Skip to content

Why we believe investors can't afford to ignore emerging markets

Apr 20, 2018

Emerging markets have long attracted investors seeking stronger growth, higher income and alpha opportunities in less efficient markets. But many were disillusioned by underwhelming returns in the post-crisis years. More recently, fund flows indicate that some investors are returning to the asset class or upping their allocations.

In the long run, we don’t think investors can afford to ignore EM. Emerging and developing economies are home to 85 percent of the world’s population—some six billion people, according to 2016 estimates form the International Monetary Fund (IMF). And in its 2018 World Economic Outlook, the IMF estimated that these economies will grow at more than double the pace of advanced economies in the coming years.

Given emerging markets’ growing importance, coupled with our belief that many investors may be under allocated to EM equity and debt, we thought the time was ripe to explore some of the opportunities, and the risks, that this vast and varied asset class presents.

To that end, we have created this collection of emerging market briefings—Q&As with BlackRock portfolio managers focused on different opportunities within the broad EM landscape. We begin the series with Jeff Shen, BlackRock’s Global Head of Emerging Markets and Co-CIO of Active Equity, who discusses how a systematic approach to equity investing may help investors navigate the EM opportunity set. Future briefings will include an overview of EM debt, a deep dive on local currency debt and in-depth looks at the opportunities in Chinese A-shares and frontier markets.

We’ll also produce a whitepaper covering the opening up of China’s onshore equity markets, as we believe that this represents a landmark event, and an important potential opportunity, for EM investors.


Q: What is the case for a long-term, systematic approach to investing in EM equities?

A: I think strong economic growth forms the foundation of a long-term investment thesis. There are a lot of EM companies—we analyze more than 2,000—that can potentially benefit from that growth, but we’ve historically seen a wide dispersion of returns within equity markets. Not all firms will prosper from the macro tailwinds that are propelling EM, so I think the ability to identify winners and losers will be critical.

But as the world becomes more interconnected, technology changes at a breakneck pace, and the amount of data to process grows exponentially, it becomes increasingly complex to analyze companies. I think this is particularly true in emerging markets, where some technological changes are playing out even faster than in the developed world.

Q: What’s one important thing that is happening faster in EM than in DM?  

A: I think the most salient example is artificial intelligence (AI). In 2016 China published more academic papers on AI than all of the countries in the EU combined, and it overtook the U.S. in terms of the number of AI papers in the top five percent of most cited research, according to a 2017 analysis by the Financial Times.

Last summer the Chinese government outlined its goal of creating a world-leading, U.S. $150bn AI industry by 2030, and the private sector is upping the ante as well: Alibaba announced in October that it would spend U.S. $15bn over the next three years on AI, quantum computing and fintech.

If you think about how AI can be most impactful, it is by intersecting with the maximum number of people. This is true both in terms of gathering data to inform the I part of AI, and in utilizing that data to impact human behaviors.

If you’re trying to sell a product or a service to Chinese consumers, you want to figure out how to provide solutions to 1.4 billion people while maintaining efficiency of scale, and the more you can leverage AI, the more of a potential edge you can gain. We see a number of EM companies working on solutions to these challenges, and we think they present some intriguing long-term investment opportunities.

Q: Many investors were originally attracted to emerging markets in part because they were viewed as inefficient. As the world grows more interconnected, is EM still an inefficient asset class?

A: The short answer is yes. But before I delve into that, I want to dispel a common misconception. Many people believe that because EM is inefficient, it’s easy to extract alpha. It isn’t. In terms of alpha added, the long-term performance of a median EM manager is not that different from that of a median U.S. large cap manager. See the table. And just as we see wide dispersion of returns within EM equity markets, we also see wide dispersion of returns across EM equity managers.

No easy pickings

Annual alpha added by active management

Annual alpha added by active management

Source: evestment, 10-year data as of 12/31/2017. For 5th and 95th percentile, performance represents the average of the managers in these brackets. Considers all 274 managers in the evestment U.S. Large Cap Core category and all 26 managers in the evestment EM Large Cap Core category with 10-year track records as of 12/21/2017. Benchmark for U.S. large cap is S&P 500; benchmark for EM is MSCI EM Index. Alpha represents returns in excess of benchmark. Past performance is no guarantee of future results. You cannot invest in an index.

While we still view emerging markets as inefficient, particularly compared to developed markets, we’ve seen a dramatic shift in the tools needed to exploit those inefficiencies. If you look at well-known signals like analyst revisions, those now get arbitraged out of the market very quickly. So as some of the old signals have gotten commoditized, we’ve turned to a new generation of signals to try to exploit market inefficiencies.

Q: What are some of the newer signals that you find particularly useful in EM?

A:  I’ll give you two examples. The first is utilizing satellite imagery to gauge economic activity. In the past, if you wanted to measure things like how many trucks are leaving a warehouse or how many people are visiting a store, you needed people on the ground to tally those metrics.

Over time, many companies in developed markets started reporting these, or similar, numbers because the investment community was interested in them, and in the process the playing field was leveled. But in many emerging markets, access to this kind of data is limited, and it isn’t practical, or cost-effective, to put boots on the ground in 24 countries. So instead, we’re using satellite images. I like to think of this as using bits (the zeros and ones in satellite images) instead of atoms (people on the ground), and the idea of bits over atoms is a big theme across our research.

The second example is natural language processing (NLP), which we use across both developed and emerging markets to analyze things like investor conference calls and analyst reports. But in emerging markets we also find NLP particularly useful to look at sources such as political speeches and articles in state-owned publications which, as you might imagine, can have a big impact in countries where government policy has outsize influence over financial markets.

These are just two of the signals that have come out of our research in the fields of big data and AI, and while I think those are tremendously promising fields, I don’t want to give the impression that they can give investors any kind of magic bullet. Big data and AI present us with a lot of opportunities, but also a lot of challenges. It’s extremely difficult to process and analyze satellite images and turn them into investment insights, and NLP algorithms need to sort through a tremendous amount of noise to find a signal. 

Q: What are some of the risks that you are most concerned about, and how do you attempt to mitigate them?

A: I’d put geopolitics and the potential for trade wars at the top of the list. As far as trying to mitigate those risks, of course we use scenario analysis and stress testing to see how portfolios might perform under different types of market shocks. But we also use natural language processing to perform systematic analyses that can provide additional insight.

For example, we can use text analysis to examine news feeds and brokerage reports to see how frequently they mention certain geopolitical risks.  See the chart. We can then factor the results into our investment process by either adjusting our risk budget or by acting on any potential opportunities that the data points to.

Ups and downs

BlackRock Geopolitical Risk Indicator, 2005-April 2018

BlackRock Geopolitical Risk Indicator, 2005-April 2018

Sources: BlackRock Investment Institute, with data from Thomson Reuters and Dow Jones, April 2018. Notes: The BGRI tracks the relative frequency of terms associated with geopolitical risks within the Dow Jones Global Newswire database and Thomson Reuters Broker Report database, adjusting the score for positive or negative sentiment in the text of articles. A positive BGRI score means market concern about geopolitical risk is elevated compared with recent history. A negative score indicates historically muted market concern and zero represents approximately neutral concern relative to recent history. The indicator is still under development and is meant for illustrative purposes only.

Q: Is there anything that you’re particularly excited about for the future?

A: There are many things, but if I had to pick one I’d say unsupervised machine learning. We’re doing a great deal of research here and the possibilities are fascinating.

With supervised machine learning, we have theories on the relationships between the things we are attempting to forecast. If we’re trying to predict future earnings based on management conference call scripts, we already have a reasonable sense that if management uses upbeat language, that could potentially indicate positive analyst revisions in the near future. Essentially we’re using machines to answer an existing question: how likely is it that analysts will upgrade their earnings estimates for a particular company?

With unsupervised machine learning, we’re analyzing large datasets and trying to discover interrelationships that we humans haven't really considered. In essence, we’re looking to machines to help us form new questions that we’ve never thought about asking. We talked earlier about bits over atoms, I like to think of unsupervised machine learning as questions over answers.

If we get to a future in which it’s not the investors with the best answers who have the most power, but those who ask the best questions, that would represent a true paradigm shift.

Jeff Shen
Global Head of Emerging Markets and Co-CIO of Active Equity
Read more