3 Pitfalls in Investing in Emerging Markets

Using Quantitative Techniques to Analyze Stocks

By GEORGE MUSSALLI/Guest Contributor

PanAgora’s Stock Selector Equity team seeks to understand companies by using quantitative techniques and unique data sets to systematically analyze stocks. PanAgora has developed proprietary quantitative models that serve to analyze a company’s strengths for nearly 25 years, initially managing U.S. equities then expanding into global equities, including emerging markets (EM).

Click chart to enlarge.

Exhibit 1 presents the decline of traditional factor performance within EM from January 2001 to June 2020. We believe generic factors, such as Quality or Value do not fully capture the return drivers within EM. Furthermore, our research has shown that the value of a company’s intangible assets can potentially be a key driver of performance.

When compared to developed markets, EM equities are typically exposed to greater uncertainties. These include geopolitical risks, liquidity risk and currency risks. While these are generally well-recognized by investors, in this paper we present three additional often overlooked EM risks. We also describe how we have sought to solve these challenges to create alpha opportunities. These risks we have identified are:

1. Volatility of economic activity
2. Increased accounting fraud
3. Quality of data

Pitfall 1 – Volatility of Economic Activity: Underlying emerging market economies are more volatile than developed economies, therefore one needs to have a strong sense of the story on the ground using real-time fundamental and economic data.

Fundamental managers typically have dedicated analysts that frequently travel to their respective country of coverage. This in-person research is key to their investment process. That said, scope and breadth of coverage is a challenge for fundamental investors given the sheer volume of information that must be included in any analysis. Instead of using analysts, we have invested heavily in collecting real-time data from EM economies to generate alpha factors. It is important to appropriately capture and reflect real-time data within our quantitative models, particularly in EM, given the region’s volatility.

One example of PanAgora’s data collection is within China. Real-time traffic data can be a leading indicator to the level of manufacturing. While most Chinese factories are in coastal cities, most labor tends to be from inland China. PanAgora has developed proprietary tools to monitor key cities’ real-time traffic data after Spring Festival to gauge potential factory order volume.

Pitfall 2 – Increased Accounting Fraud Risk: Unlike developed markets, financial statement quality is significantly lower within emerging market countries. Uneven accounting standards and frequent cases of fraud increase the need to independently verify company fundamentals and business conditions.
Relying solely on reported financial statements is particularly risky within EM. We believe that systematically analyzing companies for fraud is a key risk management tool within emerging markets, allowing us to mitigate exposure to potentially fraudulent companies.

Benford’s Law is a statistical tool that can be used to detect potential accounting fraud. Also known as the first-digit law, it observes the frequency distribution of the leading digits within any data set. Benford's Law predicts a fixed distribution of leading digits; the first digit Is expected to be '1' 30% of the time, '2' 18% of the time, etc. When accounting numbers are being manipulated, the distribution of leading digits is less likely to follow the distribution that is expected by Benford’s Law.

We use Benford's Law to measure each region’s accounting fraud risk. Exhibit 2 presents Benford's Law chi-squared statistic value across seven regions. The higher the chi-squared, the more likely the accounting fraud. As compared to developed markets, EM has a higher chi-squared statistic value. Given this, PanAgora has invested in techniques to independently verify financial results and business conditions.

Click chart to enlarge.

Pitfall 3 – Quantity of third-party data: Many data sets offered by third-party data vendors are prone to contain inaccurate or stale data sets that can negatively impact the potential for the data to be useful in constructing alpha factors.

In constructing a quantitative model, one of the most important components is consistent and high-quality data sets, as these alpha factors are only as good as the underlying data. Third-party data vendors have experienced tremendous growth in recent years. PanAgora has investigated utilizing these data sets but has too often found inconsistencies or errors within the data. For these reasons, PanAgora has invested considerable resources to limit our reliance on third-party data vendors and collects our own data directly from the source.

One example of this proprietary data collection within emerging markets is our collection of short interest data from South Korean exchange websites. Short interest is the number of shares that have been sold short by investors but have not been covered or closed out. We have found that utilizing our proprietary data collection techniques may provide us with short interest data two to three days prior to when it is available through a common data provider. Furthermore, by collecting the data ourselves, we can gather short interest data for a greater percentage of the South Korean equity universe and are not limited to the vendor's coverage universe. Our research shows that gaining access to data earlier and with greater coverage can potentially improve model accuracy.

We have identified three challenges that are not typically considered by EM investors. Additionally, we observe that there exists an opportunity to blend quantitative and fundamental approaches to create a hybrid that may be more powerful than the sum of its parts. PanAgora’s Stock Selector attempts to solve the paradox that many fundamental managers have an edge in selecting specific securities that cannot be translated across an entire broad cross-section, versus the quantitative managers that have the technological strength to test the cross-section but lack the skill to parse out the most relevant drivers for an industry. PanAgora’s ability to combine the depth typically seen in fundamental managers with the breadth of a quantitative process provided the impetus for our Stock Selector model.

PanAgora is an Associate Member member of TEXPERS. The views expressed in this article are those of the authors and not necessarily PanAgora nor TEXPERS.

About the Author

George D. Mussalli, CFA, is the Head of Equity Research and Chief Investment Officer of Equity Investments at PanAgora. He is responsible for the oversight of the firm’s Dynamic and Stock Selector Equity strategies, as well as the Equity Trading & Implementation, Data Science, and Portfolio Strategy teams. He is also a member of the firm’s Investment, Operating, Risk, and Directors’ Committees. Mussalli earned an MBA from the Sloan School of Management, Massachusetts Institute of Technology and a bachelor's degree from Tufts University.