This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

« Back to Results

Machine Learning

Paper Session

Sunday, Jan. 6, 2019 10:15 AM - 12:15 PM

Hilton Atlanta, 217
Hosted By: American Real Estate and Urban Economics Association
  • Chair: Yildiray Yildirim, Baruch College

The Information Value of Property Description

Yannan (Lily) Shen
,
Clemson University
Yiqiang Han
,
Clemson University

Abstract

This paper employs a ML-Hedonic approach to quantify the value of uniqueness, a type of soft" information embedded in real estate advertisements. We rst propose an unsupervised learning algorithm to quantify levels of semantic deviation ("uniqueness") in descriptions, the textual portions of real estate advertisements. We then estimated the impact of description uniqueness on real estate transaction outcomes using linear hedonic pricing models. The results indicate textual data disseminate information that numerical data cannot capture, and property descriptions effectively narrow the information gap between structured real estate data and the houses by conveying "soft" information about unique house features. A one standard deviation (0.08) increase in description uniqueness compared to neighboring properties leads to a 5.6% increase in property sale prices and a 2.3{day delay in the closing time, controlling for house characteristics, transaction circumstances, and agent unobservables. This paper provides theoretical and empirical insights on how to utilize the emerging Machine Learning
tools in economic research.

The Value of Curb Appeal: A Machine Learning Approach

Erik Johnson
,
University of Alabama
Sriram Villupuram
,
University of Texas-Arlington

Abstract

We develop a methodology to estimate/quantify the value of curb appeal using machine learning and data from a large metropolitan area in the United States. The most important contribution of the study is the identification of a tool that can be cost-effectively scaled across geographies to evaluate curb appeal of individual properties. We use Google StreetView to extract images of the front yard and façade of homes sold in the Denver Metropolitan Statistical Area and rank them from poor to excellent curb appeal using a tensor flow algorithm. We estimate that on average, the curb appeal premium can be a minimum of 7% of a home’s sale price. The premium can be as high as 10% for homes sold during a “cold” market. In addition, we find that curb appeal increases the liquidity of a home more so during illiquid periods such as a housing downturn when there is a larger than average supply of homes for sale.We sort neighborhoods by average curb appeal and find that homes that are in poor curb appeal neighborhoods lack the summer seasonality premium. In addition, poor curb appeal of the subject property as well as that of neighbors seems to have a limited to no impact on home values in those neighborhoods. This study has important implications for academics in that this measure of curb appeal can be employed in future hedonic models of house price estimation. This study also will have important implications for practitioners who use computer models to estimate or appraise home values.

Machine Learning, Building Vintage and Property Values

Thies Lindenthal
,
University of Cambridge
Erik Johnson
,
University of Alabama

Abstract

"This paper introduces an algorithm that collects pictures of individual buildings from Google StreetView, trains a deep convolutional neural network (CNN) to classify residential buildings into architectural styles
(vintages) and estimates the impact of vintage on sales prices for the universe of residential housing transactions for the city of Cambridge, UK, between January 1995 and June 2017. The contributions of the paper are to 1) introduce a general algorithm for capturing building specific images; 2) illustrate the efficacy of transfer learning using Google’s Inception-v3 model to classify images; and 3) provide a basic hedonic estimate of the impact of architectural style on housing sales prices. Preliminary estimates indicate a price premium of 11.6% for Georgian, 8% for Early Victorian, 22% for Late Victorian/Edwardian, 6.8% for Interwar, 6.1% for Contemporary, and 7.8% for Faux Victorian relative to the Postwar architectural style. We are currently extending the data framework to enable data collection and analysis for the entire UK."

Is Innovation Really in a Place? Accelerator Program Impacts on Firm Performance

Andrea Chegut
,
Massachusetts Institute of Technology
Schery Bokhari
,
Massachusetts Institute of Technology
Dennis Frenchman
,
Massachusetts Institute of Technology
Isabel Tausendschoen
,
University of Graz

Abstract

We investigate the impact of a nascent entrepreneurial amenity for urban agglomeration, accelerator programs, upon start-up firm's private equity performance. Accelerators are firm development programs that utilize physical space, human capital development programming, mentorship, financial capital, and community engagement to accelerate the financial feasibility of start-up firms. A sample of US accelerator treated and matched control firm's over the 2005 to 2015 period yields a study of 16,720 firms. Results indicate that there is statistically significantly more cumulative funding for accelerated firms, when taking into consideration the endogenous choice and selection of start-up firms into programs and series stage in cumulative funding. Secondly, we assess variation across accelerator participation timing and find that firms with pre-funding when coming into an accelerator leads to higher cumulative funding. Lastly, we document accelerator program's ability to cultivate agglomeration through space and programming amenities like free physical space, program length, program cohort size, investor equity stake and scale of capital injection impacts upon cumulative funding. This study supports evidence of correlation between start-up firm performance and accelerator program amenities. Accelerators can have significant impact on the life-long health of young private-equity firms.
Discussant(s)
Gianluca Marcato
,
University of Reading
You Suk Kim
,
Federal Reserve Board
Pavel Krivenko
,
Stanford University
Jiro Yoshida
,
Pennsylvania State University
JEL Classifications
  • R3 - Real Estate Markets, Spatial Production Analysis, and Firm Location
  • R2 - Household Analysis