Outperforming the benchmark: online information demand and REIT market performance
| Published date | 02 March 2015 |
| Pages | 169-195 |
| DOI | https://doi.org/10.1108/JPIF-11-2014-0069 |
| Date | 02 March 2015 |
| Author | Karim Rochdi,Marian Dietzel |
| Subject Matter | Property management & built environment,Real estate & property,Property valuation & finance |
Outperforming the benchmark:
online information demand and
REIT market performance
Karim Rochdi and Marian Dietzel
International Real Estate Business School, University of Regensburg,
Regensburg, Germany
Abstract
Purpose –The purpose of this paper is to investigate whether there is a relationship between asset-
specific online search interest and movements in the US REIT market.
Design/methodology/approach –The authors collect search volume (SV) data from “Google
Trends”for a set of keywords representing the information demand of real estate (equity) investors. On
this basis, the authors test hypothetical investment strategies based on changes in internet SV, to
anticipate REIT market movements.
Findings –The results reveal that people’s information demand can indeed serve as a successful
predictor for the US REIT market. Among other findings, evidence is provided that there is a
significant relationship between asset-specific keywords and the US REIT market. Specifically,
investment strategies based on weekly changes in Google SV would have outperformed a buy-and-
hold strategy (0.1 percent p.a.) for the Morgan Stanley Capital International US REIT Index by a
remarkable 15.4 percent p.a. between 2006 and 2013. Furthermore, the authors find that real-estate-
related terms are more suitable than rather general, finance-related terms for predicting REIT market
movements.
Practical implications –The findings should be of particular interest for REIT market investors, as
the established relationships can potentially be utilized to anticipate short-term REIT market
movements.
Originality/value –This is the first paper which applies Google search query data to the REIT
market.
Keywords Real estate, REIT, Google Trends, Information demand, Investment strategy,
Search query data
Paper type Research paper
1. Introduction
It is common knowledge that an investor’s decision about whether to invest in or divest
from the stock market is determined by a variety of factors. Besides business-related
news, it might be factors like natural disasters, the resignation of business leaders,
terrorist attacks, let alone all kinds of economic fundamentals and political reports that
make markets fluctuate. No one would seriously claim to be able to foresee such events
with any accuracy or reliability. However, broken down to the very basics, the price of
a stock is still determined by demand and supply, by one party who is willing to buy
and another who is willing to sell. Apart from trading computers, the largest share of all
financial transactions is still conducted by human beings who make a buy or sell
decision. There can be no doubt that this decision is influenced by the abovementioned
events, but in between an event or the release of certain news and a (human) financial
transaction, people gather further information. In a world of smart phones, tablets and
laptops, the internet has become the main source of this information. Therefore, big
data and search query data in particular are becoming increasingly interesting for
Journal of Property Investment &
Finance
Vol. 33 No. 2, 2015
pp. 169-195
©Emerald Group Publis hing Limited
1463-578X
DOI 10.1108/JPIF-11-2014-0069
Received 26 November 2014
Revised 26 November 2014
Accepted 11 December 2014
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1463-578X.htm
169
Outperforming
the benchmark
those researching equity markets, as it represents an appropriate instrument for
quantifying internet users’interests and motives.
In a seminal article, Preis et al. (2013a) use Google search volume (SV) data in sea rch-
interest-based trading strategies to examine the relationship between online search
behavior and stock market movements (Dow Jones Industrial Average, DJIA). They
find that online search behavior does indeed serve as an indicator of stoc k market
movements. This raises the question of whether searchers leave more (less) traces
online when researching more (less) complex asset classes, since estimating the fair
value of a non-transparent good or market is much more time-consuming and
elaborate. Thus, we hypothesize that the more research-intensive an asset, that is, the
more information is needed before making a buy or sell decision, the better the chances
of predicting the searchers’behavior with regard to their subsequent (trans)actions.
As the real estate investment trust market constitutes a relatively research-intensive
asset, it is a suitable example for further analysis. This, of course, is due to the fact that
information-gathering on REITs is considered to be more comprehensive, because both
capital and space markets have to be analyzed (Roulac, 1988). Generally, a rational
investor analyzes a REIT portfolio thoroughly before making a decision about whether
the stock is currently fairly priced. Besides fundamental equity market analysis, this
mainly includes appraising the relevant (property) markets and the future development
of rental income and yields. This relates to the discussion among researchers on the
extent to which REITs behave more similarly to property or stock markets. In order to
investigate this issue, as well as the relationship between information demand and the
US REIT market in general, we apply a methodology similar to Preis et al. (2013a). For
this purpose, two groups of keywords are compiled, one containing real-estate-related
search terms, the other one (rather general) finance-related search terms. Accord ingly,
each search term constitutes an individual information-demand-based investment
strategy whose trading signals derive from weekly changes in the underlying Google
SV. Subsequently, the overall performance of the investment strategies from the two
keyword subsets is compared with one another in order to gain knowledge about what
kind of information demand predicts the REIT market more successfully. Also, since
the past decade was characterized by turbulent markets, the time-specific dynamics of
the relevance of information demand are of particular interest. Additionally, if Google
search interest is linked to stock trading volume, as found by Preis et al. (2010),
information demand should be a particularly good predictor during phases of
exceptionally high returns or losses, which is why we also determine the strategies’
predictive ability for the 40 most extreme upward and downward market movements
between 2006 and 2013. All tested investment strategies are benchmarked against a
buy-and-hold strategy for the Morgan Stanley Capital International (MSCI) US REIT
Index, as well as the DJIA.
This paper contributes in several ways to the literature on the information demand
of real estate (equity) investors, real estate equity markets and online search behavior.
First and foremost, we find that search query data serve as a successful predictor for
the US REIT market. Moreover, the results suggest that asset-specific (real-estate
specific) search terms are better predictors for the US REIT market than finance-related
search queries. Also, this is the first paper to examine the dynamics of Google Trends
investment strategies’(GTIS) investment performance over time. The findings reveal
that particularly during the crisis of 2008-2011, a period of substantial investor
uncertainty and increased information demand, investment strategies based on the
Google data set predict the market very successfully. This is supported by the fact that
170
JPIF
33,2
To continue reading
Request your trial