Description
For investors willing to stretch their investment horizon 24+ months. I think Impinj "PI" provides an extremely compelling risk-reward. PI designs and sells an integrated circuit ("IC") in a duopoly market that has grown at compounded annual rate of over 30% the last 8 years and has the potential to grow multitudes of its current size.
Assuming the market continues to just hum along (20% unit growth), but never re-accelerates to "hyper-growth" which I define as around 40+% over a multi-year period - I think PI on a private market-basis has limited downside and trades around 7x 2019 owner's earnings to a strategic acquirer.
PI provides a RAIN radio frequency identification "RFID" tags, reader chips, readers, and software.
RAIN is a technical standard for RFID. While anyone could build RAIN tags as the standard is public, the market is concentrated between two players PI 60% and NXP 40% due to the complexity and processes of building a high performing IC within the standard.
PI RAIN tag sales have grown at a 36% CAGR over the last 7 years from 900 million tags to over 6 billion in 2016.
A tag is really just an electronic bar-code that can be read via handheld or fixed readers over a distance of up to 30 feet.
Tags are sold to value-added resellers for a little over a cent.
A reader chip enables a device to read a tag – these can be handheld readers, or fixed gateways. PI has the dominant shareof the reader chip market.
These have ASPs of about $15+.
They also sell fixed gateways that have ASPs of $100s-$1,000+.
RFID is certainly not new, but the movement to a standard – RAIN – and the inevitable ubiquity of connectivity has made RAIN tags an extremely compelling tool in multiple industries which can be seen from total tag deployments of 1.6 billion in 2010 to over 9.5 billion last year.
The largest use case today is in the apparel industry – where there are 8.5 billion units currently tagged out of an approximate market of over 80 billion apparel units.
If you talk to people in the industry the view is that at some point the majority of this apparel will be tagged. The reasons are primarily the following: item level intelligence is foundational for omni-channel retail, and at a certain threshold, 30% from my conversations, apparel suppliers move to tagging all of their inventory even if not requested by the retailer as it enables them to use the benefit of tagged inventory and mitigates need for running two supply chain processes.
In their "2016 State of RFID Adoption Among U.S. Apparel Retailers" they note the process for deploying RFID is as follows:
Proof of concept where a retailer over 3-12 months in 1-2 stores will trial a few merchandise categories.
Pilot a 1 year process across multiple stores and more categories.
Phased deployment where a retailer rolls out fixed infrastructure and tag all merchandise in a store and begin rolling out this full solution.
Full deployment fixed infrastructure and all merchandise tagged in every store.
What is fascinating is that in 2016 – full deployments grew 100% with 4% of retailers in full deployment, while phased deployments grew 18% with 39% of retailers in a form of phased deployment. Pilots grew 57% and proofs of concept grew 42%. Every large retailer is in some part of this process. To me, it is clear at some point, this industry is going to massively grow.
Other large industries utilizing RAIN RFID are logistics, supply chain, and healthcare. For example, Delta Airlines tags every checked-in bag utilizes fixed readers across its in-terminal infrastructure to provide passengers real-time updates on their luggage. In healthcare sample and pharmaceutical tracking are paramount to hospital operations.
One of the most interesting industries where it is not yet used, but is seeing increased interest is in food retail, and according to management this has progressed faster than many had anticipated. In Japan, due to labor shortages, all of the C-Stores are making investments to RFID enabled stores. Every item in the store will be tagged and the checkout process fully-automated. The Japanese C-Stores sell 100 billion items annually.
In tags, PI goes to market exclusively through in-lay providers, basically value-added re-sellers who take PI tags and embed them into apparel. The majority of their business is concentrated among Smartrac 16%, Avery Dennison 14%, Arizon 11%, Shang Yang 10%, and Blue Star 9%.
Avery Dennison, which is public, has discussed their RFID business for years, and has consistently forecasted the market significantly below its actual growth. Today they are still very bullish and see a long-runway of 15-20+% growth primarily due the large visibility they have into retailer deployment pipelines which is corroborated by the Auburn report.
Why does this opportunity exist?
In 3Q PI had disappointing earnings, one might even say very disappointing as the stock dropped 40%. The biggest concern as a long-term shareholder is the accelerated price erosion in their IC tags. The company has consistently stated they see price erosion around mid-single digits and that had edged up a little in the latest quarter. Coming off a year where they had 70+% unit growth and had some shortages into a year where there were some push-outs in deployments (management explanation) there was a little extra discounting on IC prices by the competition. The biggest risk to this investment in my opinion is accelerated price erosion and while this is worrisome from many industry conversations do not see this as a trend and management explanation seem reasonable.
The holy grail would be a moving reader chips into consumer electronics – Alexa's, your fridge, cell-phone etc. While this is 3+ years off – talking to people it is a logical place for this market to go. Moving item level intelligence into people's homes – understanding what they have purchased and when those items are no longer in there – ie. there is was a milk container in the fridge and now there is no milk in the fridge is quite interesting.
PI is making massive investments as they believe this is a market that will 5+x over the next 5 years. They are investing in all layers of their solution – with an emphasis on software that today does not contribute any revenue to their business.
I think either this is an industry that will either be 5+x as large over next few years, at 55+% gross margins, and they will have a tremendous amount of operating leverage or this kind of ho-hums along for sometime before it gets sold and rationalized.
R&D has grown from $10 million in 2013 to a run rate of $30+ million
S&M – is primarily used as they say they for hunting whales and continuing to establish this industry. They work with massive potential customers on how to deploy RFID solutions and this line item has grown from $10 million to a run rate of over $30 million.
G&A – they have built a full public company operation going from $5 million in 2013 to a run-rate of over $16 million.
There has been lots of consolidation in semiconductor space and my view is that If this market peters out then the need to keep pushing envelope on R&D and S&M is just not there. I don’t think it's crazy to bring this down to their 2013/2014 expenses structure.
Primarily stop investing in software, and move to sustaining R&D of $10-15m, $7-10m of S&M, and $5m of G&A.
If market continues to grow units 20% with 5% ASP declines with this restructured cost base get to about $60 million of EBIT.
The upside is really hard to quantify – but if end-points double over next 3 years then this business is trading 2-3x gross profit and massively growing – it is just way too cheap
I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.
Catalyst
Resumption of growth