Real estate media sentiment through textual analysis

Pages410-428
DOIhttps://doi.org/10.1108/JPIF-07-2017-0050
Published date09 July 2018
Date09 July 2018
AuthorJessica Roxanne Ruscheinsky,Marcel Lang,Wolfgang Schäfers
Subject MatterProperty valuation & finance,Property management & built environment
Real estate media sentiment
through textual analysis
Jessica Roxanne Ruscheinsky, Marcel Lang and Wolfgang Schäfers
International Real Estate Business School, University of Regensburg,
Regensburg, Germany
Abstract
Purpose The purpose of this paper is to determine systematically the broader relationship between news
media sentiment, extracted through textual analysis of articles published by leading US newspapers, and the
securitized real estate market.
Design/methodology/approach The methodology is divided into two stages. First, roughly 125,000 US
newspaper article headlines from Bloomberg,The Financial Times,Forbes and The Wall Street Journal are
investigated with a dictionary-based approach, and different measures of sentiment are created. Second, a
vector autoregressive framework is used to analyse the relationship between media-expressed sentiment and
REIT market movements over the period 20052015.
Findings The empirical results provide significant evidence for a leading relationship between media
sentiment and future REIT market movements. Furthermore, applying the dictionary-based approach for
textual analysis, the results exhibit that a domain-specific dictionary is superior to a general dictionary.
In addition, better results are achieved by a sentiment measure incorporating both positive and negative
sentiment, rather than just one polarity.
Practical implications In connection with fundamentals of the REIT market, these findings can be
utilised to further i mprove the understanding of securi tized real estate market movement s and investment
decisions. Furtherm ore, this paper highlig hts the importance of payi ng attention to new media an d
digitalization. The r esults are robust for differ ent REIT sectors and when conventional control va riables
are considered.
Originality/value This paper demonstrates for the first time, that textual analysis is able to capture media
sentiment from news relevant to the US securitized real estate market. Furthermore, the broad collection of
newspaper articles from four different sources is unique.
Keywords USA, REITs, Sentiment, Textual analysis, Dictionary-based approach, News analytics
Paper type Research paper
1. Introduction
A simple remark from him could cause the stock market and the dollar to rise or fall,
commented Abe (2011), who analysed the changes in Alan Greenspanslanguageuseduring
his period as chairman of the Federal Reserve Board. The message behind this concise
propositionis one sound reason for intensifiedresearch efforts assessing how decision-making
is often not based solely on fundamentals.
A substantial body of literature focusses predominantly on quantifying the effects of
sentiment captured through the textual analysis of stock market-related text corpora. The
most important works on text-based sentiment analysis include Tetlock (2007), Das and
Chen (2007), Tetlock et al. (2008) and Loughran and McDonald (2011), who found significant
correlations with stock returns, return volatility and trading volume. However, there is little
research investigating the role of text-based sentiment in a real estate context and in
particular none in relation to the securitized real estate market. Understanding the
behaviour of Real Estate Investment Trust (REIT) price movements, using text-based
Journal of Property Investment &
Finance
Vol. 36 No. 5, 2018
pp. 410-428
© Emerald PublishingLimited
1463-578X
DOI 10.1108/JPIF-07-2017-0050
Received 14 July 2017
Revised 15 February 2018
Accepted 16 February 2018
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1463-578X.htm
The authors would like to thank Matthias Himmelstoss for his technical support collecting the data.
The valuable feedback of the two anonymous referees is highly appreciated. Furthermore, the authors
would like to thank ERES and ARES conference participants for their valuable feedback at our
conference presentations. The authors also highly appreciate the Best Paper Award in the PhD Session
at European Real Estate Society 2016 for this research project.
410
JPIF
36,5
sentiment measures, is especially relevant for two main reasons. First, as an asset class,
REITs are information-intensive. This derives from the aspect that both stock and real
estate market characteristics must be taken into account due to the underlying asset class
on the one hand and the stock exchange listing of REITs on the other hand. Second, real
estate market information is mainly backward looking and lacks expectations about
future market conditions, for example, the NCREIF property index or real capital
analytics transactions.
Until recently, no attention has been paid to the extraction of sentiment through the
textual analysis of online text corpora related to the REIT market. Especially interesting
and promising is the investigation of online newspaper articles as a newly available source.
Hence, this paper aims at filling this research gap by analysing newspaper article headlines
from leading US financial newspapers to evaluate the question of whether news media
sentiment influences future securitized real estate market movements. The use of news
analytics is defined by Das (2014) as a special subfield of textual analysis, which is
associated with distinct advantages, in comparison to the traditional survey-based
sentiment measures. Not only the immediate availability and objectivity of results is a key
aspect, but also the option of scaling the methodology to a large data set and a wide variety
of topics. Concerning the text corpus, news headlines offer several advantages compared to
Twitter messages, blog posts or forum entries that have been explored in previous studies.
News headlines are written more professionally and therefore contain (almost) no
typographical errors, normally no slang or abbreviations, and extraction can be limited to a
specific language. Additionally, with respect to news, it is more likely that published
information is reliable and read by a broad and, equally important, a relevant audience.
The newspaper sample consists of about 125,000 market-specific US news article
headlines from Bloomberg,The Financial Times,Forbes and The Wall Street Journal. These
newspaper headlines are analysed by applying the dictionary-based approach. The
adequacy of a general psychological dictionary is compared to a domain-specific dictionary.
Subsequently, different sentiment measures are derived and tested in a vector
autoregressive model (VAR) on their linkage to the REIT market.
The empirical results suggest a significant relationship between media-expressed
sentiment and REIT returns. The findings are robust when conventional control variables
are considered. Specifically, a leading relationship of the created real estate media sentiment
by three to four months is identified. Moreover, the development of a domain-specific real
estate dictionary, leads to a superior fit of the model. The findings are relevant to various
market participants, for example, for investorsdecision-making processes, as media
sentiment is forward-looking, contrary to traditional sentiment measures.
The remainder of this paper is structured as follows. Section 2 reviews the relevant
literature in the context of textual analysis, as well as sentiment analysis in REIT markets.
Section 3 presents a description of the data set. Having described the basics and the
methodology of textual analysis in Section 4, a VAR is derived and the results are analysed
in Section 5. Afterwards, these results are tested regarding their robustness in Section 6.
Finally, Section 7 contains conclusions and the implications of the findings.
2. Literature review
2.1 Sentiment in the context of REIT market movements
Over the last few years, the theory of behavioural finance has replaced the efficient market
hypothesis, introduced by Fama in (1970), which is based on the idea that asset prices
incorporate all existing, new, and even hidden informationabout fundamental values.
Behavioural finance, which refers to as the collaboration between finance and a broader
social science perspective, has led to new insights into actual financial markets. The real
estate literature has evolved accordingly over the last years, augmenting traditional asset
411
Real estate
media
sentiment

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT