Creating thematic index/ETFs is itself a big theme now. Many firms are launching or planning to launch such products fearing if they are left behind, the early players will seize the large portion of the market share.
Working shoulder by shoulder with the largest firms to tackle these, we face two challenges: one is the accuracy and the other is the speed – speed of developing the product in the first place and speed of maintaining, rebalancing on an ongoing basis.
This blog aims to address the very first step of coming up with a rough initial universe of a theme. I summarize the themes into three broad categories: 1. traditional themes such as “luxury”, “automobile”, “natural gas”, 2. forward-looking themes such as “cloud computing”, “future mobility”, “artificial intelligence”, and 3. theme of themes such as Chinese “one belt one road”, “mobility as a lifestyle”.
- traditional themes are relatively easy and straightforward. Given a sound and granular industry classification system like GICS, anyone can boldly claim to build such a thematic index. Equipped with FatSet’s RBICS, early adopters such as Singapore Stock Exchange, STOXX and ourselves are able to leverage the primary line of business mapping to filter related companies in a quick manner. The ETF RBOT applied this approach by sifting the below 35 sectors at level6 of RBICS:
- Forward-looking themes are a bit more difficult to compose because of two reasons: one is that the taxonomy is usually lagging behind the true market status. It won’t have sectors such as blockchain, artificial intelligence included when these are trending words; two even the nodes are created in the taxonomy, companies usually won’t be primarily relying on or generating revenue from these novel businesses. To tackle this task, RBICS with Revenue and RBICS with tradenames can play a vital role. For example, the NYSE FactSet Artificial Intelligence Index we co-created for BlackRock to issue an ETF contains both RBICS focus and RBICS with Revenue. Referring to this blog https://naixianzhang.com/2018/04/23/two-more-indexes-just-launched-themed-artificial-intelligence/, you can see that we defined market leaders in this AI arena, so big companies such as Google, Facebook who have deployed certain/auxiliary resources in AI are ferreted out.
3. The theme of themes is broad and vague if we attempt with the above sector-based angle. Use “future mobility” as an example, this concept contains a wide range of industries/sectors: Alternative Energy Car Manufacturers, Powertrain Manufacturing, Heavy-Duty and High-End Batteries Manufacturing, Web Search Sites and Software, Global Positioning Systems (GPS) Manufacturing … covering not only EV makers but also EV suppliers, key battery suppliers, autonomous driving software and hardware providers. In order to cast a net and fish out relevant companies fast, we apply robo-indexing approach by starting from a list of seed companies, then decompose into relationship database to calculate relationship relevancy score, then condense to a list of related sectors, to further calculate sector_focus_score and sector_with_revenue_score. This Robo-indexing methodology will be further enhanced by adding more factors such as similarity from analyst coverage (one or a group of analysts usually cover a host of companies of same industry), market cap, price correlations, P/E, Sales/EV values etc.
The short script to find seed companies from analyst coverage are as follows:
</pre> with allt as (select r.factset_entity_id, e.entity_proper_name, ana.* from (select a.factset_person_id, d.entity_proper_name 'Participant Company Name', e.entity_proper_name 'Analyst', a.report_id from evt_v1.ce_participants a left join evt_v1.ce_reports b on a.report_id=b.report_id left join evt_v1.ce_events c on c.event_id=b.event_id left join sym_v1.sym_entity d on d.factset_entity_id=a.factset_entity_id left join sym_v1.sym_entity e on e.factset_entity_id=a.factset_person_id where a.participant_title = 'Analyst') ana, evt_v1.ce_reports r, sym_v1.sym_entity e where r.report_id = ana.report_id and e.factset_entity_id = r.factset_entity_id ), groupanalyst as (select * from allt t where t.entity_proper_name like 'ASML%' --or t.entity_proper_name like 'Broadcom%' ) select distinct a.factset_entity_id, a.entity_proper_name, a.analyst from groupanalyst t, allt a where a.Analyst = t.Analyst <pre>
Finally, to touch on this theme of theme project – “One belt one road” is another approximate example, according to Quora, “it is a project initiated by the Chinese President Xi Jinping. Its objective is to build trade routes between China and the countries in Central Asia, Europe, and Indo-Pacific littoral countries. OBOR/ BRI is a network of roads, railways, oil pipelines, power grids, ports and other infrastructural projects meant to connect China to the world both land and sea-based. Six economic corridors and one maritime route have been proposed under the OBOR:
1. New Eurasian Land Bridge. (connect Western China to Western Russia)
2. China – Mongolia – Russia Corridor (North China to Eastern Russia via Mongolia)
3. China – Central Asia – West Asia Corridor (Western China to Turkey via Central and West Asia
4. China – Indochina Peninsula Corridor (Southern China to Singapore via Indo-China)
5. China – Pakistan Corridor (South Western China to and through Pakistan)
6. Bangladesh – China – India – Myanmar Corridor (Southern China to India via Bangladesh and Myanmar)
7. Maritime Silk Road connecting Coastal China to the Mediterranean via Singapore-Malaysia, the Indian Ocean, the Arabian Sea and the Strait of Hormuz.
China has planned around $1 trillion of investment in various infrastructure projects by providing loans to the countries involved at a low cost.” So in a nutshell, this theme is more or less belongs to the first category, traditional infrastructure theme with a universe spreading beyond Chinese stock market…