Stefan Tsvetkov, founder of RealtyQuant has some amazing insights into using technology to pick properties and markets.
Finding Undervalued Properties Using AI
Using technology, he is able to screen through large numbers of listings each day from publicly available listings (Loopnet.com Crexi.com, social media) to find the gold - this is data mining at its best! Next, natural language processing (NLP) is used to grab information from the listing to suggest whether there is a value-add opportunity, Merging this data with other datasets helps to identify under-market rents.
Identifying Trendy vs. Stable Markets
Data can be very effective in identifying downside risk. One way to do this is to look for the deviation from the 20 year moving average of home price to income. This method is very accurate!
• 85% Pearson correlation with the actual drops at the state level
• 75% Pearson correlation with the actual price drop at the county level
Through studying the Great Recession, Stefan has identified how far overvalued markets fell to their lowest point, how far markets are currently overvalued, and finally how far markets are likely to fall given a similar economic crisis. He offers this data as a subscription service at his website www.realtyquant.com.
Finding Data
There is always more data available for residential than for commercial real estate. The intelligent commercial real estate investor must be willing to take the time to understand the residential market!
Stefan's Advice to His Younger Self
Get involved in private markets sooner!
#investing #datascience #realestate
