Featured Image

Thematic Indexing 2.0 Part II: Think Like an Active Fund Manager

Data Science and AI

By Jeremy Zhou  |  December 29, 2020

Earlier this year I wrote about a systematic process using the concept of “relatedness” in constructing thematic indices and why it may be superior to the traditional way of construction, which depends heavily on human experts. Today, I want to introduce another important concept in Thematic Indexing 2.0, which is “selective cloning.”  

Selective Cloning

According to NIH (National Institutes of Health), cloning describes the “processes that can be used to produce genetically identical copies of a biological entity.” This requires a thorough understanding of the original genome targeted for replicating. Applied to Thematic Indexing 2.0, selective cloning refers to a process of analyzing and decomposing existing “active” thematic funds into their most essential attributes—their genetic makeup—and then attempting to clone these attributes into a set of systematic index rules.

To begin selective cloning, one must study how active thematic fund managers think and their thought process, which generally could be distilled into the following steps:

Step 1: Idea sourcing. Managers have identified trusted information venues where they could find and develop their ideas. These venues most often include certain industry contacts, newspapers and trade magazines, blogs, Twitter feeds, sell-side or buy-side commentaries, company filing databases, etc.     

Step 2: Solidify and consolidate ideas into themes. Managers have processed the information consumed and arrived at some conclusions. From these conclusions, they develop thematic views and potential ways to express them in portfolios—most often via the selection of a set of related industries, countries, and/or companies delivering specific products and services.

Step 3: Know yourself and focus. Managers have to be honest about their strengths and weaknesses as investors and the investment operations they run and focus only on themes and strategies that best match their strengths. If a manager’s expertise is in emerging market credit analysis but the theme requires implementing a growth-based developed market equity strategy, then there is a mismatch.

Step 4: Test, measure, and repeat. This applies to both new thematic ideas that have not yet been productized as live trading funds and existing live trading thematic funds. Managers must develop objective metrics such as specific risk-adjusted returns, portfolio volatilities, maximum drawdown, return variance from benchmarks, etc. for measuring whether the performance is good or bad.

Active Thematic Funds

To implement selective cloning, we’d compile historical holding snapshots of active thematic funds and analyze them according to the four steps outlined above. For example, if we were to develop a robotics thematic index, we might identify all existing active robotics funds and study their publicly disclosed holdings to see which industries, sub-industries, product categories, and countries these holding companies belong. We might perform additional factor, style, and asset class analyses of these active thematic funds if necessary, as well as their historical performance metrics.

Not all of these active selection processes could or should be cloned, namely because the active fund managers might have some information edge, an analytical edge, and/or trading edge; or we believe that portion of the active selection is not suitable in an index format. Hence, we used selective cloning to describe our Thematic Indexing 2.0 process because exact cloning is neither possible nor desirable.

Below you can find several live trading thematic ETF/ETN that use FactSet’s thematic indices as their underlyings where both the concept of relatedness and selective cloning were used during these indices’ development process. 


In conclusion, we are seeing the disruptive force of data and technology making its presence felt in the investment management industry, leading to further convergence of active and “passive” investing. Whereas thematic investing used to be thought of as an active endeavor requiring human domain experts and judgment, thematic indexing 2.0 challenges this view with its combined use of data science, quantitative techniques, and portfolio analytics. If indeed there is an active genome, the day of it being cloned could be fast coming.


Jeremy Zhou, CFA

Vice President, Head of Indexing Solutions

Mr. Jeremy Zhou is Vice President, Head of Indexing Solutions at FactSet. In this role, his main responsibilities include alternative data sourcing, index creation and partnership, and investment strategy consulting. Prior to joining FactSet in 2013, he was a senior executive with Revere Data—a leading alternative data firm acquired by FactSet—and held positions as Head of Index Solutions, Associate Director of Product Strategy, and Director of Healthcare Equity Research. From 2011 to 2013, he was also a Portfolio Manager for Covestor Ltd.—now part of the Interactive Brokers Group. Mr. Zhou earned a Master of Public Health and a B.S. in Nutrition and Toxicology from the University of California, Berkeley. He is also a CFA charterholder and a member of the CFA Society San Francisco.


The information contained in this article is not investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.