“Properly deployed, AI-based thematic investment strategies can instantly create new strategies or power existing ones – and in a fraction of the cost and time that traditional analysis could yield.”
One of the persistent criticisms facing equity investors is their short-term view. They are characterized by adding or dropping stocks as the quarterly earnings roll in. Thematic investment, on the other hand, provides one counterpoint to earnings-focused stock picking.
Thematic investing—a strategy designed to capitalize on broad economic or social changes—has seen increasing use in recent years. In 2020, thematic funds accounted for the bulk of index fund growth, and topics ranged from artificial intelligence to cloud computing, cyber security, clean energy and cannabis.
Yet, asking questions about anything long term can be complex and often obscure. The trends and themes themselves that will reshape economies may be easy to identify, but translating them into quality investment vehicles is another matter. Using themes like clean energy, disruptive technologies, aging populations, or emerging markets to structure portfolios comes with its own unique challenges: successfully sifting genuine long-term trends from flash-in-the-pan fads – and critically, doing so early – is no easy task. Good analysis requires a massive amount of diverse data that, once structured, would facilitate thematic analysis.
A Knowledge Graph-based framework is uniquely positioned to provide both the data and analytic framework, with the inference capabilities necessary to provide actionable insights into large data sets.
A properly built Knowledge Graph describes the interrelations between real-world entities through a multidisciplinary, multidimensional correlated structure, comparing common themes and concepts across hundreds of millions of data assets over several years of correlated data embedded into the Knowledge Graph.
Such a framework can automatically calculate thousands of strategies for any investable concept an investor can think of – ranging from sustainability themes like clean energy to disruptive technologies like 5G or cloud computing.
A functional Knowledge Graph can rapidly build new, flexible strategies for thousands of concepts, deriving insights from millions of combined sources, and in ways that a typical analyst approach cannot match. Data sets can have global coverage – with strategies tailored to and applicable to multiple regions and countries – while also being highly specialized. They’re equally capable of taking in structured and unstructured data sets; everything from news reports, SEC filings, and financial or macroeconomic reports to court opinions and clinical trial data or patents. This multidimensional approach powers a dynamic point-in-time Knowledge Graph framework to produce exposure indices with precision.
Knowledge Graphs can further offer special insights in building a thematic investing portfolio through the way they look at concepts, both – quantitative (AAPL stock prices or its fundamental indicators for example) and qualitative. This offers not only the numbers behind what makes a wise investment, but also the context behind those numbers, which is especially critical when tracking themes.
Taking that capability a step further, it can also weigh data points based on the strength of their correlation to a given data set, or screen against undesirable exposure that might at first glance appear to be on theme. This scoring can be done at the entity level, offering sourced data on every point used in the process. When the process is complete, the final index that is produced has been weighed on multiple levels, accounting for variables such as market caps and liquidity for each company, and the aggregated exposures.
This type of analysis illustrates one of the key strengths of thematic investing: its concentration. Thematic investments are typically concentrated on a smaller selection of stocks, but a Knowledge Graph framework offers the opportunity to build thematic strategies based on a larger constituents basket. This pushes market analysis away from being a purely reactive prospect; through identifying anticipated changes in the world, investors can take a forward-looking approach to capitalize on opportunities as they are forming, leading to potentially greater long-term growth opportunities.
Contrast that approach with mutual funds, which are typically concentrated on 40-80 stocks in a portfolio. The emphasis is generally on diversification, which manages risk, but is not necessarily the optimal way to achieve growth.
Incorporating Yewno’s Knowledge Graph approach into a firm’s thematic investing portfolio creation has already been proven to be a game changer. The ETFs we have partnered on currently have an AUM over $1.2B, but more importantly, have easily outperformed most of their closely aligned peers. A perfect example of this would be our two Artificial Intelligence themed ETFs (Amundi Stoxx Global Artificial ETF, and XTRackers AI and Big Data ETF) have in most cases doubled, or even tripled, YTD performance when compared to other AI ETFs in the marketplace (all stats as of June 2021).
With a properly designed framework, a Knowledge Graph’s AI-based exposure engine can draw inferences to understand the dynamic market trends constantly driving returns while promoting concepts investors feel strongly about. Properly deployed, AI-based thematic investment strategies can instantly create new strategies or power existing ones – and in a fraction of the cost and time that traditional analysis could yield.