of this project are explained in detail in the final documentation and presentation.

Capital Bay is a vertically fully integrated alternative investment manager in the field of real estate. The real estate industry contributes about 20% to the German GDP – the automobile industry 4%. Further, in 2018, properties and buildings constituted 75% (EUR 15.8 trillion) of the value of all tangible assets in Germany, almost 5 times of the German GDP. However, valuation of these assets is mainly based on simplified DCF models, simply accruing and discounting rents, costs of operation and exit prices.


As the insurance industry designed models for the prediction of the occurrence and size of hailstones, we intend to create a scientific calculation machine that precisely predicts future cashflows from real estate assets based on the development of micro- and macroeconomic parameters. Only with these accurate predictions we will be able to exactly determine the value of assets. The necessary data will be provided by 21st REAL ESTATE a strategic partner of CAPITAL BAY. 21st REAL ESTATE is the market leader for real time real estate data. Its database is a complex framework of integrated data tiles that covers the entire area of Germany. From over 55 million of complex interacting analysis tiles it generates 1.1 million output tiles of 200m x 200m within cities. 21st REAL ESTATE uses the most up-to-date spatio-econometric models and machine learning algorithms to derive comparables for different usage types of real estate. These comparables, paired with an exhaustive amount of tile-specific smart data, highlighting the environmental conditions of the site, sets the foundation for each investment CapitalBAY is pursuing.

You will determine critical factors influencing asset value-driving parameters (like rent, price or CapEx etc.), create models predicting the development of such parameters during an investment period and potentially built a scenario simulator which eventually feeds the cash flow plans for real estate valuation. Such models could consider micro-, socio- and macro-economic factors as explaining variables. The tool- kit for the modelling will range from machine learning techniques like regression tasks, over time series analysis to simulations.