When Nobel Laureate Eugene Fama first formed his efficient market hypothesis, he argued that outperformance through active management was impossible—that all stocks always trade at fair value and cannot be bought at a discount or sold at inflated prices.
Nowhere is this less true than in emerging markets.
While emerging market investing is a constant source of contention and speculation—over the right time to buy, the right level of exposure, and the right weights to give individual countries and sectors—scholars, managers, and allocators alike at least agree on one thing: emerging markets are far from Fama’s ideal.
“Compared to the US, which is the most efficient [market] around, emerging markets are very inefficient,” says Geoffrey Wong, head of emerging markets and Asia Pacific equities at UBS Asset Management. “Information is less baked into the price. It’s the last place you should index—the best opportunity for active outperformance.”
But to capitalize on this opportunity requires filling in the blanks—in other words, obtaining the information that isn’t baked in.
“Some managers will outperform and some will underperform,” says Osman Ali, managing director at Goldman Sachs Asset Management (GSAM). “The defining feature of the managers who are able to consistently outperform, I think, is a consistent ability to have an informational advantage.”
In international investing, informational advantage has long been associated with having a physical presence on the ground—whether it be a full office or an individual analyst—studying the local markets. UBS does this just this: Wong says his team sends analysts to developing countries to question locals on buying decisions to better map out consumption trends.
But increasingly, managers like Ali—a senior portfolio manager with Goldman’s quantitative investment strategies team—are gaining an edge in emerging markets without racking up frequent flyer miles. They’re relying, instead, on big data.
Big data—a combination of nontraditional data sets and computational and analytics tools such as algorithms and machine learning—has slowly infiltrated investment processes over the last decade. Hedge funds, investment banks, and, according to alternative data provider Thinknum, even some of the most “forward-thinking” pension funds have adopted these data-driven strategies.
“It used to be hedge fund analysts and portfolio managers were going to stores and jotting down prices, or asking management how many employees they had and whether they were hiring,” says Justin Zhen, Thinknum’s co-founder. “Now all of this data is online, and we can build programs to automatically index that information.”
At Thinknum, Zhen and fellow co-founder Gregory Ugwi have catalogued unique data sets on hundreds of companies, public and private, by drawing from information publicly available on the internet: business locations, job listings, airline traffic, and retail prices. If it’s online, Zhen adds, it can be indexed and applied to investment analysis.
Take Facebook check-in data for example. Thinknum has found the number of users checking into a restaurant on Facebook to be highly correlated with that restaurant’s reported revenue—and, like most alternative data sets, it’s available before companies report earnings. “These data sets can be extremely predictive of how a company is doing,” Zhen says.
“Nontraditional data points can give investors an edge,” agrees Emmett Kilduff, founder and CEO of Eagle Alpha, a one-stop big-data shop providing data-driven research, analytics tools, and a directory of alternative data sources. In emerging markets in particular—where officially provided data is sparse and sometimes unreliable—Kilduff says alternative data can offer an “unbiased view,” if you can get ahold of it.
“There’s definitely less data in emerging markets than there is in developed markets. Alternative data isn’t going to be as easy to access there,” he adds.
Firms like Thinknum and Eagle Alpha that mine data from the internet, for example, will not be able to gather as much intelligence in countries where internet usage isn’t as widespread. According to a recent UBS report, internet penetration in emerging markets is “very low,” reaching less than half the population in nations like India and South Africa.
At GSAM, most of the data Ali and his team use is either fundamental in nature—“metrics that identify good businesses that are growing and stocks that are valued appropriately,” he explains—or focused on sentiment.
“Whether people like them, whether people are talking favorably about them, whether they are exposed to the right themes and trends, whether similar companies are doing well—there is more technical analysis of looking outside the four walls of a company and trying to see how other companies and investors are reacting around them,” Ali continues.
This data could come from web traffic patterns; it could come from news coverage; it could even come from computers programmed to read analyst reports and earnings call transcripts to quantify sentiment and tone.
“These are types of data that are a little unstructured, meaning it’s text and sometimes it’s hard to get, but when you put the effort into it, you can get a lot of analysis out of it that’s helpful,” Ali says. “Big data is a popular phrase right now, but what it means to us is that we’re using larger alternative data sets where we can capitalize on having computers do some interesting analysis like machine reading and machine learning.”
This type of information and analysis can “complement a ‘boots on the ground’ approach,” says Eagle Alpha’s Kilduff. But it may not be enough to replace traditional information sources entirely.
“The data in emerging markets is not clean,” says Roz Hewsenian, CIO at the $5.5 billion Helmsley Charitable Trust. “There’s still a lot of work that has to be done in order to make it useable.”
While the foundation chief uses a data-driven manager in her emerging-markets portfolio, she also employs a manager that sends research analysts to developing countries to meet with companies in person. Both managers, she says, are limited—the quantitative manager by the lack of quality data, the qualitative manager by time, travel expenses, and unreliable sources.
“Both approaches work, but neither is perfect,” she adds. “You have to diversify your sources of information.”
But as internet usage and e-commerce in emerging markets expand—something UBS’s Wong predicts will occur at a rapid rate—big data will only continue to grow. Investors would be ill-advised to ignore it.
“In the world we live in right now, most people feel like there’s an enormous amount of information and data out there than can and should be used,” Goldman’s Ali says. “Regardless of the underlying investment process, the use of more and more data to get a better informational advantage is a clear trend that will only continue.”