Research in International Business and Finance 41 (2017) 600–612

Contents lists available at ScienceDirect

Research in International Business and Finance

journal homepage: www.elsevier.com/locate/ribaf

The impact of the Arab Spring and the Ebola outbreak on African MARK
equity mutual fund investor decisions
⁎
Alfonso Del Giudicea, , Andrea Paltrinierib
a
Department of Economic and Business Management Science, Università Cattolica, Milano, Italy
b
Department of Economics and Statistics, Università di Udine, Italy

AR TI CLE I NF O AB S T R A CT

JEL classifications: Financial development and stock markets have been widely considered to be key factors in
G11 economic growth. Among institutional investors, mutual funds play a key role in providing
G23 financial resources to stock markets, particularly in developing countries. Different from other
Keywords: investments, mutual fund flows could be affected by retail investors’ behavior and their
Africa overreaction to specific events. We considered 78 equity mutual funds that are geographically
Ebola specialized in African countries and observed monthly flows and performance for the period of
Arab Spring 2006–2015. We find that two major events, Ebola and the Arab Spring, significantly affected the
Mutual funds
fund flows, controlling for fund performance, expenses and market returns. Retail investors over-
Performance
reacted to these major events, withdrawing their savings from the African mutual funds. This
Capital flows
result is particularly strong when connected to the media coverage of these events: the higher the
number of articles about Arab Spring and Ebola, the higher the withdrawals. These irrational
investors’ behavior damaged the funds’ managers market timing ability, and reduced the equity
capital injection into African stock markets. Our results have several implications for both
holders of frontier market mutual funds and the overall asset management industry.

Introduction

The effects of financial development on economic growth have been widely investigated in the literature since the last century,
but only a few studies have focused on emerging and frontier countries (Kar et al., 2011 ; Samargandi et al., 2015). The authors have
identified stock market development as a key variable in attracting private capital investment and promoting economic growth
(Levine and Zervos, 1993, 1998; Rousseau and Wachtel, 2000; Beck and Levine, 2003, among others). Hence, the importance of
institutional investors and mutual funds is to feed financial markets with private capital flows to support firm development,
particularly in emerging and developing countries. In fact, in recent years, retail and institutional investors have witnessed notable
growth in a specific sector of the asset management industry, involving emerging market equity mutual funds. With only a handful of
emerging market funds at the beginning of the 1990s (Kaminsky et al., 2001), there are currently hundreds of funds investing in
specific emerging market countries or areas. Such a development is related to the impressive rise of total private non-resident capital
flows from $25 billion in 1990 to an all-time high of $1.35 trillion in 2013, before falling to under $1.1 trillion in 2014 (Institute of
International Finance, 2015).
We are now seeing the same interest in African financial markets. Although later than in other emerging markets, during the last
decade, private capital flows to African countries have grown substantially (Bokpin, 2017). Before the 2000s, the development of the

⁎
Corresponding author.
E-mail address: alfonso.delgiudice@unicatt.it (A. Del Giudice).

http://dx.doi.org/10.1016/j.ribaf.2017.05.004
Received 2 May 2017; Accepted 2 May 2017
Available online 13 May 2017
0275-5319/ © 2017 Elsevier B.V. All rights reserved.
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Africa region was in an early stage, primarily because of political instability. This condition prevented capital flows from flooding into
the continent, just in the period when there was great interest in other emerging market areas (Asia Pacific and Latin America, etc.).
However, from the 2000s onward, African countries received growing interest from foreign institutional investors before the onset of
the subprime financial crisis caused these flows to retreat (IMF, 2014; Sugimoto et al., 2014). Then, in 2010, the recovery of financial
markets coupled with unprecedented monetary policies led to another significant spike in private capital flows to sub-Saharan
African countries, in particular. According to the IMF (2014), during the period 2010–2012, those inflows doubled from the period
2000–07, and they had a five-fold increase if we focus on Sub-Saharan frontier markets. Then, portfolio flows to African markets
experienced a significant retracement due to several factors, including geopolitical tensions and Ebola. Those exogenous shocks could
have pushed investors to withdraw liquidity from African mutual funds, causing a collapse in portfolios’ flows to emerging and
frontier countries.
Based on a sample of 4368 monthly observations of 78 African Equity mutual funds for the period of 2006–2015, the objective of
this paper is to assess the impact of Ebola and the Arab Spring on African mutual fund flows and performance.
Our paper is related to the extensive literature on financial markets anomalies and especially investors’ over-reactions (De Bondt
and Thaler, 1985). Using past returns, they predict price reversals in the case of a huge overshoot, providing evidence of the
directional effect of Brown and Harlow, 1988 (“large stock price movements will be followed by price reversals in the opposite direction”).
This is one of the first studies that fights the weak form of market efficiency theory (Malkiel and Fama, 1970), as all investors can
obtain abnormal returns taking advantage of long-term mispricing, in particular the negative serial correlation. Since then, several
papers have investigated over-reaction to a specific event in developed markets, focusing especially on abnormal stock returns
(Seyhun, 1990; Larson and Madura, 2003; Edmans et al., 2007).
However, to the best of our knowledge, only one paper examines over-reaction to specific events in emerging market economies
(Boubaker et al., 2015). Using data from the Egyptian stock exchange (EGX) over the period of 2003–2009, the authors investigate
short-term over-reaction to four major events, i.e., terrorist attacks, tensions in the Middle East region, the formation of new
governments, and the announcement of the privatization of state-owned enterprises (SOEs). They provide evidence of short-term
over-reaction and the leakage of information in the EGX, showing that the negative and significant abnormal returns post-event in
both terrorist attacks and tensions in the Middle East region are followed by strong price reversal.
Through a sample of 78 African mutual funds, we contribute to the literature investigating the over-reaction to specific major
events, namely Ebola and the Arab Spring. Differently from Boubaker et al. (2015), we decided to analyze the impact on African stock
markets rather than a single stock index, as Ebola and the Arab Spring affect more than one country. Then, we focus both on the
performance and the flows of the equity mutual funds to the African stock markets. Mutual funds are the favorite investment vehicle
of retail investors. The literature has widely reported that retail investors in developed markets do not behave rationally (Barberis
et al., 2001; Barberis and Thaler, 2003; Subrahmanyam, 2007 among others). Also in this case, only a few papers focus on emerging
markets, especially on specific countries such as Turkey (Tekce and Yilmaz, 2015; Tekce et al., 2016). In particular, Tekce et al.
(2016), analyzing Istanbul Stock Exchange transactions, found that behavioral biases are common among Turkish retail investors. to
the best of our knowledge, no study has examined the impact of specific major events on a large sample of mutual funds covering
different emerging and frontier countries in Africa.
Our main hypothesis is that retail investors in African funds have been influenced by the huge media coverage of these events,
which were (and still are) related to a portion of the entire continent. It is well known in the literature that media coverage of major
events could influence investors’ behavior (Engelberg and Parsons, 2011). Moreover, according to Tversky and Kahneman (1971),
people make erroneous judgments about the probability of an event under uncertainty. This is a particular type of bias that the
authors call representativeness heuristic, happening when people, in general, use a rule of thumb strategy to solve a specific problem
or during the process of decision making. That is why we suppose that retail investors, affected by representativeness heuristic,
generalize major events such as Arab Spring and Ebola that affect just a part of Africa to the whole continent. We find that both Ebola
and the Arab Spring have had a negative and statistically significant impact on monthly net flows, as expected. This result is
particularly strong when connected to the media coverage of these events: the higher the number of articles about Arab Spring and
Ebola, the higher the withdrawals. These exogenous shocks have caused investors over-reaction, affecting asset managers’ market
timing. They also have reduced financial resources that, through the capital flows, could have easily fed financial markets and
supported firm development, eventually promoting economic growth.
These results confirm retail investors’ biases mainly for the Ebola shock, considering that in 2014, Ebola actually affected a
geographical area of less than 10 million people (World Health Organization, 2015), representing 5% of the entire continent’s GDP
(World Bank, 2014). The overall African continent’s reputation as an investment opportunity area was damaged by the Ebola
outbreak more than the political instability due to the Arab Spring insurgency in North Africa. Our results clearly show that such a
confined exogenous shock has reduced liquidity to all African mutual funds industries, although they are able to obtain good
performance. As described above, the diminishing of mutual funds’ capital flows could severely affect economic growth.
From the policy implication standpoint, we can argue that the media coverage of such events plays a critical role in influencing
investors’ behavior. Excessively alarmist news could have induced retail investors to withdraw their savings from African mutual
funds, thus reducing financial resources to the entire continent.

Literature review and hypothesis development

The first author analyzing the effects of financial development on economic growth and providing a theoretical framework for our
work was Shumpeter (1934), followed by McKinnon (1973), while more recent empirical studies have shown a strong correlation

601
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

between those variables (King and Levine, 1993, among others). Within the financial system, Levine and Zervos (1996) discussed the
importance of stock markets in promoting economic growth. Focusing on developed markets, they show that the liquidity, size and
openness of stock markets are fundamental factors in explaining economic growth. On the supply side, their function is important to
help firms raise funds. Demigurc-Kunt and Huizinga (1993) suggest that the more stock markets are developed, the more the cost of
capital is likely to plummet, causing firms to list on the market and investors to provide financial resources to them, thus increasing
overall liquidity. Cho (1986) argues that the role of financial markets is important for financing riskier ventures and particularly in
developing countries such as African ones; they become important to attract capital from foreign investors because of their role in
pricing risk. Kenny and Moss (1998), investigating the role of stock markets in Africa, suggest that although the stock markets will
remain small and illiquid, they will be increasingly important for African economies (so called “Emerging Lions”).
Hence, it is important to analyze all the factors behind capital flows in emerging and frontier markets. In general, they are divided
into two main categories: the so-called “push” factors, related to global conditions, and the “pull” factors, related to local economies
or countries (Chuhan et al., 1993; Fernandez-Arias and Montiel, 1996). The most important push factors are interest rates in
industrialized economies and the opportunity for institutional investors to diversify their portfolios (Chuhan et al., 1993). Thanks to
globalization, improved communication and information technology, mutual funds and pension funds have flooded emerging and
frontier markets during recent years, as we described above (Institute of International Finance, 2015). Considering the pull factors,
we must highlight the strength of developing countries’ economies, firms’ profitability and the overall risk environment
(Bhattacharya et al., 1996).
Kenny and Moss (1998) show the importance for stock markets in African countries to attract capital flows. The authors analyze
the development of the stock exchanges in Africa, suggesting that although they are very small and illiquid markets, they will allow
investors to finance firms through their portfolios’ strategies, promoting the development of the countries. More recently, Moss et al.
(2007) addressed the question of investment in Sub-Saharan listed securities, examining the characteristics of 15 stock exchanges and
the portfolios’ flows from regional and global emerging market funds. They find that African countries, apart from South Africa, have
not received equity flows, primarily due to the small size and low levels of liquidity of African stock exchanges rather than to a lack of
information, perceptions of excessive risk or other unknown variables. They also show that there were very few African mutual funds
at the time (in 2006, only seven funds were active), and they were performing poorly and having difficulty attracting investors.
Moreover, large global emerging market funds did not overweight Africa in their asset allocation because of small stock market size
and liquidity. In fact, as Acquaah (2015) shows by using Ghana as a case study, sub-Saharan Africa economies have stock markets to
attract foreign capital, but there are several factors discouraging the listing of local firms, such as extensive information and
disclosure cost requirements and the loss of ownership and control.
Kaminsky et al. (2001), focusing specifically on emerging market mutual funds, explain that little is known about their
asset allocation and investment strategies, although they are key contributors to financial market globalization and are among the
main sources of emerging economies’ capital flows. They show that mutual funds’ flows to emerging markets are not stable because of
large redemptions and injections, and in particular, withdrawals from developing countries during financial crises were large. This
behavior is consistent with the financial contagion phenomenon.
From the literature above, we could easily say that mutual funds, through the stock market channel, are a key factor in providing
capital flows to emerging markets, which, in turn, promote countries’ development. Kaminsky et al. (2001) argued that the impact of
specific country/region risks on overall mutual funds’ behavior could be far more relevant for emerging markets. With this paper, we
try to fill part of that gap, assessing the impact of specific major events (the Arab Spring and Ebola outbreak) on African mutual fund
flows. As mutual funds are the favorite investment vehicle of retail investors and the literature has widely reported that individuals do
not behave rationally (Barberis et al., 2001; Barberis and Thaler, 2003; Subrahmanyam, 2007; Tekce and Yilmaz, 2015; Tekce et al.,
2016), we would like to contribute to the behavioral finance field of research, investigating the investors’ overreaction to specific
events (Seyhun, 1990; Larson and Madura, 2003; Edmans et al., 2007; Boubaker et al., 2015) and their representativeness heuristic
(Tversky and Kahneman, 1971).
Moreover, because the investment vehicles are emerging and frontier market mutual funds, we try to expand the literature on
their performance, as well. In fact, despite their growing popularity, research on their asset allocation, investment strategies and
performance has been rather poor. The vast majority of the studies have focused on the US mutual fund industry and, in general,
show the inability of mutual funds to beat the benchmark after all fees are considered (see Malkiel, 1995; Carhart, 1997; French,
2008; Droms, 2006, for an analytical review on mutual fund performance persistence). There are also many studies concerning
European industry that analyze individual countries’ mutual funds without reaching a comprehensive conclusion (Dermine and
Roller, 1992; Blake and Timmermann, 1998; Fernandez et al., 2008). Otten and Bams (2002), analyzing 506 mutual funds among five
different European countries, found that mutual fund managers generate an alpha that is not significantly different from zero. If we
then do not consider management fees, four countries produce an alpha that is significantly positive, showing their ability to pick
stocks.
In emerging markets, instead, there has been a lack of literature. As they are perceived to be less efficient than developed markets,
the opportunities to generate abnormal returns should be higher, according to Huij and Post (2011). Khorana et al. (2005) and
Ferreira et al. (2006) find that risk-adjusted performance is affected by variables such as legal institutions and the development of the
capital market, suggesting a presumable underperformance of emerging market mutual funds. In general, it seems that global mutual
funds investing in emerging markets obtain over-performance. Borensztein and Gelos (2000), investigating a sample of 309 emerging
market mutual funds during the period of 1996–1999, show that they are able to withdraw money one month before the crisis, thus
having optimal market timing. Huij and Post (2011) show that emerging market mutual funds based in the US are able to generate
returns large enough to cover all of their expenses and, in general, perform better than mutual funds investing in the US. Bialkowky

602
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

et al. (2013), examining the behavior of Turkish institutional funds, hybrid funds and foreign funds investing in Turkish stocks during
the Ramadan period, find that they obtain risk-adjusted performance higher than during the rest of the year. Only a few studies have
examined the behavior of local mutual funds investing in emerging markets, which, theoretically, should perform better because of
information advantage (Brennan and Cao, 1997; Hau, 2001) and knowledge of specific risk factors (Harvey, 1995). Bialkowsky and
Otten (2011), investigating a sample of 140 Polish mutual funds, show that asset managers are not able to add value and generate a
negative alpha.
Our paper is the first to analyze equity mutual funds investing in Africa, including sub-Saharan countries, providing evidence of
the performance of a sample of frontier market mutual funds (considering that currently in Africa, the only emerging market country
is South Africa, according to Morgan Stanley Capital International).
Following the literature, our hypotheses are that (1) African mutual fund investors over-react to specific major events, even if
related to a small portion of the whole continent, withdrawing their money from the funds; (2) the irrational withdrawals are
connected to the media coverage of these major events; and (3) unexpected withdrawals reduce fund managers’ market timing ability
to generate extra performance.

3. Data, sample and descriptive statistics

We collected data of all 149 Open-end Equity Mutual Funds for the period of 2006–2015 focusing on Africa from Thomson
DataStream. We found that African funds were mainly classified within the “Equity Emerging Market Others” category (138 funds) and
the Equity Emerging Markets Global (4), Equity Middle East North Africa (3), Equity Others (3) and Equity South Africa (1) categories.
From the sample of 149 mutual funds, we rejected those without an International Securities Identification Number (ISIN), leaving
137 funds. We considered only the funds with available data on Prices (data type: P), Total Expense Ratio (data type: TER) and Total
Net Asset (data type: TNA), so we rejected 46 mutual funds, leaving 91. Finally, we rejected those funds with less than two years of
observations because of unstable and unreliable data for costs and performance.
At the end of the process, we had a sample of 78 mutual funds with 4368 monthly observations (see Table 1).
Table 1 shows that the oldest mutual fund was established in 2006, while the most recent was established in 2013. The inception
of these funds witnesses the recent interest of institutional investors toward African stock markets, established less than 10 years ago.
In particular, after the subprime financial crisis in 2008, there was an explosion of African mutual funds, most of them established in
Europe (53), under Luxembourg law (43). They have smaller assets under management and higher total expense ratios, compared to
their non-African established peers (see Table 2).
Examining this table, we see that during the sample period, monthly net flows are relatively stable for Europe and the US, positive
for the UK and negative for the other regions. Dividing monthly net flows per year over the period 2006–2015, we note that both the
subprime financial crisis in 2008 and the European sovereign debt crisis in 2011 affected their trend. They also had negative values in
2014 and 2015 (see Table 3). Analyzing the hosting regions, we see that financial crises had an impact on both US and European
mutual funds (see Table 4), while Japan and Taiwan have always experienced negative monthly net flows.
This pattern is likely linked to the different economic cycles and, specifically, to a recovery in the developed financial markets
after the crisis periods. To study this relationship, we ran the correlation between the local market index return of the mutual fund
hosting country and the net flows, showing a positive value (0.1808). We also studied the correlation between net flows’ dynamic and
returns of African stock markets, through Thomson Reuters Africa Index (data type: XAFFLD), finding again a positive number
(0.2034). Because monthly net flows could be affected by the fund performance, we examined the correlation between those two
variables, using the performance net of the risk-free rate. Again, here, we found a positive correlation (0.2467).
Therefore, the dynamic of monthly net flows is affected by both the returns of stock markets and the fund performance. Dividing
this analysis per year, we found some anomalies, however (see Table 5).
During the sample period, we observed several years of funds’ underperformance with respect to the Africa Index: 2008, 2011,
2014 and 2015. While 2008 is a special year for all of the asset management industry because of the subprime financial crisis, the
other three years are the only ones showing the underperformance of the fund manager. To more closely examine fund managers in
the next section, we estimated funds’ performance according to a consolidated model and tried to connect fund flows with both
endogenous and exogenous variables.

Table 1
African Equity Mutual Fund Nationality and birth date.

Birth Date Funds SGR Region

2006 1 UK
2007 1 South Africa
2008 14 Europe (8), US (3), Japan (2), Taiwan (1)
2009 3 US
2010 22 Europe (13), US (5), UK (2), Japan (1), South Africa (1)
2011 6 Europe (4), US (2)
2012 20 Europe (18), US (2)
2013 11 Europe (10), US (1)

78 Europe, 53; US, 16; Japan, 3, UK, 3; South Africa, 2; Taiwan, 1

603
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 2
African Equity Mutual Fund sample descriptive statistics per region.

Region N. Obs. AUM TER Net Flows

Europe 2627 $13.03 2.44% $0.00

US 998 $27.44 1.90% $0.02
UK 236 $24.52 1.86% $0.13
Japan 234 $91.69 1.58% −$0.49
South Africa 156 $83.19 2.62% −$0.62
Taiwan 83 $38.96 2.42% −$0.06

Total 4334 $24.24 2.22% −$0.05

Observations are on a monthly basis. AUM is the average asset under management in $.000; TER is the total expense ratio; Net Flows is the average amount of cash
inflow or outflow of mutual funds in $.000.

Table 3
Net Flows descriptive statistics per year. Net flows is the average amount of cash inflow or outflow of mutual funds in $.000.

Year N. obs. Net flows mean Net flows median

2006 11 $0.6806 $1.0850

2007 15 $3.1404 $0.5530
2008 64 −$4.7823 −$0.1150
2009 215 $0.2019 $0.1101
2010 323 $3.4648 $0.5361
2011 516 −$1.5145 −$0.0403
2012 691 $0.1656 $0.0466
2013 851 $0.0480 $0.0359
2014 936 −$0.2201 −$0.0001
2015 702 −$0.4247 −$0.0764

Total 4324 −$0.0507 $0.0001

4. Empirical methodology

To better understand the dynamics of the funds’ performance, we focus our analysis on measuring African fund managers’ ability
to create wealth, both considering stock picking ability and market timing of the investment. We use a well-known methodology to
capture both effects, provided by Treynor and Mazuy (1966).
(Ri, t − Rf , t ) = αi + βi (Rm, t − Rf , t ) + γi (Rm, t − Rf , t )2 + εi, t α

where:
1-Ri,t is the funds’ return at time t;
-Rf,t is the risk-free rate1 at time t;
-Rm,t is the market return2 at time t;
2-εm,t is the white noise error at time t;
Funds’ return, netted by the risk-free rate, should be explained only by β, which represents the coefficient related to the market
premium. Theoretically speaking, all of the other coefficients should not be relevant. If they show positive or negative values that are
statistically significant, it would indicate a specific fund manager’s ability to create/destroy wealth. In particular, a positive Alpha
shows the fund manager’s stock picking ability, while a positive Gamma is a sign of good market timing in selecting securities.
Because our expectation, following the literature, is to obtain a negative Alpha because of a fund manager’s inability in security
selection and a negative Gamma because of bad market timing, the second step of our analysis is to investigate mutual funds’ net
flows. In fact, while the choice of securities is actually linked to the ability of the asset manager, the market timing could be easily
affected by investor withdrawals or injections of money.
To address these factors, we use monthly net flows (NF) as a dependent variable and both endogenous and exogenous variables as
independent ones.
NFi = a 1Rf + a 2EPi + a 3Rm + a 4Age + a 5 TER + a 6Arab Spring + a 7Ebola + εi

where:
-Rf is the monthly risk-free rate;
-EPi is the monthly extra performance of the fund computed as follows:

1
We used a global interest rate (US Gov. Bond 10 yrs, Thomson Datastream code: GBUS10Y).
2
We used Thomson Reuters African Index with returns in US $ (Thomson Datastream code: XAFFLD$).

604
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 4
Net Flows per Year in Europe and in US African Equity Mutual Funds. Net flows is the average amount of cash inflow or outflow of mutual
funds in $.000.

Year N. obs Net flows mean Net flows median

Europe
2008 8 −$0.7174 $0.0000
2009 61 −$0.1948 $0.0995
2010 102 $5.3682 $1.1254
2011 142 −$1.8940 −$0.2170
2012 159 $0.7962 $0.2000
2013 183 $0.0799 $0.1000
2014 192 −$0.4137 $0.0003
2015 144 −$0.8054 −$0.1698

Total 991 $0.2087 $0.0199

Year N. obs Net flows mean Net flows median

US
2008 21 −$3.1175 −$0.0959
2009 94 $0.9554 $0.1279
2010 153 $2.4870 $0.3997
2011 266 −$0.8361 −$0.0113
2012 424 −$0.0016 $0.0293
2013 560 $0.2387 $0.0249
2014 636 −$0.0411 $0.0000
2015 477 −$0.2839 −$0.0427

Total 2631 $0.0585 $0.0001

Table 5
Fund performance and net flows per year.

Year Local Index African Index Fund Performance Net Flow

2006 1.74% 0.45% 1.01% $0.68

2007 0.20% 2.05% 3.23% $3.14
2008 −8.98% −6.53% −8.42% −$4.78
2009 3.13% 2.95% 4.66% $0.20
2010 0.64% 2.31% 2.78% $3.46
2011 −2.29% −1.90% −2.30% −$1.51
2012 1.12% 1.00% 1.41% $0.17
2013 1.78% −0.77% 0.55% $0.05
2014 −0.64% 0.33% −0.80% −$0.22
2015 −1.72% −2.18% −2.29% −$0.42

N. Obs. 4368 4368 4368 4324

Local Index (Thomson Datastream, datatype: LI) is the benchmark local return index for an equity investment; African Index (Thomson Datastream, code XAFFLD$) is
the benchmark African return index for an equity investment; Fund Performance are obtained on monthly basis; Net flows is the average amount of cash inflow or
outflow of mutual funds in $.000.

⎛1 T ⎞2
EPi = αi + γi ⎜⎜ ∑ Rm, t − Rf , t ⎟⎟
⎝ T t =1 ⎠

where the parameters α and γ are computed according the Trenynor–Mazuy model;
-Rm is the monthly return of the Thomson Reuters African market index;
-Age is the age of the fund;
-TER is the total expense ratio of the fund;
-Arab Spring is a dummy variable that takes the value 1 in the period of Dec. 18, 2010–Oct. 20, 2011, and 0 otherwise;
-Ebola Outbreak is a dummy variable that takes the value 1 in the period from May 2014 to May 2015, and 0 otherwise.
We control for a set of variables affecting net flows considering the previous literature. We expect that the total costs of the funds
have a negative impact on monthly net flows, while the age of the funds, market returns and the total extra performance have a
positive impact on total flows. We also expect a negative sign of the risk-free rate, in terms of cost opportunity. The higher the risk-
free rate, the higher the market competition.
We also identify a couple of variables that are more event-specific. We used the Bloomberg Professional Service to collect data on

605
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 6
Treynor–Mazuy model: stock picking and market timing ability.

I II III IV V VI VII
Entire Sample Europe Japan South Africa Taiwan UK USA

Constant −0.0001 −0.0014 −0.0037 0.0019 0.0034 0.0057** 0.0029

[0.0008] [0.0009] [0.0036] [0.0044] [0.0041] [0.0023] [0.0018]
β 0.7711*** 0.6719*** 0.8007*** 0.6367*** 0.8570*** 0.9426*** 0.9425***
[0.0277] [0.0144] [0.0448] [0.0596] [0.0580] [0.0353] [0.0944]
γ −0.0282 −0.3215** 0.5799 −1.1997* 0.2146 −0.9978** 0.6452
[0.2911] [0.1575] [0.6841] [0.6232] [0.6242] [0.3939] [0.9487]

Adjusted R2 0.4598 0.5071 0.6664 0.5359 0.7779 0.8341 0.3652

Observations 4368 2658 234 156 83 236 1001

***, **, and * indicate significance at the 1, 5, and 10 percent levels.

articles from several dozen newspapers3 having in the headline the word “Ebola” (or “Viral-Ill”) and “Arab Spring”. In the first step,
we searched with the ticker “NT” for daily news from May 2014 to December 2015 on Ebola. Then, we applied the same research
criteria for Arab Spring, for the period of December 2010–October 2011. In this way, we were able to collect all news not only from
Bloomberg, but from the major info-providers and international newspapers.
These variables are defined as follows:

– Arab Spring Media Coverage, which is the natural logarithm of the monthly sum of Arab Spring daily total citations;
– Ebola Media Coverage, which represents the natural logarithm of the monthly sum of Ebola daily total citations.

In a separate model, we used these two variables instead of the dummy variables “Arab Spring” and “Ebola”. This approach helps
us to better identify the investors’ reaction to the media coverage of these major events. We expect a negative impact of the Arab
Spring and Ebola outbreak on net flows of the funds, causing large withdrawals and, eventually, very poor market timing of the fund
manager.

NFi = a 1Rf + a 2EPi + a 3Rm + a 4Age + a 5 TER + a 6Arab SpringMC + a 7EbolaMC + εi

5. Empirical results

5.1. Stock picking and market timing ability

First, we investigate fund managers’ stock picking ability using the Treynor and Mazuy (1966) model. We use as dependent
variables the equity risk premium (ERP) and its squared value to capture market timing. We calculate ERP as the return of the market
netted by the risk-free rate. Table 6 shows our results.
In the table, we report the regression for the entire sample and for the funds’ geographic origin.
For the entire sample, we find that the constant of the model (the funds’ Alpha in the Treynor–Mazuy model) is not statistically
significant, meaning no picking ability of the managers. Only UK funds show a positive and significant Alpha, indicating the
managers’ ability to create extra performance through stock picking.
If we look at the ERP coefficient (the Beta in the Treynor–Mazuy model), for the entire sample, we obtain a positive relationship
between funds’ performance and market returns as expected, considering previous literature. This relationship is positive and
statistically significant for the entire sample. It also holds considering funds’ origin.
As for market timing, the ERP2 coefficient (the Gamma of the Treynor–Mazuy model) is negative and not statistically significant
for the entire sample. However, if we focus on European, South African and UK funds, we find a negative and statistically significant
coefficient. This finding means that funds’ managers destroyed wealth because they had no market timing ability.
In summary, we find no extra performance ability for the African equity mutual funds of the sample, in terms of both market
timing and stock picking ability. If we consider the hosting country, we find that UK funds show a relevant stock picking ability but
negative market timing. European and South African mutual funds have negative and statistically significant market timing, while the
stock picking ability is not statistically significant. While we cannot draw any conclusion for the Taiwanese fund in the sample, US
and Japan funds do not show any stock picking or market timing ability.
While the ability and knowledge of the single asset manager is the most important factor impacting the stock picking ability, there
are different drivers of the market timing. Retail investors in mutual funds could withdraw their money at any moment; therefore, it is
important to analyze funds’ net flow dynamics.

3
According to Bloomberg Professional Service description, the ticker “NT” “displays news stories from several dozen of the highest quality sources, such as
Bloomberg and the New York Times”.

606
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 7
Arab Spring and Net Flows.

I II III IV V VI VII
Arab Spring Entire Sample Europe Japan South Africa Taiwan UK USA

0 −0.0007356 0.1125469 −2.375686 −0.488178 −0.1282775 −0.2392079 0.3712028

1 −0.4655933 −0.4877392 −1.607815 2.415232 0.3409273 0.2556818 −0.8772992

diff 0.46486* 0.60029*** −0.76787 −2.90341 −0.469207 −0.494887 1.24850***

st. err. [0.32624] [0.25127] [3.81135] [4.9853] [1.2999] [0.83030] [0.53028]

Observations 4324 1942 192 128 69 196 783

Net flows is the average amount of cash inflow or outflow of mutual funds in $.000. Arab Spring is a dummy variable that takes value 1 in the period Dec. 18,
2010–Oct. 20, 2011, and zero otherwise. ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively, of the t-test for the difference in means (t-
test) between the two groups of each region.

5.2. Arab Spring, Ebola and monthly net flows

Our hypothesis is that huge investors’ withdrawals are an important driver in explaining the underperformance of the funds
compared to the African Index during 2008, 2011, 2014 and 2015. However, while in 2008, the Lehman Brothers and subprime
financial crisis severely affected the global asset management industry and African markets were severely hit, as well, in 2011, 2014
and 2015, there were exogenous shocks undermining the African economy and stock markets: namely, the Arab Spring and Ebola.
Therefore, we test whether these exogenous factors have affected mutual funds’ negative net flows.
The correlation between the Arab Spring and Net Flow variables is negative (−0.0203); the correlation between Ebola and Net
Flow is also negative (−0.0132). The t-test analysis confirms that both shocks had a huge negative impact on net flows, causing their
collapse after the event. The results show a 90% statistical significance comparing net flows before and after the event (see Tables 7
and 8). If we run the t-test between the Arab Spring and Net Flows, considering the different geographical areas of the funds, we
obtain statistically significant results only for the US and Europe. Therefore, mutual funds established in those areas experienced a
significant reduction in the monthly net assets under management during the Arab Spring. We have the same results analyzing the
Ebola effect for US and EU mutual funds. To obtain more robust results, we run the correlation between funds’ net flows, extra-
performance and the variables related to Arab Spring and Ebola news, and we confirm the strong negative correlation (Table 9).
Because our objective is to understand the dynamic of monthly net flows, we combine the endogenous variables derived from the
standard models with exogenous factors such as unexpected shocks. In Tables 10 and 11, we regress monthly net flows on
endogenous variables, separately addressing Ebola and the Arab Spring. In the first table (Table 10), we consider the dummy
variables, while in the second one (Table 11), we run the regression with the media coverage variables.
Looking at the results, we find that both Ebola and the Arab Spring had a negative and statistically significant impact on monthly
net flows, as expected. In particular, the media coverage model shows a stronger negative and statistically significant impact of those
events on African mutual funds’ net flows (see Table 11). As expected, fund returns show a positive and statistically significant
coefficient, causing new flows to increase the overall assets under management, and in addition, market return also positively affects
overall flows. The total expense ratio has no impact on monthly net flows, while the risk-free rate shows a negative impact due to
higher opportunity costs.
These results confirm our main hypothesis: investors overreact to these negative major events, even if related to a small portion of
the continent, withdrawing their savings from the funds. Media coverage plays an important role: the higher the number of articles
about Arab Spring and Ebola, the higher the withdrawals. These irrational withdrawals could have damaged the managers’ ability to
generate extra performance, reducing their market timing ability (i.e., the Gamma in the Treynor–Mazuy model).
Analyzing the specific events, the Ebola outbreak was hyped by the media as an extreme event, far more than the Arab Spring.
Although Ebola was confined to a very limited geographical area in West Africa, the entire continent was incorrectly considered at

Table 8
Ebola and Net Flows.

I II III IV V VI VII
Entire Sample Europe Japan South Africa Taiwan UK USA

0 0.0020312 0.1247716 −2.60403 0.1545235 −0.0072942 0.1859036 0.3424077

1 −0.2096808 −0.1283938 −0.6017949 −1.195515 −0.3734 0.0883436 −0.2947115

diff 0.2531* 0.253434** −2.002996 1.35004 0.36616 −0.09756 0.63711***

St. err. [0.20863] [0.14348] [3.19064] [4.17595] [1.2132] [0.70041] [0.39456]

Observations 3859 2394 198 132 71 202 862

Net flows is the average amount of cash inflow or outflow of mutual funds in $.000. Ebola is a dummy variable that takes value 1 in the period May 1, 2014–May 1,
2015, and zero otherwise. ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively, of the t-test for the difference in means (t-test) between the
two groups of each region.

607
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 9
Correlations.

Net Flows EP Arab Spring News Ebola News

Net Flows 1.0000

EP 0.2467 1.0000
Arab Spring_MC −0.0865 −0.1288 1.0000
Ebola_MC −0.0207 −0.1140 −0.0761 1.0000

Net flows is the average amount of cash inflow or outflow of mutual funds in $.000. EP is the monthly fund extra performance according to the Treynor–Mazuy model.
Arab Spring_MC is the natural logarithm of the monthly sum of Arab Spring total citations (Bloomberg, ticker: NT). Ebola_MC is the natural logarithm of the monthly
sum of Ebola total citations (Bloomberg, ticker: NT).

risk. Both these exogenous shocks caused investors to withdraw money from funds because their scarce ability to estimate the
consequences of these events (representativeness heuristic, Tversky and Kahneman, 1971).

6. Robustness test

To test the robustness of our results, we also use the model from Henriksson and Merton (1981) to analyze both stock picking and
market timing ability. While Treynor and Mazuy used a magnitude timer approach in measuring timing ability, estimating the
coefficient of a squared parameter, Heriksson and Merton provided an alternative model using a direction timer approach through a
combination of two linear models.
Portfolio beta changes according to the expectations of market excess return:
max[0, −(Rm − Rf )]
β HM = βUP + (βUP − βDOWN )
Rm − Rf

-Rf,t is the risk-free rate at time t;

-Rm,t is the market return at time t;
Managers’ ability to generate excess return is estimated as a stock picking component α and a market timing factor that estimates
the ability to follow market direction.

(Rp, t − Rf , t ) = αpHM + βp (Rm, t − Rf , t ) + γpHM max[0, −(Rm, t − Rf , t )]

Analyzing stock picking and market timing ability (see Table 12), we find confirmation that Alpha, the constant of the model, and

Table 10
Multivariate regression with dummy variables.

Model 1 Model 2

Constant 0.0001 0.0067***

[0.0005] [0.0012]
Rf 0.3084 −0.9625***
[0.1910] [0.3104]
EP 0.0257*** 0.0225***
[0.0018] [0.0059]
Rm 0.0040** 0.0035
[0.0019] [0.0026]
Age −0.0001** −0.0001**
[0.0001] [0.0001]
TER 0.0028 −0.0034
[0.0085] [0.0054]
Arab Spring −0.0007**
[0.0003]
Ebola −0.0002*
[0.0001]

Adjusted R2 0.1128 0.1118

Observations 3386 3103

**, **, and * indicate significance at the 1, 5, and 10 percent levels.

Net Flow is the average amount of cash inflow or outflow of mutual funds in $.000. (in $.000) and is
the independent variable. Rf is the monthly risk-free rate (US Gov. Bond 10 yrs, Thomson Datastream
code: GBUS10Y); EP is the monthly fund extra performance according to the Treynor–Mazuy model;
Rm is the African index return (Thomson Datastream code: XAFFLD$); Age is the age of the fund (in
months); TER is the total expense ratio; Arab Spring is a dummy variable that takes the value 1 in the
period Dec. 18, 2010–Oct. 20, 2011, and zero otherwise; Ebola is a dummy variable that takes the
value 1 in the period from May 2014 to May 2015 and zero otherwise.

608
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 11
Multivariate regression with news variables.

Model 1 Model 2

Constant 0.0066 0.0021*

[0.0077] [0.0006]
Rf −0.0001 −0.0006***
[0.0016] [0.0002]
EP 0.0385*** 0.0072***
[0.0116] [0.0015]
Rm −0.0252* 0.0051***
[0.0150] [0.0012]
Age 0.0000 −0.0000
[0.0000] [0.0000]
TER 0.0423 0.0044
[0.0409] [0.0036]
Arab Spring_MC −0.0014***
[0.0005]
Ebola_MC −0.0002***
[0.0000]

Adjusted R2 0.1094 0.1043

Observations 3386 3103

***, **, and * indicate significance at the 1, 5, and 10 percent levels.

Net Flow is the average amount of cash inflow or outflow of mutual funds in $.000. (in $.000) and is the
independent variable. Rf is the monthly risk-free rate (US Gov. Bond 10 yrs, Thomson Datastream code:
GBUS10Y); EP is the monthly fund extra performance according to the Treynor–Mazuy model; Rm is the
African index return (Thomson Datastream code: XAFFLD$); Age is the age of the fund (in months); TER is
the total expense ratio; Arab Spring_MC is the natural logarithm of the monthly sum of Arab Spring total
citations (Bloomberg, ticker: NT). Ebola_MC is the natural logarithm of the monthly sum of Ebola total
citations (Bloomberg, ticker: NT).

Table 12
Heriksson and Merton model: stock picking and market timing ability.

I II III IV V VI VII
Entire Sample Europe Japan South Africa Taiwan UK USA

Constant −0.0011 −0.0018 −0.0042 0.0045 0.0026 0.0076** 0.0001

[0.0016] [0.0013] [0.0050] [0.0064] [0.0056] [0.0035] [0.0047]
β 0.7918*** 0.6611*** 0.8502*** 0.5011*** 0.8908*** 0.8335*** 1.0492***
[0.0654] [0.0288] [0.0843] [0.1170] [0.1008] [0.0709] [0.2297]
γ 0.0368 −0.0320 0.1240 −0.3104 0.0679 −0.2483* 0.2185
[0.0848] [0.0479] [0.1908] [0.2180] [0.1860] [0.1317] [0.2844]

Adjusted R2 0.4599 0.5061 0.6626 0.5247 0.7779 0.8287 0.3653

Observations 4368 2658 234 156 83 236 1001

***, **, and * indicate significance at the 1, 5, and 10 percent levels.

Table 13
Correlations.

Net Flows EP Arab Spring News Ebola News

Net Flows 1.0000

EP 0.2104 1.0000
Arab Spring News −0.0865 −0.1288 1.0000
Ebola Media Coverage −0.0207 −0.0407 −0.0761 1.0000

Net flows is the average amount of cash inflow or outflow of mutual funds in $.000. EP is the monthly fund extra performance according to the Heriksson and Merton
model. Arab Spring_MC is the natural logarithm of the monthly sum of Arab Spring total citations (Bloomberg, ticker: NT). Ebola_MC is the natural logarithm of the
monthly sum of Ebola total citations (Bloomberg, ticker: NT).

Gamma are negative but not statistically significant, so we cannot conclude that asset managers are capable of achieving extra
performance. However, we can reason that asset managers, because of strong withdrawals from the funds, face challenges in
portfolios’ rebalancing. This could lead to difficulties in generating over-performance and positive Alpha.
Running the same test with subsamples considering different asset management company hosting countries, we again obtain
different results, supporting the findings of the Treynor and Mazuy model.
The table shows that funds’ asset managers do not show either stock picking ability or market timing. The only funds with positive

609
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Table 14
Multivariate regression with dummy variables.

Model 1 Model 2

Constant 0.0005 0.0000

[0.0005] [0.0007]
Rf 0.5809*** 0.1708
[0.1968] [0.2211]
EP −0.0530** 0.0030
[0.0275] [0.0355]
Rm 0.0233*** 0.0166***
[0.0022] [0.0064]
Age −0.0001* −0.0001
[0.0001] [0.0001]
TER 0.0041 −0.0012
[0.0088] [0.0055]
Arab Spring −0.0014**
[0.0003]
Ebola −0.0003*
[0.0001]

Adjusted R2 0.0586 0.0542

Observations 3386 3103

***, **, and * indicate significance at the 1, 5, and 10 percent levels.

Net Flow is the average amount of cash inflow or outflow of mutual funds in $.000. (in $.000) and
is the independent variable. Rf is the monthly risk-free rate (US Gov. Bond 10 yrs, Thomson
Datastream code: GBUS10Y); EP is the monthly fund extra performance according to the Heriksson-
Merton model; Rm is the African index return (Thomson Datastream code: XAFFLD$); Age is the
age of the fund (in months); TER is the total expense ratio; Arab Spring is a dummy variable that
takes the value 1 in the period Dec. 18, 2010–Oct. 20, 2011, and zero otherwise; Ebola is a dummy
variable that takes the value 1 in the period from May 2014 to May 2015 and zero otherwise.

Table 15
Multivariate regression with news variables.

Model 1 Model 2

Constant 0.0032 0.0006

[0.0097] [0.0008]
Rf 0.0021 0.0000
[0.0016] [0.0000]
EP 0.0162 0.0484
[0.0208] [0.0445]
Rm −0.0261 0.0061***
[0.0220] [0.0022]
Age 0.0000 −0.0000
[0.0000] [0.0000]
TER 0.0375 0.0040
[0.0412] [0.0046]
Arab Spring MC −0.0015***
[0.0006]
Ebola_MC −0.0002***
[0.0000]

Adjusted R2 0.0730 0.0825

Observations 3386 3103

***, **, and * indicate significance at the 1, 5, and 10 percent levels.

Net Flow is the average amount of cash inflow or outflow of mutual funds in $.000. (in $.000) and is the
independent variable. Rf is the monthly risk-free rate (US Gov. Bond 10 yrs, Thomson Datastream code:
GBUS10Y); EP is the monthly fund extra performance according to the Treynor–Mazuy model; Rm is the
African index return (Thomson Datastream code: XAFFLD$); Age is the age of the fund (in months); TER is
the total expense ratio; Arab Spring_MC is the natural logarithm of the monthly sum of Arab Spring total
citations (Bloomberg, ticker: NT). Ebola_MC is the natural logarithm of the monthly sum of Ebola total
citations (Bloomberg, ticker: NT).

and statistically significant Alpha are the UK ones, which also show negative market timing ability.
After this step, we run the correlation analysis and find the same results of the Treynor and Mazuy model with a positive
correlation between the extra-performance of the fund and the net flows but a negative correlation between the same flows and both
Arab Spring news and Ebola media coverage. Table 13 summarizes the results.

610
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Considering the extra performance obtained according to the Heriksson and Merton model, we replicate the multivariate
regressions for the monthly net flows, with both dummy and news variables. Tables 14 and 15 summarize the results.
As shown, the main results concerning the Arab Spring and Ebola still hold, especially in the second regression, where both Arab
Spring and Ebola media coverage severely affect monthly net flows of African mutual funds. These robustness tests demonstrate that
even if we change the way to measure fund managers’ ability, these exogenous shocks are significantly relevant for explaining the
fund flows. Retail investors in African funds are very sensitive to this type of news.

7. Conclusions

Our paper is related to the extensive literature on financial markets’ anomalies and especially investors’ over-reaction. The
literature has widely reported that investors in developed markets do not behave rationally (Barberis et al., 2001; Barberis and
Thaler, 2003; Subrahmanyam, 2007 among others). Only a few papers focus on emerging markets, especially on specific countries
such as Turkey (Tekce and Yilmaz, 2015; Tekce et al., 2016) and Egypt (Boubaker et al., 2015). Tekce et al. (2016), analyzing Istanbul
Stock Exchange transactions, find that behavioral biases are common among Turkish retail investors. Boubaker et al. (2015)
examined over-reaction to specific events. Using data from the Egyptian stock exchange (EGX) over the period of 2003–2009, the
authors investigate short-term overreaction to four major events, i.e., terrorist attacks, tensions in the Middle East region, the
formation of new governments, and the announcement of the privatization of state-owned enterprises (SOEs).
Kaminsky et al. (2001) argued that the impact of specific country/region risks on overall mutual funds’ behavior could be far more
relevant for emerging markets. With this paper, we try to fill part of that gap, assessing the impact of specific major events (the Arab
Spring and Ebola outbreak) on African mutual fund flows. We studied the investors’ over-reaction through their African equity
mutual fund investments. The reason is twofold: first, equity mutual funds give investors the possibility to invest or withdraw their
money at any time, buying or redeeming their funds quote; second, equity mutual funds provide liquidity to the stock markets, which
are key factors for economic growth.
Through a sample of 78 African mutual funds, we contribute to the literature investigating the over-reaction to specific major
events, such as Ebola and the Arab Spring. In our models, we considered both these events as dummy variables covering the period of
their insurgency; then, we considered their media coverage as the number of articles that reported “Ebola” or “Arab Spring” in the
headline during their insurgency period. We modeled both the impact of these events on mutual fund performances and the flows.
As we expected, we find that investors overreact to these negative major events, even if related to a small portion of the continent
as Ebola, withdrawing their savings from the funds. Media coverage plays an important role: the higher the number of articles about
the Arab Spring and Ebola, the higher the withdrawals. These irrational withdrawals could have damaged the managers’ ability to
generate extra-performance, reducing their market timing ability and thus exacerbating the negative effects on the flows.
Consequently, they also reduced financial resources that, through the capital flows, feed financial markets and support firm
development, eventually promoting economic growth. The robustness tests demonstrate that even if we change the way to measure
fund managers’ ability, these exogenous shocks are significantly relevant for explaining the fund flows.
Investors in African funds are very sensitive to this type of news: the media coverage of such events plays a critical role in
influencing investors’ behavior. Excessively alarmist news could have induced retail investors to withdraw their savings from African
mutual funds, thus reducing financial resources to the entire continent.

References

Acquaah, M., 2015. Determinants of corporate listings on stock markets in Sub-Saharan Africa: evidence from Ghana. Emerg. Mark. Rev. 22, 154–175.
Barberis, N., Thaler, R., 2003. A survey of behavioral finance. In: Constantinides, G.M., Harris, M., Stulz, R. (Eds.), Handbook of the Economics and Finance. Elsevier
Science (Chapter 18).
Barberis, N., Huang, M., Santos, T., 2001. Prospect theory and asset prices. Q. J. Econ. 116 (1), 1–53.
Beck, T., Levine, R., 2003. Legal Institutions and Financial Development, NBER Working Papers 10126. National Bureau of Economic Research.
Bhattacharya, A., Montiel, P.J., Sharma, S., 1996. Private capital flows to Sub-Saharan Africa: an overview of trends and determinants. In: Iqbal, Zubair, Kanbur, Ravi
(Eds.), External Finance for Low Income Countries. IMF, Washington, pp. 207–232.
Bialkowky, J., Bohl, M.T., Kaufmann, P., Wisniewski, T.P., 2013. Do mutual fund managers exploit the Ramadan anomaly? Evidence from Turkey. Emerg. Mark. Rev.
15, 211–232.
Bialkowsky, J., Otten, R., 2011. Emerging market mutual fund performance: evidence for Poland. N. Am. J. Econ. Finance 22, 118–130.
Blake, D., Timmermann, A., 1998. Mutual fund performance: evidence for UK. Eur. Finance Rev. 2, 57–77.
Bokpin, G.A., 2017. Foreign direct investment and environmental sustainability in Africa: the role of institutions and governance. Res. Int. Bus. Finance 39, 239–247.
Borensztein, E.R., Gelos, R.G., 2000. A Panic-Prone Pack? The Behavior of Emerging Market Mutual Funds IMF Working Paper, WP/00/198.
Boubaker, S., Farag, H., Nguyen, D.K., 2015. Short-term overreaction to specific events: evidence from an emerging market. Res. Int. Bus. Finance 35, 153–165.
Brennan, M.J., Cao, H.H., 1997. International portfolio investment flows. J. Finance 52 (5), 1851–1880.
Brown, K.C., Harlow, W.V., 1988. Market overreaction: magnitude and intensity. J. Portf. Manage. 14, 6–13.
Carhart, M., 1997. On persistence in mutual fund performance. J. Finance 52 (1), 57–82.
Cho, Y., 1986. Inefficiencies from financial liberalization in the absence of well-functioning equity markets. J. Money Bank. Credit 18 (2), 191–199.
Chuhan, P., Claessens, S., Maming, N., 1993. Equity and Bond Flows to Asia and Latin America, World Bank Working Paper No 1160. World Bank, Washington, DC.
De Bondt, W.F.M., Thaler, M., 1985. Does the stock market overreact? J. Finance 40 (3), 793–805.
Demigurc-Kunt, A., Huizinga, H., 1993. Portfolio Investments in Emerging Stock Markets: Direct and Indirect Barriers in Portfolio Investments in Developing
Countries, World Bank Discussion Paper No. 228. In: Claessens, S., Gooptu, S. (Eds.), World Bank, Washington, DC, pp. 308–322.
Dermine, D., Roller, L.H., 1992. Economies of scale and scope in French mutual funds. J. Finan. Intermed. 2, 83–93.
Droms, W.G., 2006. Hot hands, cold hands: does past performance predict future returns? J. Finan. Plann. 19, 60–69.
Edmans, A., Garcia, D., Norli, O., 2007. Sports sentiment and stock returns. J. Finance 62, 1967–1998.
Engelberg, J.E., Parsons, C.A., 2011. The causal impact of media in financial markets. J. Finance 66, 67–97.
Fernandez, P., Bermejo, V.J., Bilan, A., 2008. Poor Performance of Mutual Funds in Spain 1991–2007, IESE Research Papers No. D/746.

611
A. Del Giudice, A. Paltrinieri Research in International Business and Finance 41 (2017) 600–612

Fernandez-Arias, E., Montiel, P., 1996. The surge in capital inflows to developing countries: an analytic overview. World Bank Econ. Rev. 10 (1), 51–77.
Ferreira, M.A., Miguel, A.F., Ramos, S., 2006. The determinants of mutual fund performance: A cross-country study, Swiss Finance Institute Research Paper 30.
French, K.R., 2008. The cost of active investing, social science research network, working paper available at http://ssrn.com.
Harvey, C.R., 1995. Predictable risk and returns in emerging markets. Rev. Financ. Stud. 8, 773–816.
Hau, H., 2001. Location matters. J. Finance 56 (5), 1959–1983.
Henriksson, D., Merton, R.C., 1981. On market timing and investment performance. II. statistical procedures for evaluating forecasting skills. J. Bus. 54 (2), 513–533.
Huij, J., Post, T., 2011. On the performance of emerging market equity mutual funds. Emerg. Market Rev. 12, 238–249.
International Monetary Fund, 2014. Managing Volatile Capital Flows. Experiences and Lessons for Sub-Saharan African Frontier Markets. African Department 01/
2014.
Institute of International Finance, 2015. Capital Flows to Emerging Markets. January.
Kaminsky, G.L., Lyons, R.K., Schmukler, S.L., 2001. Mutual funds investment in emerging markets: an overview. World Bank Econ. Rev. 15 (2), 315–340.
Kar, M., Nazlioglu, S., Agir, H., 2011. Financial development and economic growth nexus in the MENA countries: bootstrap panel granger causality analysis. Econ.
Modell. 28 (1–2), 685–693.
Kenny, C.J., Moss, T.J., 1998. Stock markets in Africa: emerging lions or white elephants? World Dev. 26 (5), 829–843.
Khorana, A., Servaes, H., Tufano, P., 2005. Explaining the size of the mutual fund industry around the world. J. Finan. Econ. 78 (1), 145–185.
King, R., Levine, R., 1993. Finance and Growth: Schumpeter Might Be Right. Mimeo, World Bank, Washington, DC.
Larson, S.J., Madura, J., 2003. What drives stock behavior following extreme one-day returns. J. Finan. Res. 26 (1), 113–127.
Levine, R., Zervos, S., 1993. What we have learned about policy and growth from cross-country regressions? Am. Econ. Rev. 83 (2), 426–430.
Levine, R., Zervos, S., 1996. Stock market development and long run growth. World Bank Econ. Rev. 10 (2), 323–340.
Levine, R., Zervos, S., 1998. Capital control liberalization and stock market development. World Dev. 26 (7), 1169–1183.
Malkiel, B.G., Fama, E.F., 1970. Efficient capital markets: a review of theory and empirical work. J. Finance 25 (2), 383–417.
Malkiel, B.G., 1995. Returns from investing in equity mutual funds 1971–1991. J. Finance 50 (2), 549–572.
McKinnon, R., 1973. Money and Capital in Economic Development. The Brookings Institutions, Washington DC.
Moss, T., Ramachandran, V., Standley, S., 2007. Why Doesn’t Africa Get More Equity Investment? Frontier Stock Markets, Firm Size and Asset Allocations of Global
Emerging Market Funds, Center for Global Development, Working Paper Number 112, February.
Otten, R., Bams, D., 2002. European mutual fund performance. Eur. Finan. Manage. 8 (1), 75–101.
Rousseau, P.L., Wachtel, P., 2000. Equity markets and growth: cross-country evidence on timing and outcomes, 1980–1995. J. Bank. Finance 24 (12), 1933–1957.
Samargandi, N., Fidrmuc, J., Ghosh, S., 2015. Is the relationship between financial development and economic growth monotonic? Evidence from a sample of middle
income countries. World Dev. 66–81.
Seyhun, H.N., 1990. Overreaction or fundamentals: some lessons from insiders’ response to the market crash of 1987. J. Finance 45, 1363–1388.
Shumpeter, J., 1934. The Theory of Economic Development. Harvard University Press, Cambridge, MA.
Subrahmanyam, A., 2007. Behavioral finance: a review and synthesis. Eur. Finan. Manage. 14 (1), 12–29.
Sugimoto, K., Matsuki, T., Yoshida, Y., 2014. The global financial crisis: an analysis of the spillover effects on African stock markets. Emerg. Market Rev. 21, 201–233.
Tekce, B., Yilmaz, N., 2015. Are individual stock investors overconfident? Evidence from an emerging market. J. Behav. Exp. Finance 5, 35–45.
Tekce, B., Yilmaz, N., Bildik, R., 2016. What factors affect behavioral biases? Evidence from Turkish individual stock investors. Res. Internat. Bus. Finance 36,
515–526.
Treynor, J.L., Mazuy, K.K., 1966. Can mutual funds outguess the market? Harv. Bus. Rev. 45, 131–136.
Tversky, A., Kahneman, D., 1971. Belief in the law of small numbers. Psychol. Bull. 76 (2), 105–110.
World Bank, Annual Report 2014, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/20093
World Health Organization, World Health Statistics 2015, http://www.who.int/gho/publications/world_health_statistics/2015/en/.

612