REIT excess dividend and information asymmetry: evidence with taxable income
| Pages | 221-236 |
| DOI | https://doi.org/10.1108/14635781011048867 |
| Date | 27 April 2010 |
| Published date | 27 April 2010 |
| Author | Ming‐Te Lee,Bang‐Han Chiu,Ming‐Long Lee,Kevin C.H. Chiang,V. Carlos Slawson |
| Subject Matter | Property management & built environment |
REIT excess dividend and
information asymmetry:
evidence with taxable income
Ming-Te Lee
Department of Accounting, Ming Chuan University, Taipei, Taiwan
Bang-Han Chiu
Department of Finance, Yuan Ze University, Chung-Li, Taiwan
Ming-Long Lee
Department of Finance, National Dong Hwa University, Shoufeng, Taiwan
Kevin C.H. Chiang
School of Business, University of Vermont, Burlington, Vermont, USA, and
V. Carlos Slawson Jr
Department of Finance, Louisiana State University, Baton Rouge,
Louisiana, USA
Abstract
Purpose – US real estate investment trusts (REITs) typically distribute more dividends than
required by tax regulations. This paper aims to focus on discretionary dividends, and examines the
impact of information asymmetry on this excess component of dividends.
Design/methodology/approach – This paper considers a set of US REITs with reported taxable
income figures over the 2000-2007 period, and employs regression analysis to examine the influence of
information asymmetry on the excess component of dividends. The explained variable is specified as
excess dividends scaled by total assets. Excess dividends are dividends paid over the mandatory
dividend payments calculated with taxable income, instead before-tax net income. Following the REIT
studies of Hardin and Hill and Han, this study employs Tobin Qas the proxy for asymmetric
information.
Findings – Contrary to Hardin and Hill’s conclusion, but consistent with dividend signaling theory
as well as agency cost explanations, the results indicate that REITs with higher level of asymmetric
information pay out significantly more excess dividends. Nevertheless, in contrast to Deshmukh’s
study on manufacturing firms, the REIT results are against the prediction of the pecking order theory.
Originality/value – The paper is one of the few studies that explicitly examine the factors
influencing REIT decision on discretionary dividends. Contrast to previous studies, this study is able
to obtain taxable income and compute the discretionary dividends more accurately. Furthermore this
paper is able to provide evidence against the pecking order theory, which is not investigated in the
existing REIT dividend studies.
Keywords Dividends, Realestate, United States of America
Paper type Research paper
Introduction
To avoid paying corporate income taxes and maintain their pass-through tax status,
real estate investment trusts (REITs) must pay out most of their taxable income as
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1463-578X.htm
REIT excess
dividend and
information
221
Received November 2009
Accepted March 2010
Journal of Property Investment &
Finance
Vol. 28 No. 3, 2010
pp. 221-236
qEmerald Group Publishing Limited
1463-578X
DOI 10.1108/14635781011048867
dividends. The dividend payout requirement was initially set at 90 percent, increased
to 95 percent in 1980 (Wang et al., 1993), and then reduced back to 90 percent in 2001
(Chan et al., 2003; Hartzell et al., 2005). This high payout requirement leaves little room
for REITs to use retained earnings to fund positive-NPV investment projects. One
would then think that REITs may particularly cherish this scare source of financing
and choose to pay out only what is required, but this is often not the case. The stylized
fact of excess dividends has long been recognized in the literature (Bradley et al., 1998;
Ghosh and Sirmans, 2006; Hardin and Hill, 2008; Wang et al., 1993).
Why would REITs pay out more dividends than is required while they
simultaneously go on the capital market to raise monies to fund growth opportunities?
A long list of existing studies addresses this important and interesting question by
focusing on the determinants of REIT dividend policy, e.g. Bradley et al. (1998), Ghosh
and Sirmans (2006), and Wang et al. (1993), to name just a few. Most of the studies
explain REIT dividend distribution as either a principal-agency problem solver or a
signaling mechanism.
The empirical underpinning of these earlier studies, however, is not completely
satisfactory for three reasons. First, Hardin and Hill (2008) point out the need to
explicitly differentiate between the mandatory and discretionary components of REIT
dividends[1]. The mandatory component is the payment to retain the tax-favored
regulatory status. It is the excess payment over and above the required component that
one can attribute to management discretion and decisions. Thus, understanding the
determinants of excess dividends is the key to addressing the dividend puzzle in the
REIT industry (Hardin and Hill, 2008).
Second, althoughHardin and Hill (2008) correctly point out thatit is the discretionary
component of dividends that is of more interest, the existing studies have not yet
provided true measurements of excess dividends. It is well understood that the excess
component of dividendsis defined relative to a certain percentageof taxable income. The
existing empirical analyses, however, are based on either a comparison of actual
dividend payment to before-tax net income or an analysis of the difference between
actual dividend payment and 90 percent (or 95 percent prior to 2001) of before-tax net
income. The reasonfor using before-tax net income as a proxy for taxableincome is that
taxable income, is not required to be disclosed, and was rarely available in the vintage
REIT era. However, as the REIT industry becomes more competitive and transparent
over time, the availability of tax income observations has improved recently.
Note that the difference between before-tax net income and taxable income can be
considerable. For example, Boston Property, Inc. has a before- tax net income of
$1,324.690 million and a taxable income of $924.061 million in 2007. The disparity is
$400.629 million! A possible main reason for this disparity is that, for accounting
reporting purposes, managers are able to pick and choose different accounting
treatments for property depreciation (Edelstein et al., 2007). In contrast, this flexibility
is largely absent for tax purposes under the governance of the Internal Revenue Service
(IRS). To the extent that firms with higher degrees of information asymmetry tend to
manage their net incomes, the use of before-tax net income may blur the empirical
relationship between excess dividends and information asymmetry [2].
Third, both agency cost and signaling theories imply that REITs convey more
information by paying higher dividends (Downs et al., 2000). Interestingly this
implication obtains no direct empirical support from the existent studies. According to
Hardin and Hill (2008, p. 345), “Given the overall dependence REITs have on the capital
JPIF
28,3
222
To continue reading
Request your trial