This Wednesday Impinj (NASDAQ: PI) is expected to price 4.6m shares at $12-14. RBC is leading the charge with four other banks. (Everyone should read the Impinj prospectus and IPO Candy subscribers can peruse the Impinj IPO slide deck.)
We have to give the company and the banks credit for creative positioning. Their roadshow is all about them as the core technology behind the real "Internet of Things" movement. The truth is that Impinj is actually an RFID company. If you haven't heard of it before RFID stands for Radio Frequency Identification and it's been around for a long time. RFID tags are very small electronic devices (think grain of sand) that are passive but become active when strobed with a radio signal in the right frequency. They can then transmit information back to an RFID reader.
RFID tags have been used for myriad applications already - toll collection, tracking of goods, luggage handling, timing sporting events, access and ID cards, tracking animals and even insects for research purposes.
To their credit Impinj has a strong position in RFID but whether or not they should be considered an IoT player is more of an open question. IoT means different things to different people. With Impinj the sensors are passive. They can transmit information when strobed but are not programmable. Much of the thrust of IoT applications require powered and programmable sensors that actively gather information, process it to some level and communicate appropriately.
For example the huge acquisition of ARM Holdings (ARMH) is being justified based on the massive expansion of low-power computing embedded everywhere. In many cases with dozens or even scores of processors in a single system like an automobile. Most of the press coverage so far has highlighted large companies like ARM, Google, NXP Semiconductors (NXPI) where their direct exposure is positive but small in terms of their overall business.
There have been some more interesting opportunities in vertical industries like Fitness with Fitbit (FIT) and home security with Alarm.com (ALRM). There's even a very small horizontal play (which we like) in Adesto Technologies (IOTS). Adesto makes ultra-low-power memory devices that work well in IoT applications where power is a major limitations. Their design wins have accelerated dramatically, they've secured improved financing and are below their original $5 IPO price at just under $4. We'd encourage investors to take a look at that one.
Returning to Impinj it's not clear that they really belong in the IoT space. But there's even a more interesting way to think about Impinj which occurred after they had to write their prospectus and finalize their slide deck - Pokemon GO. Now we're not suggesting that Impinj is a Pokemon GO play, they are not. However the idea of embedding passive tags in things and places has much broader application in an augmented reality (AR) world.
Augmented reality has been advancing towards the mainstream of some time. In fact the first thought I had when the Pokemon GO craze hit was how Clay Shirky at NYU gave a talk on how the future wasn't about putting games on phones but about putting phones into games. He even did a field experiment with some of his NYU students that was presaged the Pokemon GO phenomenon.
Having digital sensors like the RFID tag in items and in places is a good way to get real things into the digital or virtual world. Of course there are other options including simple image recognition and QR codes but there are many applications where RFID tags are the only way to go since line of sight isn't needed and they are very efficient.
A big question is whether or not management sees the opportunity in AR. They do allude to consumer offerings that would "open the power" of being able to scan and use tags in any environment. However it's not clear if these would involve any radically new applications or use cases. With that Impinj is more about RFID and process automation rather than a real IoT play.
From a pure investment standpoint Impinj looks like an attractive IPO. Our IV suggests substantial upside from the proposed price. They have several advantages they highlight in their roadshow presentation (see the IPO Candy roadshow transcript for more detail):
- When it comes to RFID tags Impinj is a market-share leader. They are growing slightly above the industry growth rate so at least maintaining if not expanding market share.
- Their indirect multi-channel model is capital efficient. At relatively small revenue levels the company is profitable. With 50%+ gross margins and cost of sales close to 20% they can invest in R&D and remain profitable. We'd like to see G&A decline more as a % of sales but time will tell there.
- Because their sensors are so cheap many are consumed rather than reused. That means they have pretty good visibility around near-term business trends and quarterly performance.
- Impinj has a broad array of partnerships with OEMs (Avery, Zebra, Checkpoint, SMARTRAC), resellers (Siemens, Tyco) and smaller ISVs and VARs. This helps the company scale faster and more broadly geographically with less risk.
- RFID-based items offer solid ROI by making distribution more efficient and lowering inventory costs. With only 5% penetration in apparel alone the market has a long way to grow. Similar dynamics are in play in healthcare and the enterprise.
- Margins are solid and improving. As the business mix shifts towards more software and connectivity, in addition to high function end-points margins will expand further.
Valuation is reasonable. At the mid-point the market capitalization will be $240M which 2.4x our estimated 2016 revenues of $102M. Our IV estimate based on the longer-term potential of Impinj comes to $20.29/share versus the current $13 midpoint-of-the-range.
Below is our IV model:
Note to self: Should look more deeply into LOGM, CTRL, SSNI, PTC and DGII in the IoT space. Also put a comparable table together.