Analysis of Smart Beta Investment Strategies in Emerging Markets

INTRODUCTION

In 1981, the World Bank coined the term “emerging market” (Vihang & Errunza, 1983). At the time, foreign direct investment in poorer nations of the world had become a standard business practice, but foreign portfolio investment into these nations was virtually unprecedented. Around this time, a trend began in which many of these poorer nations took steps to significantly liberalize their nation’s financial markets to bring in new portfolio inflows from foreign entities (Eun & Janakiramanan, 1986). Particularly, investment through foreign institutional investors has been coveted due to the large size of their transactions (Bekaert, 1995).

Portfolio managers have stopped pondering whether or not to invest in emerging markets; instead, they ask “What percentage of the portfolio should be allocated toward emerging markets?” and “Which emerging nations are the most promising?” (Bekaert & Harvey, 2014). Since the 1980s, it has become significantly cheaper and easier than ever for investors to navigate the complex landscape of emerging equity markets, particularly through the advent of passive investment approaches (Malkiel, 2003). Passive investing has historically been defined as index funds that mimic traditional market-cap weighted indices, but passive investing has also come to include electronically traded funds which trade on a variety of factors (Malkiel, 2003). One such type of fund that has risen in popularity is known as the smart beta fund, which uses an alternative weighing scheme to attempt to take advantage of systematic biases or inefficiencies in the market.

Understanding emerging market investment is crucial for portfolio managers in the developed world considering new investments in the emerging markets have become a staple asset allocation among investors. Existing literature suggests that active management is still superior to traditional passive strategies when it comes to emerging market investment (French, 2008; Malkiel, 2003). Our study attempts to build upon the passive vs. active debate in emerging markets by analyzing the performance of passive funds that take advantage of these non-traditional weighting schemes against a well-known emerging markets index. This study attempts to answer whether there is a significant difference in the performance of emerging market smart beta strategies and traditional passive strategies. Higher returns of smart beta funds would mean that the smart beta strategies are beginning to capture the market return that active managers have historically prided themselves on their ability to capture.

We select five smart beta funds and evaluate their performance relative to the emerging market index fund. We find that none of the smart beta funds outperforms the emerging market index fund at the conventional significance level. As we continue to define traditional beta as any ETF that tracks an investable index solely based on market cap, geography, or sector, smart Beta is a new, popular financial product that attempts to beat indexed funds. Despite the lack of sufficient evidence that smart beta can outperform traditional indexed funds, the innovation of non-traditional ETFs is a growth factor for the ETF industry. With more than 1,000 ETFs on the market today, most smart beta ETF strategies did not outperform, and even for those that do, it may take a long enough investment time horizon to capture that potential outperformance. Just because a certain strategy is doing well or poorly in a particular arbitrary time frame doesn’t mean it’ll continue to do so going forward: Factors perform and underperform in cycles (Roy, 2019).

The remainder of this paper is organized as follows. Section 2 reviews the literature and develops a hypothesis. Section 3 explains the process of fund selection and the methodology used for analyzing the fund performance. Section 4 reports the statistical results. Section 5 includes the conclusion, limitations, and avenues for further research.

LITERATURE REVIEW AND HYPOTHESES

Characteristics of Emerging Equity Markets

Mainstream financial analysts took note of the emergence of capital markets in underdeveloped nations during the 1980s (Vihang & Vihang & Errunza, 1983). A country is deemed to be “emerging” if per capita GDP falls below a threshold; this threshold has changed over time since its inception. The tier above emerging markets is referred to as “developed,” while the tier below is referred to as “frontier.” Researchers state the goal of Emerging Market Economies (EMEs) is to “emerge” from an underdeveloped nation and join the developed nations of the world, a phenomenon known as convergence (Bekaert & Harvey, 2002). For this paper, these nations are referred to as Emerging Market Economies.

Due to the lack of available statistics and reliable data on EMEs in the 1980s, the International Finance Corporation (IFC) of the World Bank created a new data bank in the 1980s. The data bank consists of stock market statistics from 15 EMEs, starting with data from the year 1976 (Bekaert & Harvey, 2002). This data bank is referred to as the Emerging Market Database (EMDB). As a result of this database’s creation, the year 1976 tends to be the starting point for most research into EME capital markets. This database has since been sold to Standard & Poor (S&P), a global leader in index fund products. We use the MSCI Emerging Markets Index as a benchmark passive index. The footnote contains a list of the emerging markets this fund uses^[1].

Some of the earlier work in international finance attempted to model the impact of global integration on financial markets and security prices (Eun & Janakiramanan, 1986; Stulz, 1981). Historically, many underdeveloped nations shielded themselves from outside economic influence by instating capital controls that limited foreign investment. Furthermore, asymmetric information, fear of expropriation, and discriminatory taxation are all barriers that academics have identified as being deterrents to foreign investment (Eun & Janakiramanan, 1986). In recent decades, many EMEs have removed these barriers to stimulate economic growth. An example of this is China, its government began financial deregulation efforts in 1978 by increasing autonomy for state-owned banks to allocate credit, the gradual removal of restrictive state-owned ownership structures, and the relaxation of legal and geographical barriers to new financial intermediaries. Studies have shown that economic growth has been bolstered in China as a direct result of these more efficient allocation channels (He, 2012).

Financial markets in developed nations such as the United States are characterized by researchers as the most developed in the world and they present little liquidity risk to investors (Amihud, 2002; Amihud et al., 2005; Amihud & Mendelson, 1986, 1988; Bekeart & Harvey, 2000; Levine, 1997; Stoll, 1978). Liquidity can be defined as the ease with which a security can be traded (Bekaert & Harvey, 2002). For example, common stock is far more liquid than real estate, and equity can be traded on an American stock market more easily than on a Vietnamese stock market. Academic research explaining the role of liquidity began in the 1960s (Demsetz, 1968) and subsequent research has maintained that equity markets that can minimize the cost of trading an asset attract far more attention from investors (Amihud & Mendelson, 1986, 1988; Chordia et al., 2005; Levine, 1997; Stoll, 1978). Some researchers consider liquidity as the most significant distinction between developed and emerging capital markets, as EMEs operate in a highly illiquid environment where investors can expect to pay an added premium in transaction costs simply due to the lack of activity taking place (Frenkel & Menkhoff, 2004).

In both developed and emerging markets, some researchers argue that informational inefficiencies exist, which investors can take advantage of. The Stiglitz-Grossman Model is one of the first models to detail this phenomenon. The researchers who constructed this model state that sophisticated investors are compensated for their information-gathering with higher returns (Stiglitz & Grossman, 1980). Additionally, researchers have built on this model by concluding that less liquid markets have higher informational inefficiencies as well (Amihud et al., 2005). Therefore, illiquidity and high informational inefficiencies are characteristics of the emerging markets.

Bekaert, et al. (1995; Bekaert et al., 2003, 2004, 2007; Bekaert & Harvey, 2002, 2000) have kept their selection of EMEs fairly consistent throughout their research. Typically, they have studied 20 countries with the most data available in the EMDB (since at least 1990). What has been noted throughout all academic research on emerging markets is that the definition of “emerging” is constantly being changed as new markets migrate in and out of the classification of “emerging.” An innumerable number of factors go into the development of any one economy, so it can be very difficult to conduct meaningful research on EMEs. The MSCI Emerging Markets Index currently features 23 EMEs and is the most prominent emerging market benchmark by which emerging market funds measure their performance. It is estimated that around $1.6 trillion in funds track this index, placing it in elite investment among global equity investors (Kynge, 2017).

Performance of Active and Passive Management

Historically, emerging markets have been an esoteric, but rewarding investment destination for investors who understand their unique characteristics. The rise of quantitative modelling and electronically traded funds have made emerging markets a more accessible investment destination, but research shows that humans may still be the best stock pickers in these markets (Kremnitzer, 2012).

Institutions tend to rely on portfolio managers to do the bulk of their investing. Historically, these managers have been active stock pickers, primarily basing their stock choices on fundamental analysis. Quantitative modelling and behavioural techniques have also been employed by active managers in recent decades to narrow the universe of stock selection. Many researchers believe portfolio management activities can be decomposed into three tasks: (1) asset allocation, (2) stock selection, and (3) market timing (Lakonishok et al., 1992). The premise of active investing is to rely upon informational inefficiency between informed and uninformed investors, as the informed investor would be better able to distinguish when an asset is underpriced and creates an arbitrage opportunity. Active managers are likely to refer to the Stiglitz-Grossman model as justification for their activities. Researchers have shown that active managers seldom beat the market index in their stock selections across developed markets (Conversano & Vistocco, 2010). This can be illustrated by the study that concludes that a passive market portfolio on average would have performed 67 basis points higher each year than the aggregate of all active portfolios from 1980-2006 (French, 2008).

Passive management is identified by researchers as a portfolio that mimics an index. Researchers who are advocates of the efficient market hypothesis, which hypothesizes that security prices are random and not influenced by past events, tend to be the strongest advocates for passive investing over active investing (Fama, 1998; Malkiel, 2003). Researchers can trace passive investing to the 1970s when the first mutual fund index arose. These index funds are groups of securities that strictly track known market indices. However, most research on passive investment begins with the advent of the Exchange-Traded Fund (ETF). The first ETF, SPDR, was launched on January 22, 1993 (Elton et al., 2002). ETFs are groups of securities designed to mirror certain indices and sectors. Though they vary slightly in technical structure, ETFs and mutual fund indices are the primary ways passive investing takes place, and they are universally cheaper than actively managed funds (Park et al., 2014).

Passively managed funds have grown significantly in popularity. Vanguard estimates that in 2014, passive net investment inflows totalled $392 billion, topping the $66 billion of funds that went to actively managed funds (Rawson & Johnson, 2015). The active vs. passive management debate is among the most divisive within the financial community. As previously noted, there has been indisputable evidence that passive investing is cheaper and that the index has beaten active investing on average (Fama, 1998; French, 2008; Malkiel, 2003), but as ETFs and index funds rise in popularity, researchers have noted their drawbacks.

The shortcomings of traditional cap-weighted passive indices are that they are “the market” in essence, and therefore investors are going to get the market’s return. If the market is inefficient, meaning some companies are being priced too high (or too low), then the traditional passive indices will naturally contain a disproportionate amount of assets that are overvalued (or undervalued), while lacking undervalued (or overvalued) assets. Researchers note that as uniformed traders switch to passive investing, there is a decrease in price efficiency and an increase in transaction costs. Active managers specialize in seeking costly information to make decisions that would give them a better view of an underlying security, and less uninformed traders mean that there is less of a supply of willing market participants to trade this security. The result is a decrease in the impact of future earnings response coefficients and less efficient markets (Israeli et al., 2016).

While the majority of studies on active management conclude that active managers underperform the index, studies that focus on this comparison in emerging markets have different findings. One study indicates that active managers in emerging markets outperform their benchmark by an average of 49 basis points net of fees. It goes further to state that the less efficient the emerging market is, the more excess return is reaped by the active manager. The study concludes that active management’s value comes not only from the skill of the active manager, but also from how efficient the underlying market is (Dyck et al., 2013).

Smart Beta: A Hybrid Product

Ground-breaking financial researcher William Sharpe (1964) coins the terms “beta” and “alpha.” He defines the “Beta” as the measure of non-diversifiable risk in a portfolio, measured against a traditional benchmark market index, while “alpha” as the residual return that is not attributed to beta. These original definitions are important, as beta and alpha have come to be loosely defined within financial circles. Willis Tower Watson, the global investment consulting firm, first coined the term “smart beta” in the early 2000s (Philips et al., 2015). The definition of smart beta can vary significantly depending on who is being asked. For this paper, smart beta is defined as a rules-based investment strategy that seeks to provide exposure to risk premiums through various weighting mechanisms instead of a traditional market-cap weighting (Philips et al., 2015). These strategies, in essence, are passively managed but give the feel of an actively managed strategy without the high fees. Some examples of criteria that might be used to construct these strategies include volatility, dividends, and gross domestic product. Additionally, some smart beta strategies use more than one criterion to deviate away from a market-cap-weighted benchmark; these are commonly referred to as multi-factor strategies (Philips et al., 2015).

The founders of smart beta describe the strategies as containing the following passive characteristics: transparency, where the principles of the portfolio’s construction are plainly stated; rules-based, where the methodology is systematic and mechanically executed; low-cost, where the strategies are more affordable compared to active management; highly liquid, to accommodate easy entrance and exit; and well-diversified, where investors are not exposed to sector and industry mis-valuations (Arnott & Kose, 2014). A study by researchers at Vanguard found that the “excess return” generated by smart beta strategies can be somewhat explained by time-varying exposures to risk factors, such as style and fund size (Philips et al., 2015).

Hypothesis Development

The emerging markets present a unique angle to the active vs. passive debate. Passively managed strategies have made investing in these markets significantly easier than they would have been in the past. As previously noted, emerging markets are characterized as being more informationally inefficient and more illiquid than developed markets. As a result, some researchers indicate that returns from active managers in emerging markets are superior to returns by emerging market ETFs on a risk-adjusted basis (Kremnitzer, 2012).

Given the body of literature supporting passive investing (Fama, 1998; French, 2008; Malkiel, 2003), it is easy to understand why passive investing has become so popular in recent decades. It is less clear whether or not passive funds can significantly outperform actively managed funds in the emerging market investment space over a long period. Smart beta emerging market ETFs are an even less studied phenomenon given their recent introduction to the financial community. If smart beta strategies can provide significantly different returns from traditional indices in emerging markets, then these strategies will continue to narrow the divide between active and passive management in emerging markets, resulting in higher levels of efficiency in this investment space. We develop the following hypothesis to evaluate the performance of selected smart beta funds relative to the MSCI emerging markets index fund.

FUND SELECTION AND METHODOLOGY

All data used for this paper’s analysis has been retrieved from the Yahoo Finance Historical Prices database. This study analyzes the performance of five emerging market alternative-indexing funds, i.e. smart betas, which have been trading for at least 2 years. These funds are compared against the MSCI Emerging Markets Index for performance evaluation. The MSCI Emerging Markets Index is held in high regard among global equity investors and is considered the benchmark for measuring general emerging market performance. The data analysis is conducted using R and Excel. The multi-factor emerging market funds being studied and their factor criterion are listed in Table 1.

The exact valuation and weighting scheme for each fund are confidential, so we are unable to know exactly how much influence the multi-factor approach has on the actual trades that take place.

Monthly historical returns have been pulled for each of the five smart beta strategies from January 1, 2016, to January 1, 2018. Given the recent introduction of smart beta to the financial community, with most being launched sometime in 2015, this is the longest range of data available for study. Adjusted closing price is used to calculate returns to account for any dilution among the returns. The purpose of using the return instead of the closing price is to normalize the results. The monthly returns are used to calculate the monthly arithmetic mean and the monthly geometric mean. The geometric mean is the best comparison to use in investment comparisons because it considers the effect that compounding has on the investment and the fluctuation in the percentage return from year to year. The variance, standard deviation, and risk-adjusted return are also calculated for the MSCI Emerging Markets index and smart beta funds.

RESULTS

Figure 1.Visual comparison of 2-year monthly returns

The descriptive statistics reveal that overall the smart beta funds are quite comparable to the MSCI Emerging Markets Index in terms of performance. The two funds that outperform the geometric average over the two years are PXH and FNDE, with PXH edging out FNDE as the best-performing fund of the group. Both of these funds use cash flow as a factor; the others do not. Additionally, FNDE only trades large-cap stocks, which signals that large-cap over-perform throughout these two years. On a risk-adjusted basis, JPEM is the highest-performing fund given that it has much less variation in its monthly return compared to the other funds across these two years.

The results of the regression show that there is a clear, strong correlation between each smart beta fund and the MSCI Emerging Markets Index. This evidence for the correlation is reflected in the R-square values. Of these five funds, EMGF, JPEM, and GEM have extremely high R-squared values over 90%; this means that over 90% of the variation in these smart beta funds can be explained by the variation in returns from the MSCI Emerging Markets Index. PXH & FNDE, are the two funds which display the higher geometric monthly averages and lower R-squared values in the 70-80% range. The factor that sticks out from these two funds compared to the others is the cash flow factor, which indicates that the cash flow yield each quarter may be a factor that leads to stock outperformance in emerging markets. FNDE also has a tilt toward large-cap stocks, which may indicate that the size is an indicator of outperformance in emerging markets. We also conduct a t-test between the monthly returns of smart betas and the returns of the MSCI Emerging Market Index besides the regression analysis. The high p-values across all tests suggest that we have no reason to reject the null hypothesis that the smart beta funds are statistically significantly different from the MSCI Emerging Markets Index.

It is worth noting that although regressing testing and t-test methods have, historically, been developed along separate tracks, most statisticians consider them as special cases of the same General Linear Model (GLM). Both methods can do a mean-difference test between or in more than two groups and result in the same conclusion.

CONCLUSION, LIMITATIONS, AND AVENUES FOR FURTHER RESEARCH

The rise and success of passively invested index funds has completely transformed the way that institutions and individuals invest their assets. Unsophisticated investors have been given more access to the markets than ever before, and existing researches suggest that passive investing provides a better return on average than actively managed funds. Despite the growing popularity of passively managed funds, researchers have shown that emerging markets remain one of the few asset allocations where actively managed funds consistently outperform local benchmark index funds. As passive investing products become more sophisticated, researchers believe that active managers will continue to face competition from cheaper, electronically traded investment products. One such sophisticated product that has gained traction over the past decade is smart beta funds.

The smart beta funds examined are highly correlated to the MSCI Emerging Markets Index. In particular, JPEM, EMGF, and GEM deviate very little from the index as far as performance and monthly return variation. PXH and FNDE also display a high level of correlation in the index, but much less than the other three funds and both of these funds display a higher monthly average return and 2-year holding period return. Investors looking for a more unique investment in emerging markets are likely to consider these two funds over the other three.

The most notable drawback to this research is the timeframe. Two years of data isn’t an ideal timeframe for comparison purposes. There are 24 observations for each sample, based on the 24 monthly returns for each fund, and by increasing this sample size we would be able to have a more accurate picture of how closely these funds mimic the market. The period of January 1, 2016, to January 1, 2018, is a period of fairly stable, moderate growth across the emerging markets. In particular, a timeframe that includes a global market downturn would be an interesting addition to future regression models, as it would show how well these smart-beta funds are protected on the downside. The scope of this paper is limited in nature due to the restrictive time frame. Additional research will need to include a longer time frame once that data becomes available.

Future research could include a regression model that adds controls for each factor smart beta funds claim to utilize to more accurately assess how each factor affects overall monthly return. Instead of comparing smart beta returns to an index, researchers could also compare these funds to an aggregate of emerging market mutual funds. Comparing the net of fee returns could give investors an accurate picture of the value of utilizing a smart beta fund over a mutual fund, as well as the risks associated with a longer holding period.

Alexander Merryman received his MBA with a certificate in Data Analytics from Robert Morris University - Pittsburgh, as well as a Bachelor of Science in Business Administration with major concentrations in Economics and Finance. He currently is a Digital Product Manager at PIMCO.

Jianyu Ma is a Professor of Finance at the Robert Morris University - Pittsburgh. He received a PhD in Finance at the University of Texas-Pan America. His research appears in Int. J. Revenue Management, Int. J. Commerce and Management, Int. J. Business, Int. J. Business and Systems Research, among others. His current research interests include high-technology performance evaluation, international investment, and mergers and acquisitions.