Kensho, founded in 2013, a well-known technology company dedicated to applying machine intelligence to disrupt existing commercial landscape has made some splashes on Wall Street recently by entering the index business, primarily thematic indices.
They gathered considerable steam by claiming the capacity to build trendy theme such as drones, space, robotics, cybersecurity, wearable technology, 3D printing etc. Early this year, they partnered with State Street and launched three ETFs based on Kensho indices: SPDR Kensho Intelligent Structures ETF (XKII), SPDR Kensho Smart Mobility ETF (XKST), and SPDR Kensho Future Security ETF (XKFS).
Their methodology is based on the New Economy Classification Standard (NECS). This white paper –Innovation Patterns: Upgrading Sectoral Classification
for the Fourth Industrial Revolution – touches on the genesis and gist of this NECS. They theorize into a hierarchy framework as below to capture key elements in the Fourth Revolution.
Then apply a “Bidirectional Mapping Algorithm for Deriving Innovation Patterns
● Step 1: Aggregate and parse technology trend reports and forecasts
● Step 2: Tally and rank technologies/trends from Step 1 by popularity (frequency)
● Step 3: Select top-ranked unmapped technology/trend; when none remain, then go
to Step 7
● Step 4: If technology/trend selected in Step 3 does not cleanly map to any of the
Innovation Axes, then discard it and go to Step 3; otherwise, go to Step 5
● Step 5: Decompose technology/trend selected in Step 3 into the Innovation Axes
for which it is most relevant by identifying how it is amplifying innovation
along each relevant Axis
● Step 6: For each relevant Axis selected by Step 5, perform the following: if no
other technologies/trends are mapped to that Axis, then proceed to the next
relevant Axis; otherwise, condense all technologies/trends currently
mapped to the selected relevant Axis by identifying the greatest common
denominator (GCD) among them, and saving this GCD before proceeding
to next relevant Axis; when no more relevant Axes remain, go to Step 3
● Step 7: End iterations and output condensed list of GCDs
This algorithm forms “The Kensho New Economies℠, consisting of a family of indices that capture the industries and sectors which are the driving force of economic growth in the so-called Fourth Industrial Revolution. The family includes industry level indices, such as autonomous vehicles, drones, cyber security etc, sector aggregates, such as Smart Transportation, Intelligent Infrastructure etc, which are the equivalent level of granularity as level 1 GICS sectors, and the New Economies Composite which is a tracking index of all companies involved in the New Economies” (page5 of the methodology). Index’s security selection is mainly based on the economic families pre-defined in this ECS supplemented by applying natural language processing to sift products/services mentioned in 10K/20Fs.
To examine whether this fancy method is a viable approach for creating thematic index, I randomly picked the 3D printing index on their website, top 10 components are: Proto Labs, 3D Systems Corporation, Stratasys Ltd., FARO Technologies, K2M Group Holdings, HP Inc., Flex Ltd., STMicroelectronics N.V., Autodesk, Eastman Kodak Company. Using RBICS mapping as a reference, I found only Proto Labs, 3D Systems Corporation, Stratasys Ltd. fit into 3D business, while HP, Autodesk are quite a stretch as they primarily rest on 2D printing, not to mention Eastman Kodak, an outdated Printing and Publishing Equipment Manufacturer. I suspect Kensho has been able to employ any “intelligent scouring” but literal textual searching, thus, such a misfit as Kodak will be chosen to be 3D printing stock.