Available online at www.sciencedirect.

com

Journal of Policy Modeling 35 (2013) 29–44

Is market integration associated with informational

efficiency of stock markets?
Chee-Wooi Hooy a , Kian-Ping Lim b,∗
a Finance Section, School of Management, Universiti Sains Malaysia, Malaysia
b Labuan School of International Business and Finance, Universiti Malaysia Sabah, Malaysia
Received 13 April 2012; received in revised form 29 June 2012; accepted 14 September 2012
Available online 23 September 2012

Abstract
This study addresses the question of whether a more integrated stock market is associated with a higher
degree of informational efficiency. We employ the adjusted pricing error from an equilibrium international
asset pricing model as a proxy for market integration. The aggregate country-level price delay serves as
an inverse measure of informational efficiency, as it captures the relative speed with which each aggregate
stock market reacts to global common information. Using data from 49 countries, we find robust evidence
supporting the hypothesis that markets more integrated with the world are also more efficient, and this
positive association is only significant in the sub-sample of emerging stock markets. The results provide
additional insight on the factors facilitating the transmission of global information and yield important policy
implications.
© 2012 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved.

JEL classification: F36; G14; G15

Keywords: Informational efficiency; Market integration; Financial liberalization; Price delay; Pricing error

1. Introduction

Since the late 1980s, many developing countries have actively pursued financial liberaliza-
tion policies, with the anticipation that the opening of capital account will deliver higher rates

∗ Corresponding author at: Labuan School of International Business and Finance, Universiti Malaysia Sabah, Jalan

Sungai Pagar, 87000 F.T. Labuan, Malaysia. Tel.: +60 87460513; fax: +60 87460477.
E-mail address: kianpinglim@yahoo.com (K.-P. Lim).

0161-8938/$ – see front matter © 2012 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jpolmod.2012.09.002
30 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

of economic growth. Their broad liberalization packages also include the removal of statutory
restrictions on foreign ownership of domestic equity securities (for stock market openings, see
Bekaert & Harvey, 2000; Henry, 2000). As a result, most emerging market economies experienced
surges in the volume of international capital flows over the next two decades. However, a series of
financial crises in the 1990s and the recent global financial turmoil have triggered intense debate
in both the academic and policy circles on the desirability of full liberalization of capital flows.
Providing further ammunition to their critics is the lack of conclusive evidence on the positive
growth effect of financial openness, as highlighted by several survey papers (see Edison, Klein,
Ricci, & SlØk, 2004; Henry, 2007; Kose, Prasad, Rogoff, & Wei, 2009). Despite the controversy
on its growth-enhancing benefit, a consensus emerges that these liberalization policies have fur-
ther integrated stock markets with the world, though the process is still far from the aspired level
of perfect integration (see Carrieri, Errunza, & Hogan, 2007; Carrieri, Chaieb, & Errunza, 2011;
Bekaert, Harvey, Lundblad, & Siegel, 2011).
The objective of this paper is to further determine whether the increased integration with
world capital markets goes hand in hand with a higher degree of informational efficiency. The
efficiency effect deserves greater attention in policymaking than it presently receives for at least
two reasons. First, promoting informational efficiency, in which the dissemination of relevant
information is timely reflected in the price formation process, is one of the key objectives for
securities market regulators.1 In fact, efficient price discovery is one main function of stock
market. Unfortunately, after thousands of published articles spanning over several decades, little
is known about how, when and why the stock price becomes efficient. Only in recent years that
the academic literature pays relatively more attention to the efficiency effect of various market
reforms undertaken worldwide, such as securities laws (Daouk, Lee, & Ng, 2006), insider trading
laws (Fernandes & Ferreira, 2009), short-selling regulations (Bris, Goetzmann, & Zhu, 2007; Saffi
& Sigurdsson, 2011), private property rights protection (Morck, Yeung, & Yu, 2000), corporate
transparency (Jin & Myers, 2006), trade opening (Lim & Kim, 2011) and financial liberalization
(Bae, Ozoguz, Tan, & Wirjanto, 2012; Li, Morck, Yang, & Yeung, 2004). An understanding of
the key drivers behind informational efficiency is likely to provide useful input to policymakers
and stock exchange regulators.
Second, the informational efficiency of a stock market warrants the attention of policymakers
to avoid substantial misallocation of resources that has a negative impact on long-term eco-
nomic growth. Morck, Shleifer, and Vishny (1990) point out that market efficiency would not be
important if the stock market does not affect economy activity. According to Majumder (2012),
wrong investment strategies may disrupt the optimal allocation of resources. To establish the link
between an efficient stock price and efficient allocation of investment resources, Dow and Gorton
(1997) develop a theoretical model which shows that the stock market indirectly guides managers’
investment decisions by transferring information about investment opportunities and information
about managers’ past investment decisions. Following their work, the feedback effect from stock
prices to real investment decisions has attracted a sizeable theoretical literature (for recent stud-
ies, see Dow, Goldstein, & Guembel, 2011; Goldstein & Guembel, 2008). Empirically, several
studies find that the stock market is not an economic sideshow, but instead, efficient stock prices
enhance the efficiency of capital allocation (Wurgler, 2000), induce higher productivity and faster
economic growth (Durnev, Li, Morck, & Yeung, 2004), facilitate more efficient corporate capital

1 The objectives and principles of securities regulations can be downloaded from the website of International Orga-

nization of Securities Commissions (IOSCO), the leading international grouping of securities market regulators, at
http://www.iosco.org/library/pubdocs/pdf/IOSCOPD154.pdf (retrieved on 02.05.12).
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 31

investments (Durnev, Morck, & Yeung, 2004), and exert strong positive effect on the sensitivity
of corporate investment to stock prices (Chen, Goldstein, & Jiang, 2007).2 Given the profound
implications of an efficient stock market to the real sector, it is important to explore the efficiency
effect of macroeconomic or financial reforms, an aspect largely ignored by policymakers and
previous academic studies.3
Our present investigation provides useful input to the formulation of capital market develop-
ment plans, given that enhancing greater integration and promoting market efficiency are always
at the core of policy goals. A pertinent question is whether there exists a trade-off between these
two objectives. If increased stock market integration resulted from official liberalization is nega-
tively related to informational efficiency, then policymakers might need to reconsider their prior
commitment to financially integrate with the global capital markets. In such a scenario, the policy
is only worth pursuing if the cost-benefit analysis shows that the growth benefits are large enough
to outweigh the costs where the latter will now include efficiency loss. Conversely, if a greater
level of market integration is associated with improved informational efficiency, the finding then
throws further support to policy reforms aim at attaining full financial integration. In fact, the
efficiency gain in which global information is incorporated swiftly will be an additional benefit
not reported by previous studies. Finally, the absence of a significant association indicates that
both policy goals are independent from one another.
It is worth highlighting that market integration and informational efficiency has remained sep-
arate subjects in finance. Rockinger and Urga (2001) and Schotman and Zalewska (2006) straddle
the two disparate streams of burgeoning literature to simultaneously test for the varying degrees
of market integration and informational efficiency over time. However, they do not examine the
empirical relationship between these two key policy variables, a void in the extant literature that
motivates the present paper. To the best of our knowledge, only Li et al. (2004) and Bae et al.
(2012) examine the efficiency effect of financial liberalization at the country and firm levels,
respectively. Both studies find that a greater level of stock’s accessibility to foreign investors
improves the informational efficiency of domestic stock markets. However, their use of investible
weight from the Standard & Poor’s Emerging Markets Database is subject to criticism as the
indicator measures the intensity and evolution of de jure stock market openness. It is well docu-
mented that the removal of statutory foreign investment restrictions alone is not sufficient to foster
complete financial integration with the world capital markets. The commonly cited impediment
to effective integration is the existence of indirect barriers that discourage foreign investment,
such as the availability of information, differential accounting standards, investor protection, and
country risk (for details, see Bekaert, 1995; Nishiotis, 2004; Carrieri et al., 2011).4 Motivated

2 Interested readers can refer to Bond et al. (in press) for a thorough discussion and literature review of the potential
real effects of financial markets.
3 The connection between the stock market and real sector activity is not something new to the literature. Fama (1990)

and Schwert (1990), for instance, find that current stock returns are highly correlated with future production growth rates,
indicating that information about future cash flows is impounded in stock prices. A significant link between stock price
and economy activity has implication on monetary policy targeting asset prices (see, for example, Bordo & Jeanne, 2002;
Leduc & Natal, 2011; Roubini, 2006).
4 We do note that several studies define financial integration in terms of the elimination of legal barriers to capital or

portfolio flows (see, for example, Bonfiglioli, 2008; Buch & Pierdzioch, 2005; Schindler, 2009). These official liberaliza-
tion measures only capture the ability of foreign investors to invest in the domestic country (direct barriers) but disregard
the existence of indirect barriers that affect foreign investors’ willingness to invest. Market integration also differs from
the concept of economic integration which has attracted a huge literature (see, for example, Bouët et al., 2012; Kim, 2005;
Norén, 2011).
32 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

by this concern, our study directly measures the degree of stock market integration because the
indicator, by construction, has taken into account both the direct and indirect forms of investment
barriers. In other words, the present focus is on financially integrated rather than financially open
markets.
The rest of this paper is organized as follows. Section 2 discusses the research framework in
this paper. Following that, we describe the sample and present some preliminary analysis. Section
4 provides the empirical results for our hypothesis. Concluding remarks are given in the final
section.

2. Hypothesis, measurement of variables and model specification

2.1. Hypothesis development

The hypothesis put forward in this paper is that there exists a positive association between stock
market integration and informational efficiency, where the latter is measured in terms of the speed
with which each aggregate stock market reacts to global common information. To rationalize
the hypothesized positive relationship, we compare the proportion of stock returns explained by
global over local factors under three different scenarios. First, if a market is fully segmented from
the rest of the world, its stocks are exposed purely to local market shocks, and hence we would
expect significant delayed response to global information. Second, in the other extreme case of
perfect integration, the market is sensitive only to world events, which implies that common global
factors account for all the variation in stock returns. It is then logical to expect stocks in a fully
integrated market to respond swiftly to global information. Third, in between the two extremes,
stock returns are determined by a combination of local and global factors, with the importance of
the latter increasing with the degree of integration. The incorporation of global information, in
this intermediate case of partially integrated market, will be faster than a completely segmented
market but slower than the speed in perfect integration. To sum up, our line of reasoning suggests
that when a market becomes more globally integrated, the explanatory power of global over local
factor increases, leading to a faster incorporation of global information into stock prices and hence
a higher degree of informational efficiency. In other words, integrated markets respond more and
faster to global factors.

2.2. The empirical measure for stock market integration

We use a simple yet widely accepted equilibrium asset pricing model to measure market
integration—a single factor International Capital Asset Pricing Model (ICAPM) by Stehle (1977),
which takes the following form:

rtm = α + βrtw + εt (1)

where rtm is the local market excess return at week t, and rtw is the world excess return at week t.
If a market is perfectly integrated with the world, the intercept in the above model should be
equal to zero, indicating no mispricing in the ICAPM. Korajczyk (1996) shows that the mispricing
measure (α) is positively correlated with higher official barriers, tax on international asset trading,
transaction cost and larger impediments on firm information.
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 33

Following Levine and Zervos (1998), we take the absolute value of the pricing error in Eq. (1),
since both positive and negative deviations are indicative of market segmentation. Our empirical
measure for market integration is taken as the absolute pricing error multiplied by −1:
INTEGRATE = − |α| (2)
Hence, a higher value of INTEGRATE indicates a greater level of integration between the
domestic stock market and the rest of the world.

2.3. The empirical measure for stock market informational efficiency

Our metric of informational efficiency is the stock price delay measure. For cross-country study,
it is more appropriate to use global market-wide public news so that the price delay captures the
relative speed with which each aggregate stock market reacts to this common information set.
Hence, we utilize the country-level price delay measure proposed by Lim and Hooy (2010), which
involves the following unrestricted version of ICAPM:

4
rtm = α + βrtw + w
δk rt−k + εt (3)
k=1

Eq. (1) is the restricted version of Eq. (3) where the coefficients on the lagged world market
returns are constrained to zero. The R-squares from both equations are used to calculate the price
delay measure:
R2restricted
DELAY = 1 − (4)
R2unrestricted
The larger the value of the DELAY, the more variation in the domestic market index returns
that is explained by lagged world market returns, indicating a greater delay in the response of
domestic stock market to global market-wide news that has common effects across countries.
Hence, the price delay is an inverse measure of informational efficiency, where a higher value of
DELAY indicates a lower degree of efficiency.

2.4. Model specification

We employ pooled ordinary least squares (OLS) to examine the empirical relationship between
market integration and informational efficiency. In comparison with the literature on market
integration, the factors that affect price delay are well-grounded theoretically and receive strong
empirical support (see the survey paper by Lim, 2009). We therefore control for the effects of these
cross-sectional determinants of price delay in our regression. Due to data availability, our analysis
focuses on the country-level counterparts for firm size (SIZE), trading volume (VOLUME), short
sales restrictions (SHORTING) and the degree of investibility (INVEST).
Hence, we estimate the following pooled cross-sectional OLS regression model:

DELAYi,t = μ + δINTEGRATEi,t + Xi,t β + vi,t (5)
Our parameter of interest, δ, measures the impact of market integration on informational
efficiency. The four control variables (SIZE, VOLUME, SHORTING, INVEST) are reflected in the
 with the coefficient vector β. v is the error term that captures all other omitted factors
vector Xi,t i,t
34 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

in the regression. The standard errors are robust to heteroskedasticity in the variance–covariance
matrix, and they allow for clustering at the country-level.

3. Sample

3.1. Data sources

We collect country indices at the weekly frequency (Wednesday closing prices) for 49 stock
markets over the sample period of 1995–2007 from Morgan Stanley Capital International (MSCI).
These series are in common currency (the U.S. dollar) to alleviate exchange rate noise. We use
the value-weighted MSCI All-Country World Index as a proxy for global factors and the rates
for U.S. 3-month Treasury bill as the global risk free rates. All the stock indices are transformed
into continuously compounded percentage returns. For the empirical measure of stock market
integration, we estimate Eq. (1) annually using weekly indices and then calculate INTEGRATE
based on Eq. (2). This gives us 13 annual observations for each country i. In a similar vein, Eq. (3)
is estimated using weekly indices and the DELAY is calculated based on Eq. (4) for each country
and year.
The sources for our four control variables are as follows. First, the natural logarithm of market
capitalization of listed companies is used to proxy for the size of the stock market. Second, our
proxy for trading volume is the natural logarithm of one plus the turnover ratio. The panel data for
market capitalization and turnover ratio are collected from the World Bank’s World Development
Indicators (WDI). Third, we utilize the dataset on the legality and feasibility of market-wide short
selling assembled by Daouk and Charoenrook (2005) based on their survey of stock exchange
officials. The short sales dummy variable, which reflects the ability of investors in taking short
positions, equals one if either short selling or put options trading is feasible in a given country
and year, zero otherwise. Finally, we construct the country-level degree of investibility using the
ratio of the number of firms in the Global Index (IFCG) and Investable Index (IFCI), extracted
from the Emerging Stock Markets Factbook (1996–2002) and the Global Stock Markets Factbook
(2003–2007). However, the annual data are only available for the period up to 2006. This ratio
provides a quantitative measure of the availability of the country’s equities to foreigners, with
values ranging from zero (completely closed to foreign investors) to one (completely open market
with no foreign restrictions).

3.2. Descriptive statistics

The descriptive statistics and the correlation matrix of all the variables are reported in Table 1.
The most important information from the descriptive statistics reported in Panel A is the varia-
tion of each panel series in total, between markets and within each market. Apparently, there is
considerable variation in both between and within countries, hence justifying the use of pooled
cross-sectional regression. Panel B reports the correlation matrix. Note that all the explanatory
variables are negatively correlated with stock price delay, which is consistent with the firm-level
evidence. Comparing to the four control variables, market integration has the lowest degree of
correlation with informational efficiency. Among the explanatory variables, the highest corre-
lation is between short sales and the degree of investibility. It is worth highlighting that all the
reported correlation coefficients are statistically significant but the degree of correlation is not
strong, indicating that multicollinearity is not a major concern.
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 35

Table 1
Descriptive statistics and correlation matrix.
DELAY INTEGRATE SIZE VOLUME SHORTING INVEST

Panel A: descriptive statistics

Mean 0.3365 −0.4228 25.5893 3.9514 0.7127 0.7900
Median 0.2379 −0.2865 25.5744 4.0325 1.0000 0.9860
Maximum 1.0000 −0.0010 30.6241 6.6631 1.0000 1.0000
Minimum 0.0017 −4.0099 20.7948 0.8502 0.0000 0.0000
Overall Std dev 0.3011 0.4310 1.8056 0.9560 0.4529 0.3096
Between Std dev 0.2149 0.3468 0.6109 0.4659 0.0781 0.0971
Within Std dev 0.2814 0.3849 1.7436 0.9418 0.4519 0.3083
No. of observations 637 637 636 637 637 588
Panel B: correlation matrix
DELAY 1.0000
INTEGRATE −0.2733 1.0000
SIZE −0.6148 0.2820 1.0000
VOLUME −0.3440 0.0973 0.5325 1.0000
SHORTING −0.5360 0.2670 0.5795 0.3482 1.0000
INVEST −0.5457 0.2327 0.6278 0.4259 0.6627 1.0000

Notes: Std dev refers to the standard deviation. The sample period is from 1995 to 2007, covering 22 developed and 27
emerging markets.
DELAY is the country-level price delay and serves as an inverse measure of informational efficiency, where a higher value
indicates a lower degree of informational efficiency.
INTEGRATE is measured by the negative-sign pricing errors, where a higher value (i.e., a value closer to zero) indicates
a greater level of domestic stock market integration with the world.
SIZE is the natural logarithm of market capitalization of listed companies, proxying for the size of the stock market.
VOLUME stands for trading volume, proxied by the natural logarithm of one plus the turnover ratio.
SHORTING is a short-sales dummy variable that reflects the ability of investors to take short positions. It equals one if
either short selling or put options trading is feasible in a given country and year, zero otherwise.
INVEST measures the degree of foreign investibility, with values ranging from zero (completely closed to foreign investors)
to one (completely open market with no foreign restrictions).
Developed Markets: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Hong Kong, Ireland,
Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland, United Kingdom, and
United States.
Emerging Markets: Argentina, Brazil, Chile, China, Colombia, Czech Republic, Egypt, Hungary, India, Indonesia, Israel,
Jordan, Korea, Malaysia, Mexico, Morocco, Pakistan, Peru, Philippines, Poland, Russia, South Africa, Sri Lanka, Taiwan,
Thailand, Turkey, and Venezuela.

3.3. Behavior of the two key policy variables

The computed values of INTEGRATE and DELAY are averaged over the sample period from
1995 to 2007 for each of the 49 countries, as plotted in Fig. 1. As expected, nearly all the developed
markets exhibit a greater level of market integration than their emerging counterparts, indicating
that developed stock markets are more integrated with the world market. The stock price delay
measure behaves quite similarly, supporting the conventional wisdom that developed markets are
more efficient in incorporating information into stock prices.
Fig. 2 plots the time-series evolution of the computed values of INTEGRATE and DELAY
over the sample period 1995–2007, averaged across all countries for two groups: developed and
emerging markets. The figure shows that, in all of our sampled 13 years, developed markets exhibit
a greater level of integration and a higher degree of informational efficiency than their emerging
counterparts. However, the gap between these two groups in both aspects has been narrowing
36 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

Fig. 1. Cross-sectional variation in market integration and informational efficiency. Notes: The figure plots the average
values of INTEGRATE and DELAY over the sample period 1995–2007 for each of the 49 sampled countries. Market
integration is measured by the negative-sign pricing errors, where a higher value (i.e., a value closer to zero) indicates a
greater level of domestic stock market integration with the world. Informational efficiency is measured in terms of the
stock price delay, where a higher value of DELAY indicates a lower degree of informational efficiency.
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 37

Developed Market Efficiency Emerging Market Efficiency

Developed Market Integration Emerging Market Integration
0.8

0.3

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
-0.2

-0.7

-1.2

Fig. 2. Time-series variation in market integration and informational efficiency. Notes: The figure plots the time-series
evolution of the computed values of INTEGRATE and DELAY over the sample period 1995–2007, averaged across all
countries for two groups: developed and emerging markets. Market integration is measured by the negative-sign pricing
errors, where a higher value (i.e., a value closer to zero) indicates a greater level of domestic stock market integration
with the world. Informational efficiency is measured in terms of the stock price delay, where a higher value of DELAY
indicates a lower degree of informational efficiency.

since 2004. It is difficult to infer from Fig. 2 on the existence of a long-term trend, as we do
observe reversals at times.
To get a preliminary view of the relationship between market integration and informational
efficiency, Fig. 3 plots the time-series averages of DELAY against the average scores of INTE-
GRATE for each of the 49 sampled countries. It is clear that there is a relatively strong negative
relationship between the stock price delay and market integration. As the former is an inverse
0.8

JOR
EGY PAK MAR
LKA
0.6

IDN COL
VEN
IND
CZE
PHL
AUT
PER
TUR NZL
MYS PRT
0.4

HUN TWN
THA
RUS POL
CHN
KOR SGP DNK
ARG BEL
NOR
CHL
JPN CHE
0.2

ISR AUS
DELAY = 0.07 - 0.64 INTEGRATE BRA ZAF IRLHKG
FIN ESP ANT
FRA
MEX ITA GBR
SWE
CAN DEU
USA
0.0

-1.0 -0.8 -0.6 -0.4 -0.2 0.0

Pricing Error (Market Integration), 1995-2007

Fig. 3. Scatter plot of informational efficiency against market integration. Notes: The figure plots the time-series averages
of DELAY between 1995 and 2007 against the average scores of INTEGRATE over a similar sample period. Market
integration is measured by the negative-sign pricing errors, where a higher value (i.e., a value closer to zero) indicates a
greater level of domestic stock market integration with the world. Informational efficiency is measured in terms of the
stock price delay, where a higher value of DELAY indicates a lower degree of informational efficiency.
38 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

measure of informational efficiency, this implies a positive association between the two core pol-
icy variables. To evaluate its statistical significance, we turn to the formal panel analysis in the
subsequent section.

4. Empirical results

4.1. Baseline pooled OLS results

Table 2 presents the pooled OLS results with the country-level price delay measure as the
dependent variable. To begin with, we run a simple univariate pooled regression and report the
output in column (1). The regressor of INTEGRATE is statistically significant at the 1% level.
The negative coefficient suggests that a greater level of market integration is associated with a
lower value of price delay, and hence a higher degree of informational efficiency. Next, we add
a time trend to the model to address the concern that our earlier significant coefficient may pick
up the common trend in both policy variables. The result reported in column (2) shows that the
inclusion of time trend does not subsume the explanatory power of INTEGRATE. The time trend
has a significant negative coefficient of −0.0172, but its inclusion adds only 4.5% to the R2 of the
regression.
We next enter the control variables one by one into columns (3) to (6). Market size, trading
volume, short sales restrictions and the degree of investibility are all highly significant with the
expected negative coefficients, suggesting that these variables remain important determinants of

Table 2
The relationship between market integration and informational efficiency.
(1) (2) (3) (4) (5) (6) (7)

Intercept 0.2558 0.3670 2.7525 0.7177 0.4867 0.7997 2.2391

(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
INTEGRATE −0.1909 −0.1720 −0.0723 −0.1566 −0.1199 −0.0898 −0.0564
(0.0000) (0.0000) (0.0018) (0.0000) (0.0000) (0.0001) (0.0109)
SIZE −0.0939 −0.0648
(0.0000) (0.0000)
VOLUME −0.0924 0.0045
(0.0000) (0.6922)
SHORTING −0.1958 −0.0067
(0.0000) (0.7633)
INVEST −0.4901 −0.2724
(0.0000) (0.0000)
Trend −0.0172 −0.0071 −0.0137 −0.0164 −0.0191 −0.0114
(0.0000) (0.0072) (0.0000) (0.0000) (0.0000) (0.0000)
No. of observations 637 637 636 637 637 637 636
No. of countries 49 49 49 49 49 49 49
R2 0.0747 0.1197 0.3961 0.2032 0.2195 0.3602 0.4428
Adjusted R2 0.0732 0.1169 0.3932 0.1994 0.2158 0.3571 0.4375

Notes: The dependent variable is DELAY constructed for each country in each year over 1995–2007, thus capturing the
delay with which each aggregate stock market responds to global common information.
Descriptions for all the explanatory variables are provided in the Notes to Table 1. The data for all variables are available
for the full sample period 1995–2007, with the exception of INVEST which is available only up to 2006.
The regressions are estimated using pooled OLS. Entries in parentheses are the two-tailed p-values. The standard errors
are robust to heteroskedasticity in the variance–covariance matrix, and they allow for clustering at the country-level.
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 39

how fast stock prices adjust to common information at the country level. It is worth noting that
among the four control variables, the inclusion of SIZE contributes the most to the R2 of the
regression, at approximately 28%. This is followed closely by INVEST, with a 24% contribution.
Our key variable of interest, INTEGRATE, is still negatively and significantly related to price
delay in all the model specifications. Finally, we determine whether market integration can still
explain the time-series and cross-sectional variation in price delay after controlling for all the
four competing variables. As reported in column (7), INTEGRATE retains its negative sign and
statistical significance along with SIZE and INVEST, whereas the other two regressors lose their
predictive power.
One result emerges from Table 2 worth our mentioning is that INTEGRATE has a significantly
negative coefficient even after controlling for INVEST. This indicates the former has an incremental
effect on price delay, further implying the degree of investibility is an imperfect measure of
market integration.5 The result is not surprising because the removal of legal restrictions alone is
not sufficient to foster complete financial integration with the global capital markets due to the
presence of indirect investment barriers. The strongest empirical support for our interpretation
comes from the recent work of Carrieri et al. (2011). Using data from 22 emerging markets,
these authors compute the degree of integration for investible and non-investible indices, with the
difference being explicit barriers have been removed for the first group. Carrieri et al. (2011) find
that the returns on both segments are determined by a combination of local and global factors.
In the case of investible index, the significance of local risk factor indicates that the reduction in
explicit barriers in conjunction with financial liberalization does not lead to full market integration.
Instead, their analysis shows that the lowering of implicit barriers is associated with a higher degree
of market integration. More specifically, better institutions, stronger corporate governance, and
more transparent markets are found to jointly explain 15–20% of the variation in the integration
measure.

4.2. Alternative estimation methods

The baseline pooled OLS results provide strong evidence to support our hypothesis of a positive
association between market integration and informational efficiency. Table 3 checks the robustness
of our core finding across different estimation methods. Column (1) uses two-way fixed effects
to account for time-series and cross-sectional dependence. Column (2) employs random effects
model which takes the form of random intercepts assuming the regressors are uncorrelated with
the country effect, and could provide more efficient estimates than the fixed effects estimator.
Column (3) considers the population-average GLS estimator that adjusts for residuals correlation.
Column (4) employs a standard Fama and MacBeth (1973) two-step procedure which estimates the
same cross-sectional regression for each year and then takes the time-series averages of the OLS
estimated coefficients. Column (5) instead estimates the standard Fama–MacBeth regression using

5 To ensure that market integration has an independent influence on price delay that goes beyond the effect of INVEST,

we conduct two additional tests as follows. First, we extract the pure market integration component that is orthogonal to
investibility by regressing INTEGRATE on INVEST, and construct RESID INTEGRATE using the intercept plus residuals
from this regression. This new variable can be interpreted as representing the dimension not captured by INVEST, which
is implicit investment barriers. We replace INTEGRATE with RESID INTEGRATE in column (7) of Table 2 and find
that the latter is still strongly and negatively related to the price delay measure. Second, we test the null hypothesis that
the INTEGRATE coefficient is equal to the INVEST coefficient. The null hypothesis is rejected even at the 1% level of
significance. All the results are available upon request from the authors.
40 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

Table 3
Alternative estimation methods.
(1) (2) (3) (4) (5) (6)
Fixed effects Random GLS Fama- WLS Fama- Dynamic
effects population MacBeth MacBeth system GMM
averaged

INTEGRATE −0.0514 −0.0648 −0.0677 −0.0865 −0.0866 −0.0801

(0.0706) (0.0149) (0.0107) (0.0471) (0.0470) (0.0437)
SIZE −0.1321 −0.0733 −0.0688 −0.0616 −0.0617 −0.1127
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0330)
VOLUME −0.0386 −0.0077 0.0045 0.0117 0.0117 −0.0098
(0.1869) (0.7041) (0.7990) (0.3998) (0.4014) (0.8161)
SHORTING −0.0181 −0.0104 0.0004 −0.0122 −0.0121 0.1195
(0.7654) (0.7559) (0.9910) (0.6351) (0.6379) (0.3348)
INVEST −0.1004 −0.1973 −0.2260 −0.2775 −0.2776 −0.1738
(0.3159) (0.0035) (0.0001) (0.0005) (0.0005) (0.2866)
No. of observations 636 636 636 636 636 587
No. of countries 49 49 49 49 49 49

Notes: The dependent variable is DELAY constructed for each country in each year over 1995–2007, thus capturing the
delay with which each aggregate stock market responds to global common information.
Descriptions for all the explanatory variables are provided in the Notes to Table 1. The data for all variables are available
for the full sample period 1995–2007, with the exception of INVEST which is available only up to 2006.
The table presents coefficient estimates for the explanatory variables using different estimators. Model (1) is a two-way
fixed effects estimator. Model (2) is a random effects estimator. Model (3) is GLS population-averaged linear estimator,
where we specify the error structure to follow an AR(1) process. Models (4) and (5) employ the Fama and MacBeth (1973)
two-pass regression approach, estimated using OLS and WLS, respectively, with the standard errors adjusted using the
Newey–West procedure. Model (6) is a dynamic system GMM estimator with a lagged dependent variable in the model.
Entries in parentheses are the two-tailed p-values. To conserve space, the diagnostic results are not reported but are
available upon request from the authors.

weighted least-squares (WLS) method, which provides precision-weighted time-series averages

of the cross-sectional regressions coefficients. The final column (6) includes a dynamic system
GMM estimator, which allows the delay process to be dynamic (with the inclusion of the lagged
delay) and uses the first-differences and the levels equations as instruments. Our core finding is
not affected by these different estimators. The INTEGRATE coefficient is still significant with the
negative sign, and its effect is independent of the role played by INVEST.

4.3. Sub-sample analysis

We investigate whether the 1997/1998 Asian financial crisis affects the positive association
between market integration and informational efficiency. The sample is split into two, crisis and
non-crisis (i.e., excluding the years of 1997 and 1998). Table 4 reports the pooled OLS results.
The key findings remain unaffected, with only one exception that SIZE losses its explanatory
power during the crisis period.
The next analysis determines whether our key results are consistent across two sets of countries
with different characteristics and environments. We re-estimate Eq. (5) using the pooled OLS for
developed and emerging markets samples separately. Table 4 shows that neither INVEST nor
INTEGRATE is statistically significant in developed markets, indicating that financial liberaliza-
tion is not a main factor for the efficient functioning of their stock markets. Instead, the results
establish a link between financial development and informational efficiency in these industrial
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 41

Table 4
Sub-sample analysis.
Non-crisis Crisis Developed markets Emerging markets

Intercept 2.3633 1.6207 2.5696 2.3126

(0.0000) (0.0002) (0.0002) (0.0000)
INTEGRATE −0.0874 −0.0857 −0.0107 −0.0611
(0.0049) (0.0087) (0.8373) (0.0246)
SIZE −0.0708 −0.0283 −0.0542 −0.0698
(0.0000) (0.1327) (0.0000) (0.0000)
VOLUME 0.0103 −0.0090 −0.0708 0.0192
(0.4016) (0.7636) (0.0050) (0.1741)
SHORTING −0.0082 −0.0327 0.0038 −0.0368
(0.7345) (0.5381) (0.9138) (0.2599)
INVEST −0.2327 −0.4385 −0.5801 −0.2686
(0.0000) (0.0001) (0.3901) (0.0000)
Trend −0.0146 −0.0882 −0.0069 −0.0116
(0.0000) (0.0447) (0.0531) (0.0065)
No. of observations 538 98 286 350
No. of countries 49 49 22 27
R2 0.4654 0.4471 0.3025 0.3687
Adjusted R2 0.4594 0.4106 0.2875 0.3576

Notes: The dependent variable is DELAY constructed for each country in each year over 1995–2007, thus capturing the
delay with which each aggregate stock market responds to global common information.
Descriptions for all the explanatory variables are provided in the Notes to Table 1. The data for all variables are available
for the full sample period 1995–2007, with the exception of INVEST which is available only up to 2006.
The table presents coefficient estimates using a pooled OLS for two sets of sub-samples based on (1) 1997/1998 Asian
financial crisis and (2) market status.
Entries in parentheses are the two-tailed p-values. The standard errors are robust to heteroskedasticity in the
variance–covariance matrix, and they allow for clustering at the country-level.

countries, in particular market size and liquidity. We conclude that our findings from previous
full sample analysis are mainly driven by the sub-sample of emerging markets, given that SIZE,
INVEST and INTEGRATE remain negatively and significantly associated with the price delay. The
results are not surprising since all developing countries had some forms of investment restrictions
in place and only gradually began their liberalization process in the late 1980s. In fact, most
previous studies on financial liberalization and integration generally confine their investigation to
emerging markets.

5. Policy discussions and concluding remarks

This paper extends the empirical literature on market integration by exploring its associa-
tion with informational efficiency, an issue that is still largely unexplored in previous academic
studies. Our proxy for market integration is the adjusted pricing error from an equilibrium inter-
national asset pricing model, which by construction, has taken into account both the direct and
indirect forms of investment barriers. Informational efficiency, on the other hand, is proxied by the
country-level price delay measure which captures the relative speed with which each aggregate
stock market reacts to global common information. Using data from 49 national stock markets,
our empirical results show robust evidence supporting the hypothesis that the level of market
integration is significantly and positively related to the degree of informational efficiency. Further
analysis reveals that the positive association is established only in the sub-sample of emerging
42 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

stock markets. This implies that policy efforts to deepen financial integration complement those
strategies aim at promoting informational efficiency since globally integrated stock markets
respond more and faster to global market information. Such prescription is mainly for emerg-
ing market economies in which the present efforts to integrate with the world are concentrated
on the opening of their stock markets to foreign investors. In light of the policy debate on the
desirability of full financial liberalization, it is important to consider the efficiency gain on top
of economic growth when comparing with the costs of macroeconomic and financial stability
risks. This is pertinent given the potential effects of efficient stock market to the real sector, in
particularly on the allocation of investment resources.
However, the prescription for developing countries goes beyond the mere removal of statutory
foreign restrictions when devising policies toward full financial integration. Our present empirical
results highlight the importance of removing indirect investment barriers in order to reap the
associated efficiency benefit. The significance of investibility and market integration in the same
model specification indicates the complementary roles of both policies. More specifically, if a
country is to realize the efficiency-enhancing benefit of financial integration, the focus should
not be placed exclusively on decrees that remove direct investment barriers. Instead, it must
be accompanied by policy reforms to reduce implicit barriers that keep foreign investors from
investing in the domestic stock market. This is in line with the recommendation of Carrieri et al.
(2011) that improvement in corporate governance, transparency and macro institutions would
complement market liberalization policies in further integrating emerging markets. Our indirect
inference on the importance of institutional reforms for the efficient functioning of stock markets
is consistent with the findings of Morck et al. (2000), Durnev et al. (2004a) and Lim and Brooks
(2010). One specific reform highlighted by these three studies is the strengthening of private
property rights institutions in emerging market economies.
Nevertheless, we invite alternative explanations for our results, and hope that this study will
set the stage for a richer theoretical exploration on the economic forces that underlie the positive
association between market integration and informational efficiency. We stress here that the evi-
dence presented should be interpreted as identifying an association rather than causality. Given
the absence of a theoretical model on the interdependence between these two policy variables,
their relationship is obviously an empirical question and we leave it for future research to iden-
tify the causal mechanisms. The existence and direction of causality between market integration
and informational efficiency have implications for setting policy priorities. This is particularly
important for those developing countries with scarce administrative capacity and political capital.
In fact, the issue of causality is exactly analogous to the debate on the optimal sequencing of
liberalization between trade and financial openness.

Acknowledgements

The authors would like to thank anonymous referees of the journals for their con-
structive comments and suggestions that have improved the manuscript substantially. The
first author acknowledges financial support from Universiti Sains Malaysia (RU grant no.
1001/PMGT/816135).

References

Bae, K. H., Ozoguz, A., Tan, H., & Wirjanto, T. S. (2012). Do foreigners facilitate information transmission in emerging
markets? Journal of Financial Economics, 105, 209–227.
 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44 43

Bekaert, G. (1995). Market integration and investment barriers in emerging equity markets. World Bank Economic Review,
9, 75–107.
Bekaert, G., & Harvey, C. R. (2000). Foreign speculators and emerging equity markets. Journal of Finance, 55, 565–613.
Bekaert, G., Harvey, C. R., Lundblad, C. T., & Siegel, S. (2011). What segments equity markets? Review of Financial
Studies, 24, 3841–3890.
Bond, P., Edmans, A., & Goldstein I. The real effects of financial markets. Annual Review of Financial Economics, in
press, http://dx.doi.org/10.1146/annurev-financial-110311-101826
Bonfiglioli, A. (2008). Financial integration, productivity and capital accumulation. Journal of International Economics,
76, 337–355.
Bordo, M. D., & Jeanne, O. (2002). Monetary policy and asset prices: Does ‘benign neglect’ make sense? International
Finance, 5, 139–164.
Bouët, A., Berisha-Krasniqi, V., Estrades, C., & Laborde, D. (2012). Trade and investment in Latin America and Asia:
Perspectives from further integration. Journal of Policy Modeling, 34, 193–210.
Bris, A., Goetzmann, W. N., & Zhu, N. (2007). Efficiency and the bear: Short sales and markets around the world. Journal
of Finance, 62, 1029–1079.
Buch, C. M., & Pierdzioch, C. (2005). The integration of imperfect financial markets: Implications for business cycle
volatility. Journal of Policy Modeling, 27, 789–804.
Carrieri, F., Chaieb, I., & Errunza, V. R. (2011). Do implicit barriers matter for globalization? SSRN Working paper.
Available at http://dx.doi.org/10.2139/ssrn.1730206
Carrieri, F., Errunza, V. R., & Hogan, K. (2007). Characterizing world market integration through time. Journal of Financial
and Quantitative Analysis, 42, 915–940.
Chen, Q., Goldstein, I., & Jiang, W. (2007). Price informativeness and investment sensitivity to stock price. Review of
Financial Studies, 20, 619–650.
Daouk, H., & Charoenrook, A. (2005). A study of market-wide short-selling restrictions. SSRN Working paper. Available
at http://dx.doi.org/10.2139/ssrn.687562
Daouk, H., Lee, C. M. C., & Ng, D. (2006). Capital market governance: How do security laws affect market performance?
Journal of Corporate Finance, 12, 560–593.
Dow, J., & Gorton, G. (1997). Stock market efficiency and economic efficiency: Is there a connection? Journal of Finance,
52, 1087–1129.
Dow, J., Goldstein, I., & Guembel, A. (2011). Incentives for information production in markets where prices affect real
investment. Working paper. London Business School.
Durnev, A., Li, K., Morck, R., & Yeung, B. (2004). Capital markets and capital allocation: Implications for economies in
transition. Economics of Transition, 12, 593–634.
Durnev, A., Morck, R., & Yeung, B. (2004). Value-enhancing capital budgeting and firm-specific stock return variation.
Journal of Finance, 59, 65–105.
Edison, H. J., Klein, M. W., Ricci, L. A., & SlØk, T. (2004). Capital account liberalization and economic performance:
Survey and synthesis. IMF Staff Papers, 51, 220–256.
Fama, E. F. (1990). Stock returns, expected returns, and real activity. Journal of Finance, 45, 1089–1108.
Fama, E. F., & MacBeth, J. D. (1973). Risk, return and equilibrium: Empirical tests. Journal of Political Economy, 81,
607–636.
Fernandes, N., & Ferreira, M. A. (2009). Insider trading laws and stock price informativeness. Review of Financial Studies,
22, 1845–1887.
Goldstein, I., & Guembel, A. (2008). Manipulation and the allocational role of prices. Review of Economic Studies, 75,
133–164.
Henry, P. B. (2000). Stock market liberalization, economic reform, and emerging market equity prices. Journal of Finance,
55, 529–564.
Henry, P. B. (2007). Capital account liberalization: Theory, evidence, and speculation. Journal of Economic Literature,
45, 887–935.
Jin, L., & Myers, S. C. (2006). R2 around the world: New theory and new tests. Journal of Financial Economics, 79,
257–292.
Kim, Y. H. (2005). The optimal path of regional economic integration between asymmetric countries in the North East
Asia. Journal of Policy Modeling, 27, 673–687.
Korajczyk, R. A. (1996). A measure of stock market integration for developed and emerging markets. World Bank
Economic Review, 10, 267–289.
Kose, M. A., Prasad, E., Rogoff, K., & Wei, S. J. (2009). Financial globalization: A reappraisal. IMF Staff Papers, 56,
8–62.
44 C.-W. Hooy, K.-P. Lim / Journal of Policy Modeling 35 (2013) 29–44

Leduc, S., & Natal, J. M. (2011). Should central banks lean against changes in asset prices? Working paper. Federal
Reserve Bank of San Francisco.
Levine, R., & Zervos, S. (1998). Stock markets, banks, and economic growth. American Economic Review, 88, 537–558.
Li, K., Morck, R., Yang, F., & Yeung, B. (2004). Firm-specific variation and openness in emerging markets. Review of
Economics and Statistics, 86, 658–669.
Lim, K. P. (2009). The speed of stock price adjustment to market-wide information. SSRN Working paper. Available at
http://dx.doi.org/10.2139/ssrn.1412231
Lim, K. P., & Brooks, R. D. (2010). Why do emerging stock markets experience more persistent price deviations from a
random walk over time? A country-level analysis. Macroeconomic Dynamics, 14(S1), 3–41.
Lim, K. P., & Hooy, C. W. (2010). The delay of stock price adjustment to information: A country-level analysis. Economics
Bulletin, 30(2), 1609–1616.
Lim, K. P., & Kim, J. H. (2011). Trade openness and the informational efficiency of emerging stock markets. Economic
Modelling, 28, 2228–2238.
Majumder, D. (2012). Towards an efficient stock market: Empirical evidence from the Indian market. Journal of Policy
Modeling, http://dx.doi.org/10.1016/j.jpolmod.2011.08.016
Morck, R., Shleifer, A., & Vishny, R. W. (1990). The stock market and investment: Is the market a sideshow? Brookings
Papers on Economic Activity, 157–215.
Morck, R., Yeung, B., & Yu, W. (2000). The information content of stock markets: Why do emerging markets have
synchronous stock price movements? Journal of Financial Economics, 58, 215–260.
Nishiotis, G. P. (2004). Do indirect investment barriers contribute to capital market segmentation? Journal of Financial
and Quantitative Analysis, 39, 613–630.
Norén, R. (2011). Towards a more integrated, symmetric and viable EMU. Journal of Policy Modeling, 33, 821–830.
Rockinger, M., & Urga, G. (2001). A time-varying parameter model to test for predictability and integration in the stock
markets of transition economies. Journal of Business and Economic Statistics, 19, 73–84.
Roubini, N. (2006). Why central banks should burst bubbles. International Finance, 9, 87–107.
Saffi, P. A., & Sigurdsson, K. (2011). Price efficiency and short selling. Review of Financial Studies, 24, 821–852.
Schindler, M. (2009). Measuring financial integration: A new data set. IMF Staff Papers, 56, 222–238.
Schotman, P. C., & Zalewska, A. (2006). Non-synchronous trading and testing for market integration in Central European
emerging markets. Journal of Empirical Finance, 13, 462–494.
Schwert, G. W. (1990). Stock returns and real activity: A century of evidence. Journal of Finance, 45, 1237–1257.
Stehle, R. (1977). An empirical test of the alternative hypotheses of national and international pricing of risky assets.
Journal of Finance, 32, 493–502.
Wurgler, J. (2000). Financial markets and the allocation of capital. Journal of Financial Economics, 58, 187–214.