By Tim Atwill, Ph.D., CFA, Head of Investment Strategy, Tianchuan Li Quantitative Analyst, and Mahesh Pritamani, Ph.D., CFA, Senior Researcher
It is this perspective which informs Parametric's Emerging Markets strategy. We observe that in the emerging markets, country-risk is paramount, and each country experiences high levels of volatility, and exhibits low correlation with other countries. Further, given the dynamic state of market leadership, where the best performing country one year can easily be the worst performing country the next, we maintain that assuming each country has a similar expected return is entirely reasonable. Taken together, these observations argue for the superiority of an equal-weighted country portfolio in the emerging-markets asset class. However, there are large implementation obstacles to holding an equal-weighted portfolio, given the high transaction costs and vast differences in market capitalizations across emerging-market countries. Accordingly, Parametric commonly refers to its strategy as a modified equal-weighted portfolio, which attempts to gain the advantages of the theoretical equal- weighted portfolio, while managing these implementation realities.
In recent conversations, we have encountered skepticism about the superiority of an equal-weighted portfolio in the emerging markets, as many investors view diversification as primarily a tool for volatility reduction, and not as a potential source of excess return. In this brief, we use historical data to show that an equal-weighted country portfolio in the emerging markets generates both lower volatility and higher returns than the market-cap index. We also show a similar result for frontier markets. It is this superior risk/return profile of the equal-weighted country portfolio that Parametric targets with its emerging-markets strategy.
Country Weighting: Equal Weight Versus Market Capitalization
For our analysis, we use MSCI country-level index returns and market-capitalizations. Each month, we then calculate two return streams from this data, which correspond to applying two distinct weighting schemes to the country-level index returns:
The first is the equal-weighted average of the month's country returns, for each country in the MSCI Emerging MarketsSM Index as of the prior month-end. This is equivalent to an equally- weighted country portfolio, rebalanced monthly back to equal weights.
The second is the market-cap weighted average of the month's country returns, for each country in the MSCI Emerging Markets Index at the prior month-end. Note that this will approximate the Index returns, but is not identical due to corporate actions and index reconstitutions which take place intra-month.1
These return figures are calculated for every month over a 15-year period from 2001 to 2015. We also perform the same analysis for frontier markets, but due to data availability, we can only analyze the 10-year period from 2006 to 2015. Note that transaction costs are not reflected in this analysis, as we are not claiming such a portfolio is implementable; only that it possesses desirable risk/return attributes. The results of these calculations are presented in Figure 1.
Figure 1 gives ample evidence of the power of equal weighting in emerging and frontier markets. In both cases, the equally-weighted portfolio outperformed its market-cap equivalent by over 3% (annualized), and, in both cases, this outperformance was achieved with lower volatility. These attractive characteristics come about entirely through the simple application of the naive equal- weighting diversification scheme, as the only difference between the two return streams was the choice of weighting. As such, Figure 1 demonstrates how diversification can be a material source of excess return in markets with highly differentiated and volatile country components.
To get a better understanding of the performance pattern between equal-weighting and cap-weighting countries, in Figure 2 we plot the annual performance differences between the two portfolios, for the case of emerging markets.
Click to enlarge For emerging markets, the equal-weighted country portfolio demonstrates remarkable consistency, outperforming the market-cap weighted portfolio in twelve of the fifteen years. The three years of underperformance correspond to notable outperformance by large constituents in the index, with Brazil's dramatic 128% rise in 2009, and China's market-beating rallies in 2013 and 2015 being the key detractors from relative performance.
For frontier markets, the equal-weighted portfolio demonstrates a quite different quality of consistency, as shown in Figure 3.
Click to enlarge In this case, the equal-weighted portfolio demonstrates a lower degree of consistency, only outperforming in four of the ten years. This reflects the extreme concentration which has been characteristic of the MSCI Frontier Markets Index, where a material part of the Index has been contained in just three to five countries. In addition, this concentration has primarily been in economies which are heavily reliant on the export of crude oil for their economic growth. Specifically, the top five constituents have typically included U.A.E, Qatar, and Kuwait (all Persian Gulf states) and Nigeria (the largest African exporter of crude oil). This dynamic in frontier markets makes the excess return of the equal-weight portfolio more of a "feast or famine" situation, dependent upon the relative strength of the oil-based economies. The reduced consistency in returns reflects this fact, but we note that the magnitude of the equal-weighted portfolio's outperformance dwarfs that of its underperformance, resulting in the equal-weighted portfolio outperforming strongly over the entire sample period.
Why do Equal-Weighted Portfolios outperform?
There are many viewpoints as to the source of the above outperformance. As mentioned, it may be the natural outcome of modern portfolio theory, applied to an environment with attractive volatility and correlation patterns which make it ripe for benefiting from diversification. Another perspective is to note that an equal-weighted portfolio implicitly assumes a monthly rebalancing trade, and so this outperformance may simply reflect the harvesting of a rebalancing premium2. A further perspective sees this outperformance as simply mining the "small country premium", as smaller countries have historically outperformed versus larger countries in the emerging markets3. It is unlikely that any of these are complete explanations, but instead each thesis focuses on different aspects of the same phenomena: that diversifying a portfolio's country exposure, and then rebalancing to keep this diversification intact, can yield superior performance in the emerging markets.
At the outset of the paper, we noted that we have encountered skepticism about the superiority of an equal-weighted portfolio in the emerging markets, as many investors view diversification as primarily a tool for volatility reduction, and not as a source of excess return. We have shown, using historical data, that an equally-weighted country portfolio in the emerging markets generates both lower volatility and higher returns than the market-cap index. Parametric's Emerging Markets strategy follows a similar investment philosophy as the equal-weighted methodology presented here and differs from it only for pragmatic implementation reasons. It uses a tier-based weighting scheme to increase country-level diversification while reflecting the vast dispersion in country market sizes. It also uses a trigger-based rebalancing methodology to maintain this increased level of diversification, while acknowledging the high transaction costs encountered in the developing world. In this way, the strategy tries to lock in the risk/return benefits of equal-weighting countries demonstrated in Figure 1, while avoiding the elevated implicit and explicit costs such a naive portfolio would incur in reality.
1. Our monthly data series tracks the performance of the official MSCI Emerging Markets and MSCI Frontier MarketsSM Indexes closely at an annualized tracking error of 0.18% and 0.85% respectively. As the objective of the paper is to show how changing the country weighting scheme from cap-weighting to equal-weighting affects performance and risk characteristics relative to one another, the fact that our data doesn't track the official index returns perfectly is not a source of concern.
2. Bouchey, P., Nemtchinov, V., and Wong, L., "Volatility Harvesting in Theory and Practice", Journal of Wealth Management, Vol. 18, No. 3, Winter 2015: 89-100.
3. Li, Tianchuan, & Pritamani, M., "Country Size and Country Momentum Effects in Emerging and Frontier Markets", The Journal of Investing 24.1, 2015: 102-108.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.