/rent-or-buy

Propose optimal business strategy for a volatile real-estate market. Used machine learning algorithms to forecast listing and rental prices in local markets based on openly available market data

Primary LanguageJupyter NotebookGNU General Public License v3.0GPL-3.0

  • Rent or Buy

Here I use regression analysis on real-estate market to determine the optimal time of buying a house in a case of coming recession. As a historical data on house prices I use a data set from Zillow, which provides monthly median listing prices and rental rates for units of different sizes and locations in the period from 2011 till 2018. However, this period does not include the period of recession when house prices where falling rapidly.

To capture the period of the house bubble bust, I use some historical government data: house price index (HPI) and Fair monthly rate (FMR). The HPI represents an aggregated index house prices in a regions, while the FMR provides an aggregated data for rental rates across cities and counties.

The idea of my approach, is to use the government indexes as predictors for listing prices and rental rates in a particular local market. By a local market I consider houses having a particular number of bedrooms and located in an urban, suburb or rural population density area. As an example, I provide analysis for Eastern Massachusetts, where I currently reside.

The central result of this work is an interactive dashboard in which one can bet an estimate of the house market drop, select which type of house one is currently renting and which type one wants to buy. The app calculates an estimated cost the purchase, considering the future paid rent and the house listing price.

dashboard

To run an interactive notebook click below:

Binder

DISCLAIMER:

THIS PROGRAM IS NOT DESIGNED TO GIVE ANY FINANCIAL ADVICE. TO INVEST ON THE REAL-ESTATE MARKET CONTACT A QUALIFIED REALTOR. I AM NOT RESPONSIBLE FOR ANY FINANCIAL LOSS OR LOSS OF PROFIT SHOULD YOU DECIDE TO USE THIS PROGRAM.

rent-or-buy Copyright (C) 2019  Ivan Lisenkov
This program comes with ABSOLUTELY NO WARRANTY;
This is free software, and you are welcome to redistribute it
under certain conditions;