International Journal of Business and Social Science Vol. 7, No.

7; July 2016

Using Volatility to Improve Momentum Strategies

Omar Khlaif Gharaibeh

Al al-Bayt University
P.O.BOX130040, Mafraq 25113
Jordan

Abstract
This paper attempts to enhance momentum strategy by using volatility effect. To achieve this objective, double
sorting portfolio is used and data is collected from 10 Arabic market indices over the period of 1990-2014. A
simple modification to the traditional momentum strategy provides highly profitable results in Arabic market
indices. While traditional momentum alone provides significant abnormal raw return of 1.16% per month over
the six-month holding period, new momentum strategy based on double sort suggested by this study represented
via recent winners with low-volatility outperform recent losers with high-volatility and it provides significant
abnormal raw returns of 2.60% per month over the same holding period. Finally, either traditional momentum or
momentum with volatility strategies can’t be explained by two factor model
Keywords: volatility, momentum, strategy, two-factor model
1. Introduction
Over the past two decades, momentum effect has been considered a big challenge to efficient market theory, since
in efficient markets the investor cannot achieve profit over the long- term by investing in stocks that have done
well over 3, 6, 9 and 12 months. According to efficient markets theory, investors cannot achieve additional returns
without bearing additional risk. Behavioral finance has assisted in explaining how the momentum anomaly could
exist. Therefore, this paper will address this gap by investigating whether there is a momentum effect on the 10
Arabic indices.
To examine whether the momentum strategy can be improved, this study uses double sort 10 Arabic indices
employing a momentum strategy as the first sort variable and a volatility strategy as the second sort variable.
Previous studies related to momentum use past 3, 6, 9 and 12-month returns to sort securities into portfolio, this
study adopt this practice for first sort variable. The volatility strategy was calculated based on past 6 and 12
months in double sorts. In other words, the strategy in the current study is formed by selling the portfolio that
includes indices that have achieved poorly over the past 3, 6, 9 and 12 months with relatively high volatility
“SLHV”. On the other hand, buying the portfolio that includes those indices that have achieved well over the past
3, 6, 9 and 12 months with relatively low volatility “SWLV”.
This study addresses 10 Arabic market indices while previous studies address either developed or developing
market indices without giving the attention for these Arabic market indices such as Bornholt and Malin (2013).
unlike the Gharaibeh and Al-Eitan (2015) this paper adopts momentum and momentum with volatility strategies
rather than momentum and 52wk high strategies. While Ejaz and Polak (2014) use 7 stock markets of 6 Middle
East countries to investigate only momentum effect, this paper comprehensively investigates momentum and
momentum with volatility effects at the level of 10 Arabic market indices. In addition, the current study considers
the international two-factor model applied by Balvers and Wu (2006) to risk-adjust raw returns rather than
applying the capital asset pricing model (CAPM) like Ejaz and Polak (2014).
This paper contributes to the literature in two ways. Firstly, while previous studies examine momentum at the
level of developed or emerging market indices, this study investigates the topic at the 10 Arabic indices level. The
main finding from this paper is that there is strong evidence of momentum at the level of Arabic indices.
Secondly, this paper compares and contrasts the momentum strategy with the new double sort strategy based on
momentum with volatility suggested by the current study.

99
ISSN 2219-1933 (Print), 2219-6021 (Online) © Center for Promoting Ideas, USA www.ijbssnet.com

The results show that the new momentum with volatility strategy has consistently larger profits than the
momentum strategy. The rest of this paper is arranged as follows. Section 2 presents the relevant literature. The
data and methodology used in the current study is discussed in Section 3. Section 4 analyzes the results for both
raw and risk adjusted returns. Finally, Section 5 concludes the paper.
2. Literature Review
Ang, Hodrick, Xing, and Zhang (2006) investigate the relationship between the pricing of aggregate volatility risk
and the cross-section of stock returns. They follow Breeden, Gibbons, & Litzenberger (1989) and Lamont (2001)
methodology to construct a factor to mimic innovations in market volatility. They show that there is a negative
relationship between stocks that have sensitive to innovations in aggregate volatility and idiosyncratic volatility
with average returns. In the other words, they find that stocks that have high sensitivities to innovations in
aggregate volatility and high idiosyncratic volatility relative to Fama & French (1993) achieve low average
returns.
Berrada and Hugonnier (2008) explain the relationship between idiosyncratic volatility and stock returns by
developing a model of company investment under incomplete information assumption. This model has a new
implication for the cross-section of stock returns. Particularly, the model predict that the idiosyncratic volatility
anomaly can be mitigated by controlling for earning forecast errors and this implication is consistent with the
result of Jiang, Xu & Yao (2007). They find that idiosyncratic volatility is negatively related to stock returns in
the length of expansions and recessions and this finding is consistent with findings documented by Jiang et al.
(2007), Brockman and Yan (2008) and Ang et al. (2006, 2008). Surprisingly, they show that the idiosyncratic
volatility is positively (negatively) related to stock returns after good (bad) news.
Blitz and Van Vliet (2008) examined whether classic cross-sectional return patterns, that are documented at the
security level, can also exist across asset classes. They used a single global tactical asset allocation (GTAA)
model to directly compare the attractiveness of a broad range of asset classes, as well as to develop a given
strategic asset allocation. Blitz and Van Vliet (2008) found that value and momentum strategies across 12 asset
classes earn economically and statistically significant abnormal returns. In particular, 7% to 8% return premiums
have been documented for the 1-month, 12-months momentum and value GTCAA strategies over 1986-2007.
They confirmed the findings of previous studies that there is a 1-month momentum effect at industry level, in
contrast to the existence of a 1-month reversal effect documented at the individual stock level. They reported an
alpha of 12% when combining momentum and value factors for their GTCAA strategy. Their findings challenge
the concept of market efficiency because they show that the value and momentum effects extend across other
asset classes.
George and Hwang (2010) have constructed three idiosyncratic volatility measures by using monthly return based
on past five years, daily returns of the previous year and daily returns of the prior month. They confirm the
finding of the AHXZ (2006) that the return is negatively related to idiosyncratic volatility after controlling firm
size and controlling for January. This relationship is assigned to mispricing of the subsample of high-volatility
stocks, which are not generally followed by analysts. Outside of low coverage stocks, idiosyncratic volatility is
insignificantly or positively related to returns when measured as in AHXZ. They point out that the AHXZ finding
related to distinct patterns in earnings can be explained by the market overestimates the persistence of earnings
growth for low-coverage stocks with high idiosyncratic return volatility. They report a comparable finding for
stocks sorted on the volatility of share turnover, recommending that Chordia, Subrahmanyam & Anshuman
(2001) document that the return is negatively related to turnover volatility also attribute to mispricing coupled
with limits of arbitrage and information uncertainty.
Recently, Bornholt and Malin (2011) investigated whether each index’s recent volatility can be employed to
enhance the profitability of the standard momentum approach. In other words, a double-sorting procedure was
used to test whether a momentum/volatility strategy outperforms the standard single-sorted momentum strategy.
They used international indices grouped as developed and emerging markets. Bornholt and Malin (2011) found
that the momentum/volatility strategy provided only small enhancements over pure momentum in the case of
developed markets, however the new strategy performed surprisingly well when applied to emerging markets.
Recent high volatility winners outperformed recent low volatility losers on an average annualized basis by 17.4%
during the holding period.

100
International Journal of Business and Social Science Vol. 7, No. 7; July 2016

On the other hand, the long portfolio of the pure momentum strategy outperformed the short portfolio by 9.1%.
Furthermore, for the case of emerging markets, they showed that high volatility winners achieve an average
annualized return of 28.3%, and an alpha of 21.1%. Asness, Moskowitz and Pedersen (2013) investigated whether
a combined value/ momentum strategy based on individual assets provides abnormal returns within countries and
across different asset classes. They showed that the value/momentum strategy generates abnormal returns across
markets and asset classes and that the profitability of the value (momentum) strategy in one asset class is
positively related to the profitability of the value (momentum) strategy in other asset class. Moreover, they found
that value and momentum are negatively correlated within and across asset classes. While liquidity and macro
risks seem to be important common components of value and momentum, they leave a significant portion of both
unexplained.
Looking at the Arabic market region, Gharaibeh and Al-Eitan (2015) investigates whether there is momentum and
52 wk high strategies at the level of 10 Arabic market indices. They find that momentum effect is present and it
economically significant while the 52 wk high effect is unprofitable. They conclude that the 52 wk high effect is
not as regular as the momentum effect. In the Amman stock exchange market, Gharaibeh (2015) examines
whether there is a momentum effect on Jordan firm returns. He finds that momentum effect is only statistically
significant for the large-sized portfolios at Jordan firm returns. Using data of Morocco Stock Exchange from 1995
through 2014, Gharaibeh (2016) shows strong evidence of momentum effect. He confirmed that the momentum
effect is still statistically significant when applying to sub-period sample.
3. Data and Methodology
Monthly returns are calculated from monthly prices with reinvested gross dividend of 10 Morgan Stanley Capital
International (MSCI) indices downloaded from Data stream. The time frame of the study extends from February
1988 through September 2014. Table 1 summarizes the 10 Arabic countries, together with average monthly
return, standard deviation, skewness and kurtosis data for each index.
Table 1 details descriptive statistics over the period February 1988 through September 2014 for the 10 Arabic
indices, showing average monthly returns, standard deviation, Skewness and Kurtosis for each index. Table 1
shows big difference in the mean and standard deviation of average returns. Egypt and Lebanon have the biggest
monthly average (over 2% per month). On the other hand, the Bahrain has the lowest average at -0.33. The 10
Arabic indices have an average monthly return of 1.40% and an average standard deviation of 8.18%. The study
compares and contrasts the pure momentum and the momentum with volatility strategies applied to 10 Arabic
indices. The next two sections detail the pure momentum and the momentum with volatility strategies used in this
paper.
3.1 Momentum strategy
The momentum portfolios have been formed as follows. At the beginning of each month t, the 10 Arabic indices
in Table 1 are classified based on their past J-month returns (J = 3, 6, 9 and 12 months). For a given J, the short-
term winner (SW) portfolio includes the 50% of indices that have the highest past J-month returns whereas the
short-term loser (SL) portfolio includes the 50% of indices that have the lowest past J-month returns. The
momentum strategy (SW-SL) buys the short-term winner portfolio and sells the short-term loser portfolio.
Portfolios are held for K-month holding periods, where K = 1, 3, 6, 9 and 12 months. For this single-sort strategy,
this study maintains a 1-month gap between the end of the J-month formation period and the beginning of the K-
month holding period. A gap of one month is consistent with previous studies such as Jegadeesh and Titman
(1993). Jegadeesh and Titman (1993) found that skipping the first one month after the end of the formation period
improves the performance of the momentum strategy and provides stronger results since this practice helps avoid
any short-term reversals being compensated by the short-term continuation of returns.
3.2 Momentum with Volatility Strategy
In the momentum strategy, the investor buys a portfolio of short-term winners and sells a portfolio of short-term
losers. The success of such a strategy is dependent on the indices in the portfolios continue their short-term past
performances. One problem with this strategy is that these indices may not all be equally ready to continue.
Having a one month gap between the end of the formation period and the beginning of the holding period may
increase continuation impacts but it does not guarantee that the short-term winner and loser indices improve the
continuation by the end of the 1 month gap.

101
ISSN 2219-1933 (Print), 2219-6021 (Online) © Center for Promoting Ideas, USA www.ijbssnet.com

As discussed in the beginning of this paper, Bornholt and Malin (2011) suggest momentum with volatility
strategies in order to select those short-term winners and losers that are related to the volatility. In other words, the
momentum with volatility approach uses the volatility performances of the short-term winners and losers in a
double-sort procedure. Particularly, the momentum with volatility strategy buys short-term winners with
relatively high volatility and sells short-term losers with relatively low volatility.
The momentum with volatility approach is a double dependent sort procedure, and is described as follows. The
first sort is the same as the momentum strategy sort. The 10 indices are classified at the beginning of each month
based on their most recent past J-month returns. For a given J, the short-term winner portfolio (SW) includes the
50% of indices with the lowest past J-month returns, while the short-term loser portfolio (SL) contains the 50% of
indices with the highest past J-month returns. The indices in the SW and SL portfolios are further classified in the
second stage based on their past J2-month volatilities, where J2 = 6 or 12 months. This means that these J2-
volatilities are from the last J2 months of the J-month formation period. For a given J and J2, the SWHV
portfolio contains the 50% of SW indices with the largest J2-month volatilities (here HV indicates ‘high
volatility’). Similarly, the SLLV portfolio contains the 50% of SL indices with the lowest J2-month volatilities
(where LV denotes ‘low volatility’).
This procedure means that out of the total of 10 Arabic indices, the short-term winner and short-term loser
portfolios of the momentum strategy each contain 5 indices, while the momentum with volatility SWHV and
SLLV portfolios each contain 2 indices. The momentum with volatility strategy (SWHV-SLLV) buys short-term
winners with relatively high volatility (SWHV) and sells short-term losers with relatively low volatility (SLLV).
Bornholt and Malin (2011) shows that this strategy provide marginally larger profits than will the corresponding
momentum strategy.
Alternatively, the current study suggests another way to improve the momentum strategy; therefore, it constructs
the double-sorts in opposite way in terms of volatility as the second variable. Thus, The double-sorts is also
constructed by buying the portfolio that includes indices that have performed well over the past recent J-month
returns (J = 3, 6, 9 and 12-month) and that have also demonstrated relatively low volatility (SWLV), and selling
the portfolio that includes those indices that have performed poorly over the past recent J-month returns (J = 3, 6,
9 and 12-month) with relatively high volatility (SLHV). The key insight of the momentum with volatility
approach is this strategy should outperform the corresponding momentum strategies.
As with the momentum strategy, all portfolios in the momentum with volatility strategy are held for a K-month
holding period, where K = 1, 3, 6, 9 or 12 months. While a 1-month gap is employed between the end of the
formation period and the beginning of the holding period for the momentum strategy, the momentum with
volatility strategy in this study follows the method of Bornholt and Malin (2011) by having only a one-month gap
between the end of the formation period and the beginning of the holding period. The current study uses
Jegadeesh and Titman’s (1993) overlapping portfolio method for the holding period returns of all strategies to
avoid overlapping returns, and to enhance test power. For expositional convenience, the 6-month holding period
case (K = 6) will be the major focus of this paper comments about the empirical results in the next section.
Table 1: Summary statistics of stock index returns
country Mean% S.D% Skewness Kurtosis
Egypt 2.56 9.95 1.00 4.62
Lebanon 2.18 9.24 1.31 5.54
Morocco 1.89 5.77 0.46 2.55
Qatar 1.84 8.74 -0.13 1.66
Kuwait 1.37 7.06 -0.16 0.76
Jordan 1.33 5.21 -0.10 2.05
Oman 1.19 6.17 -1.31 5.01
UAE 1.09 11.00 0.19 1.40
Sudia Arabia 0.86 11.46 -0.08 -0.62
Bahrain -0.33 7.18 -0.61 2.99
AVERAGE 1.40 8.18

102
International Journal of Business and Social Science Vol. 7, No. 7; July 2016

4. Results
This section analyses the results for both the momentum and momentum with volatility strategies in terms of raw
and risk-adjusted results.
4.1 Momentum results
Table 2 report results for the long (SW), short (SL), and long-short (SW-SL) momentum portfolios for several (J,
K) combinations. Table 2 contains the results for formation period lengths of J = 3, 6, 9 and 12 months. Each table
presents the equal-weighted average monthly portfolio returns in percentages for K-month holding periods (K = 1,
3, 6, 9 and 12 months) in columns 3 through 6. The momentum results in Table 2 show that the strategy profits
(SW-SL) are positive over all K-month holding periods if J = 3, 6, 9 or 12 months. Table 2 shows significant
momentum SW-SL profits for all J = 3 to 6 months and all K. As an example, for the 6-month formation period
and 6-month holding period (K=6) case, the difference between the average monthly returns of the SW portfolio
and the SL portfolio is 1.16% per month (t-stat 2.60), which is statistically significant. In general, the holding
period returns in Table 2 provide evidence of momentum effect at the Arabic indices level. Consequently, next
consider the results for using volatility effect based on double-sort in the Table 3.
Table 2: Profitability of Momentum Strategies.
Holding Months
J Portfolio K =1 K =3 K =6 K =9 K =12
3 SW 1.33% 1.38% 1.21% 1.06% 0.91%
(2.93) (3.13) (2.86) (2.58) (2.21)
SL 0.04% -0.20% -0.01% 0.02% 0.07%
(0.09) (-0.44) (-0.02) (0.04) (0.14)
SW-SL 1.29% 1.57% 1.22% 1.04% 0.84%
(2.53) (3.66) (3.32) (3.31) (2.9)
6 SW 1.48% 1.44% 1.08% 0.87% 0.73%
(2.88) (2.84) (2.35) (2.02) (1.69)
SL -0.34% -0.30% -0.07% -0.02% -0.03%
(-0.73) (-0.6) (-0.14) (-0.04) (-0.05)
SW-SL 1.82% 1.74% 1.16% 0.90% 0.76%
(3.45) (3.48) (2.6) (2.37) (2.1)
9 SW 1.16% 0.95% 0.69% 0.56% 0.42%
(2.21) (2.03) (1.62) (1.31) (0.97)
SL -0.25% -0.13% -0.01% 0.04% -0.02%
(-0.46) (-0.25) (-0.02) (0.07) (-0.03)
SW-SL 1.40% 1.09% 0.70% 0.53% 0.43%
(2.42) (2.08) (1.55) (1.3) (1.12)
12 SW 0.77% 0.61% 0.33% 0.15% -0.04%
(1.38) (1.21) (0.71) (0.33) (-0.08)
SL -0.15% -0.11% 0.06% 0.02% -0.06%
(-0.29) (-0.21) (0.11) (0.04) (-0.1)
SW-SL 0.91% 0.72% 0.27% 0.13% 0.02%
(1.49) (1.25) (0.51) (0.27) (0.04)
4.2 The momentum with volatility strategy results
One of the objectives of the current study is to examine whether the momentum with volatility approach can
improve the performance of the traditional momentum strategy. The momentum with volatility strategy suggested
by Bornholt and Malin (2011) is based on buying those short-term winners with relatively high volatility
performances and selling those short-term losers with relatively low volatility performances. Momentum and
volatility strategies proposed by Bornholt and Malin (2011) enhanced the traditional momentum strategies. This
section reports momentum with volatility results.

103
ISSN 2219-1933 (Print), 2219-6021 (Online) © Center for Promoting Ideas, USA www.ijbssnet.com

Table 3 provides the results of the momentum with volatility strategies for the Arabic indices showing the average
monthly returns of the longs (SWHV), shorts (SLLV), and the arbitrage (SWHV-SLLV) portfolios, together with
their t-statistics. The results in Table 3 demonstrate considerable differences from the results in Table 2. Table 3
shows insignificant momentum with volatility SWHV-SLLV profits for all J1 and J2 months and all K. For the 6-
month formation period case with a six-month holding period (K= 6), for example, past short-term winners with
high volatility produce an average of 0.27% per month whereas past short-term losers with low volatility provide
an average of only 0.10 % per month over the same period. The resulting SWHV-SLLV difference of 0.18% per
month is statistically insignificant (t-stat 0.38).
Overall, the results in Table 3 show that the momentum with volatility strategies suggested by Bornholt and Malin
(2011) does not enhance the profitability of the momentum strategies. Therefore, the current study proposes new
momentum with volatility strategy based on buying those short-term winners with relatively low volatility
performances and selling those short-term losers with relatively high volatility performances. This new strategy
suggested by this paper is different from Bornholt and Malin’s (2011) strategy in terms of volatility. While the
arbitrage portfolio of Bornholt and Malin’s (2011) strategy based on (SWHV-SLLV), the current study prefer to
use this new strategy in opposite way in terms of volatility as (SWLV-SLHV). Tables 4 and 5 consider the results
for the new momentum with volatility suggested by this paper.
Table 4 demonstrates that the average monthly returns of the momentum with volatility strategies SWLV-SLHV
are larger than 2% per month and they are statistically significant for all formation and holding periods, except of
J/J2 = 3/6 case with a six-month holding period where the average monthly returns is 1.69% per month and it is
weakly significant. For example, the J/J2 = 6/6 case with a six-month holding period (K= 6). Short-term winners
that are low volatility provide an average of 1.11% per month. In contrast, short-term losers that are high volatility
generate an average of -1.49 per month over the same period. Consequently, the momentum with volatility
strategy (SWLV-SLHV) produces a significant 2.60% per month (t-stat 2.89). Generally, a comparison of Table 4
with Table 2 demonstrates that the momentum with volatility strategy outperforms the corresponding J = 3, 6, 9
or 12 momentum strategy with for all holding periods.
The robustness of these findings to the choice of range of past volatility can be checked by replacing the past one-
year volatility instead of past six-month volatility with momentum strategy. Thus, whereas momentum with past
six-month volatility in Table 4, momentum with past one-year volatility in Table 5. The result in Table 5 tells a
similar story for the momentum with volatility strategies. Table 5 confirms the previous findings in Table 4. The
SWLV-SLHV returns provide statistically significant profits for all holding periods. Consider, for example, the
6/12 case in Table 6 there is a significant K =6 returns of 2.92% (t-stat 3.00) which is larger than traditional
momentum strategy presented in Table 2. Comparing this result with the corresponding results in Table 2
demonstrates that the momentum with volatility approach has much larger returns. This proposes that the
momentum with volatility approach suggested by the current study achieve larger holding period profitability is
by avoiding indices that have high volatility.

104
International Journal of Business and Social Science Vol. 7, No. 7; July 2016

Table 3: Profitability of Momentum/Momentum with 6-month Volatility Strategies

Holding Months
J1 J2 Portfolio K =1 K =3 K =6 K =9 K =12
3 6 SWHV 0.62% 0.61% 0.45% 0.30% -0.04%
(1.14) (1.16) (0.9) (0.6) (-0.09)
SLLV 0.34% 0.21% 0.26% 0.32% 0.30%
(0.67) (0.48) (0.6) (0.71) (0.64)
SWHV-SLLV 0.29% 0.40% 0.19% -0.02% -0.35%
(0.51) (0.82) (0.45) (-0.07) (-1.09)
6 6 SWHV 0.67% 0.65% 0.27% 0.01% -0.32%
(1.06) (1.1) (0.52) (0.02) (-0.63)
SLLV 0.29% 0.01% 0.10% 0.13% 0.06%
(0.58) (0.02) (0.23) (0.3) (0.12)
SWHV-SLLV 0.38% 0.64% 0.18% -0.13% -0.37%
(0.61) (1.19) (0.38) (-0.33) (-1.06)
9 6 SWHV 0.85% 0.58% 0.32% -0.01% -0.26%
(1.54) (1.17) (0.66) (-0.02) (-0.54)
SLLV 0.06% 0.09% 0.27% 0.26% 0.20%
(0.12) (0.2) (0.64) (0.59) (0.45)
SWHV-SLLV 0.79% 0.49% 0.04% -0.27% -0.47%
(1.36) (0.99) (0.09) (-0.66) (-1.23)
12 6 SWHV 0.53% 0.49% 0.05% -0.20% -0.44%
(0.88) (0.95) (0.11) (-0.42) (-0.93)
SLLV 0.25% 0.26% 0.23% 0.14% 0.03%
(0.51) (0.57) (0.52) (0.31) (0.06)
SWHV-SLLV 0.29% 0.24% -0.17% -0.33% -0.47%
(0.47) (0.48) (-0.39) (-0.84) (-1.23)
Table 4: Profitability of Momentum/Momentum with 6-month Volatility Strategies
Holding Months
J1 J2 Portfolio K =1 K =3 K =6 K =9 K =12
3 6 SWLV 1.32% 1.42% 1.34% 1.11% 1.00%
(2.44) (2.74) (2.63) (2.16) (1.96)
SLHV -0.36% -0.95% -1.08% -1.40% -1.62%
(-0.38) (-0.96) (-1.07) (-1.38) (-1.55)
SWLV-SLHV 1.69% 2.37% 2.42% 2.51% 2.62%
(1.91) (2.94) (3.38) (3.51) (3.43)
6 6 SWLV 1.44% 1.23% 1.11% 1.00% 0.94%
(2.57) (2.25) (2.08) (1.94) (1.81)
SLHV -1.18% -1.41% -1.49% -1.67% -1.99%
(-1.14) (-1.27) (-1.28) (-1.45) (-1.74)
SWLV-SLHV 2.62% 2.64% 2.60% 2.67% 2.93%
(2.9) (2.87) (2.89) (3.11) (3.34)
9 6 SWLV 1.29% 1.48% 1.18% 1.05% 0.92%
(2.22) (2.6) (2.14) (2) (1.75)
SLHV -1.31% -1.39% -1.51% -1.74% -2.19%
(-1.1) (-1.17) (-1.34) (-1.6) (-1.97)
SWLV-SLHV 2.61% 2.87% 2.69% 2.79% 3.11%
(2.36) (2.83) (2.86) (3.15) (3.43)
12 6 SWLV 1.35% 1.28% 1.36% 1.18% 1.03%
(2.51) (2.38) (2.51) (2.17) (1.91)
SLHV -1.30% -1.30% -1.28% -1.69% -2.30%
(-1.18) (-1.08) (-1.06) (-1.39) (-1.88)
SWLV-SLHV 2.65% 2.58% 2.64% 2.87% 3.33%
(2.59) (2.49) (2.64) (2.88) (3.3)

105
ISSN 2219-1933 (Print), 2219-6021 (Online) © Center for Promoting Ideas, USA www.ijbssnet.com

Table 5: Profitability of Momentum/Momentum with 12-month Volatility Strategies

Holding Months
J1 J2 Portfolio K =1 K =3 K =6 K =9 K =12
3 12 SWLV 1.43% 1.41% 1.26% 0.99% 1.11%
(2.53) (2.57) (2.32) (1.81) (1.96)
SLHV -0.45% -1.32% -1.21% -1.77% -2.36%
(-0.45) (-1.23) (-1.07) (-1.52) (-1.98)
SWLV-SLHV 1.88% 2.73% 2.47% 2.75% 3.48%
(2.11) (3.12) (3.04) (3.36) (4.18)
6 12 SWLV 1.89% 1.69% 1.43% 1.26% 1.22%
(3.17) (3) (2.72) (2.44) (2.26)
SLHV -0.62% -1.26% -1.49% -1.95% -2.43%
(-0.56) (-1.07) (-1.18) (-1.5) (-1.79)
SWLV-SLHV 2.51% 2.95% 2.92% 3.21% 3.65%
(2.48) (2.97) (3.00) (3.3) (3.56)
9 12 SWLV 1.66% 1.52% 1.14% 1.12% 0.97%
(2.81) (2.6) (2.07) (2.05) (1.72)
SLHV -0.85% -1.43% -1.49% -1.77% -2.56%
(-0.69) (-1.15) (-1.21) (-1.47) (-2)
SWLV-SLHV 2.52% 2.95% 2.63% 2.89% 3.53%
(2.29) (2.75) (2.54) (2.92) (3.39)
12 12 SWLV 1.09% 1.12% 1.10% 0.96% 0.84%
(2.06) (2.12) (2.09) (1.8) (1.57)
SLHV -1.17% -1.44% -1.43% -1.92% -2.63%
(-1.05) (-1.18) (-1.16) (-1.53) (-2.03)
SWLV-SLHV 2.26% 2.56% 2.53% 2.87% 3.47%
(2.22) (2.42) (2.46) (2.74) (3.19)
Figure 1: Cumulative Returns of Strategies
This graph provides the cumulative returns for the pure momentum strategy (SW-SL), the Momentum with 6-
month volatility strategy (SWLV-SLHV) and the Momentum with 12-month volatility strategy (SWLV-SLHV)
for the 36 months following the end of the formation period.
100%
Momentum Strategy
90%
80%
70% Momentum with 12-month
Cumulative Returns

Volatility Strategy
60%
50% Momentum with 6-month
40% Volatility Strategy

30%
20%
10%
0%
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
Event Months

The post-formation behavior of the momentum and momentum with volatility strategies’ profits is also illustrated
in Figure 1. Figure 1 shows the post-formation cumulative returns of the momentum strategy (SW-SL), the
momentum with past six-month volatility strategy (SWLV-SLHV) and the momentum with past 12-month
volatility strategy (SWLV-SLHV) using non-overlapping (K = 1) for the 36 months following the end of the
formation period. For the momentum strategy depicted, it is evident that the momentum performance shows no
signs of slowing down by the end of the first 12 or 36 post-formation months. For momentum with volatility
strategies depicted, they are both evident either momentum with past 6-month volatility or momentum with 12-
month volatility that these strategies enhance the traditional momentum performance.

106
International Journal of Business and Social Science Vol. 7, No. 7; July 2016

Table 6: Risk adjusted Momentum/ Momentum and volatility profits

Portfolio Two-factor model

 bwld bvmg Adj R 2

Panel A: Pure Momentum Strategy
SW 1.0108 0.3854 -0.4414 3.80%
(160.64) (2.16) (-1.75)
SL 0.9941 0.9341 -0.8163 16.30%
(124.16) (2.96) (-1.94)
SW-SL 0.0167 -0.5487 0.3749 11.60%
(3.15) (-2.74) (1.43)
Panel B: Momentum and Volatility Strategy
SWLV 1.0127 0.4172 -0.4729 8.10%
(193.19) (2.76) (-2.11)
SLHV 0.9785 1.5073 -1.421 20.20%
(79.56) (3.01) (-2.17)
SWLV-SLHV 0.034 -1.09 0.948 17.40%
(3.6) (-2.91) (1.89)
4.3 Risk adjustments
To test whether the profits of these strategies should be considered a reward for bearing risk, the profits of the
momentum and momentum with past 12-month volatility strategies are risk adjusted employing the Two-factor
model. The two-factor model comprises of a market factor and a value minus growth factor (VMG) as follows:
R pt  Rft   p   p,mkt (Rmt  Rft )   p,vmgVMGt   pt
(1)
where the dependent variable Rp, t - Rf,t is the monthly excess return of a portfolio of interest (whether it’s the
long, short or the arbitrage portfolio of a strategy), Rp,t indicates to the monthly return of portfolio p at time t and
Rf,t the monthly risk free rate at time t represented by the one-month US T-bill return. The independent factors for
the two models are as follows: Rwld, t-Rf,t corresponds to the excess return on the MSCI World market portfolio
at time t and VMGt or Value minus Growth is the return on the MSCI World Value Index minus the return on the
MSCI World Growth Index at time t. The monthly values of the MSCI world market index and the MSCI world
value and growth indices were obtained from Data stream. The monthly returns for the Ibbotson and Associates
one-month treasury-bill risk free rate were downloaded from Kenneth French’s website. The two-factor model
risk adjustment covers the period from their first available months January 1989 until September 2013. The
coefficients βp and νp are the regression loadings corresponding to the factors of the model, while the intercept αp
or (alpha) refers to the risk-adjusted abnormal profits of the portfolio over the estimation period. The t-values
corresponding to the regression coefficients are corrected for heteroskedasticity using White’s (1980) test.
Table 6 shows the estimated regression coefficients and the associated t-statistics for the long, short and long-
short portfolios for the momentum and momentum with volatility strategies with six-month holding periods (K =
6). The alphas of the two strategies either momentum or momentum with volatility zero-cost portfolios (SW-SL,
SWLV-SLHV, respectively) in Panel A and B are large (1.67%, and 1.34% per month) and statistically
significant (t-stat 3.15 and 3.60, respectively). Clearly, the results in Panel A and B show that there are
momentum and momentum with volatility effects that cannot be explained by the two-factor model.
5. Conclusion
This paper has investigated whether momentum strategies applied to 10 Arabic indices can be enhanced by taking
into account each index’s volatility. The momentum with volatility strategy based on double-sorts provides big
improvements over the traditional momentum strategy. The momentum strategy provides statistical significant
profit of 1.16% per month. Short-winner with low past six-month volatility outperforms short-loser with high
past six-month volatility on an average basis by 2.60% per month. While these findings are important to
practitioners, they also suppose opportunities for future research into whether past volatility can improve
momentum strategy either at the level of stocks or industries. In addition, there is implication for academic
research. The returns of 10 Arabic market indices of both momentum and momentum with volatility strategies
cannot be explained by the Fama-French three-factor model. This means that these momentum and volatility
effects at the 10 Arabic market indices level needs further investigation in order to understand what is driving the
inter-index momentum and momentum with volatility effects.
107
ISSN 2219-1933 (Print), 2219-6021 (Online) © Center for Promoting Ideas, USA www.ijbssnet.com

References
Ang, A., Hodrick, R. J., Xing, Y., & Zhang, X. (2006). The cross section of volatility and expected returns. The
Journal of Finance, 61(1), 259-299.
Asness, C. S., Moskowitz, T. J., & Pedersen, L. H. (2013). Value and momentum everywhere. The Journal of
Finance, 68(3), 929-985.
Balvers, R. J., & Wu, Y. (2006). Momentum and mean reversion across national equity markets. Journal of
Empirical Finance, 13(1), 24-48.
Berrada, T., & Hugonnier, J. (2008). Incomplete information, idiosyncratic volatility and stock returns. Swiss
Finance Institute Research Paper Series.
Blitz, D., & Van Vliet, P. (2008). Global tactical cross-asset allocation: applying value and momentum across
asset classes. Journal of Portfolio Management, 35, 23-38.
Bornholt, G., & Malin, M. (2011). Using volatility to enhance momentum strategies. The Finsia Journal of
Applied Finance(2), 16.
Bornholt, G. N., & Malin, M. (2013). Strong and weak momentum components: Evidence from international
market indices. Available at SSRN 2315993.
Breeden, D. T., Gibbons, M. R., & Litzenberger, R. H. (1989). Empirical test of the consumption-oriented CAPM.
The Journal of Finance, 44(2), 231-262.
Brockman, P., & Yan, X. S. (2008). The time-series behaviour and pricing of idiosyncratic volatility: Evidence
from 1926 to 1962. under revision at Journal of Banking and Finance.
Chordia, T., Subrahmanyam, A., & Anshuman, V. R. (2001). Trading activity and expected stock returns. Journal
of financial economics, 59(1), 3-32.
Ejaz, A., & Polak, P. (2014). Short Term Momentum Effect: A Case of Middle East Stock Markets. Ejaz, A, 104-
112.
Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds* 1. Journal of
financial economics, 33(1), 3-56.
George, T. J., & Hwang, C. Y. (2010). Why Do Firms with High Idiosyncratic Volatility and High Trading
Volume Volatility Have Low Returns?
Gharaibeh, O. (2015). Interaction of Size and Momentum Effects in Jordan Firms: 2005-2014. International
Review of Management and Business Research 4(1), 1-16.
Gharaibeh, O. (2016). Evidence of the Momentum Effect in the Morocco Stock Market: 1995-2014 International
Review of Management and Business Research, 5(1), 200-2011.
Gharaibeh, O. K., & Al-Eitan, G. N. (2015). Is the 52 Week High Strategy as Pervasive as Momentum? Evidence
from Arabic Market Indices. Research Journal of Finance and Accounting.
Jegadeesh, N., & Titman, S. (1993). Returns to buying winners and selling losers: Implications for stock market
efficiency. Journal of Finance, 48(1), 65-91.
Jiang, G. J., Xu, D., & Yao, T. (2007). The information content of idiosyncratic volatility. Journal of Financial
and Quantitative Analysis, 44(01), 1-28.
Lamont, O. (2001). Economic tracking portfolios: National Bureau of Economic Research Cambridge, Mass.,
USA.
White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for
heteroskedasticity. Econometrica: Journal of the Econometric Society, 48, 817-838.

108