Adesto is more than meets the eye

Adesto (NASDAQ: IOTS) doesn’t look like much on the surface – flat revenues, consistent losses in a heavily commoditized business – computer memory. So it’s not surprising that they just reduced their proposed price to $7/share with their tier-2 underwriting team of Needham, Oppy and Roth. It’ll be a very small capitalization company at the new price – just over $100M.

But some further work is justified because underneath flat revenue growth there has been very rapid growth from new products with a steep drop in revenue from legacy products. At this point the revenue is almost all the new stuff. That means revenue growth will accelerate dramatically over the next few quarters. The ramp in design wins from 32 in 2013 to 65 in 2014 to what will be between 100 and 150 in 2015 makes the acceleration in revenue growth a given.

There also is a fundamental need for faster, much lower power electronics for embedded, wearable and mobile devices. Adesto has a technology advantage that allows them to serve this growing niche better than any of the larger competition. (see technology discussion below)

Unfortunately Adesto didn’t make their roadshow presentation public which is a bad decision and amplifies the limited reach a small tier-2 group of underwriters provides. So building a decent IV model for them isn’t easy due to the timing between design wins and revenue, price and volume trends and not even knowing what the target model is.

The memory business is a generally known quantity though so based on comparable companies and results we were able to put together a fairly solid model. The main parts we will have to watch for is how rapidly they can grow the top line, can gross margins get back to 45% and how much will they need to invest in R&D and S&M to support the growth.

Our IV suggests a near-term objective of $9 which would expand dramatically in 2017 if the company executes well enough for long-term numbers to be believable.

We published a research note on Adesto as a private company back in January of 2013. It’s worthwhile background and context for where the company is today. Here is a major excerpt:

Adesto Technologies which is a next-generation memory company that has now moved from the development stages to commercialization and expansion. We expect the company to capture a significant portion of the non-volatile memory outside of the commodity data storage market and enjoy profitable royalty licensing from a number of other markets.

Adesto has spent the last five years developing a non-volatile memory (NVM) that delivers an order of magnitude improvement in speed with lower power consumption. It’s a simple technology in concept: a solid electrolyte between two conductors can be “turned off or on” with a charge that creates or removes a conductive channel in the material. The invention is “obvious” but hard to execute. There’s an industry consensus that this is the most promising new emerging memory technology for mainstream applications. From the start Adesto has focused on commercialization by attracting strategic funding and licensing agreements, securing intellectual property, sticking to a profitable business model. This is the combination that can drive high consistent returns on invested capital. Presently the company is shifting from the development to the commercialization stage after investing $55M in capital from a mix of strategic partners and top-tier VC firms. The strategic investors and technology licensees have helped to develop the manufacturing process and have strong interests in end market adoption. Adesto acquired substantial revenues and relationships with nearly every major OEM from their purchase of the Serial FLASH business from Atmel. Having this product line and distribution channel provides a solid business foundation and accelerates the process of bringing the new technology into the marketplace.

The management team is made up of industry veterans with experience that spans many leading companies including AMD, Silicon Storage Technology, National Semiconductor, Broadcom, Transmeta, Cirrus Logic, Fairchild, AMCC, and Tower Semiconductor. Over the next 12 to 18 months we expect the company to sign additional licensing deals, and begin to experience a ramp in their revenue growth while margins expand at the same time.

Technology, memory and storage has a long way to go in order to deliver the performance, capability and efficiency we’re going to need in our computational systems. Complicating the challenge is the shift to mobile and wireless. This means that power consumption is as important as speed unless and until the “battery problem” is solved in some profound way. Memory close to the processor in mobile devices is a critical component for which disruptive advances in the speed/density/power envelope can create significant market opportunities for semiconductor companies. This helps to explain the dizzying array of R&D pro-jects aimed at producing the “ultimate memory.” There is a great chasm between the breathless descriptions of memory technology “breakthroughs” in R&D labs and devices that can be shipped by the millions and embedded in consumer products. It’s fun to talk about molecular memories and storing bits as spinning electrons but when the discussion turns to manufacturability, data retention, cost, stability and endurance, excitement evaporates quickly. The technology Adesto has been working on is called Conductive Bridging RAM (CBRAM) and has been gaining adherents around the industry (a few very large semiconductor firms have licensed the technology already from Adesto.) Adesto has tackled the key obstacle to mainstream CBRAM adoption, which is manufacturability.

The advantages from CBRAM come from simplicity and the relatively low impetus needed to create and destroy connections or “bits” of information. The lower impetus means greater durability and lower power requirements. Simple design improves density and per-formance. Adesto has worked through advanced manufacturing partnerships to develop the expertise needed for effective scale manufacturing. For example, Altis Semiconductor is an investor and manufacturer that is based in Europe and has key manufacturing platform expertise gained from a history involving both IBM and Infineon.

Market Opportunity Memory in general is a very large existing market on the order of $70B/year in sales. Ini-tially, however, Adesto will be offering discrete products in the lower density but “sticker socket” used for code rather than data storage. This part of the market is relatively small but still in excess of $5B/year. The licensing model of the Adesto business plan gives them access to the much larger (but highly capital intensive and competitive) data storage market. As far as we’re concerned this is the only way to go after the memory market and achieve high returns. Any market that is highly competitive, commodity-like and capital intensive is not good for equity investors. Going after the attractive low density segment and licensing the rest is the right approach. Adesto investors should look at the market in terms of the margin opportunity. By making their own devices in the lower-density code storage market and licensing in data storage the company can capture a meaningful total share of the available gross and operating margins in the total memory business over time. The next few years of execution will dictate the ultimate revenue opportunity and operating margins that can be achieved in the longer term.

Competition & Industry Dynamics No major innovation in the memory market is going to slip by unnoticed. Most of the large firms in the market and some that are not have ongoing R&D initiatives that are closely related to the Adesto CBRAM technology. Right now Adesto is recognized as the leader and has a substantial IP base upon which to build. However, we expect other memory-makers to offer at least something competitive.

In terms of semiconductor industry dynamics we’ve seen this many times before. Some examples include Invensense INVN (motion processing chips), Nvidia NVDA (graphics processing), and OmniVision OVTI (image sensors). All of these companies have competed against much larger, broader semiconductor firms and maintained their advantages to build meaningful independent public companies. We expect the Adesto timeline to show additional licensing agreements, next generation products and a strong revenue and profit ramp.

Probably the biggest risk to the story is that the major makers of memory have their own competitive products which are generally inferior to what Adesto has for special applications but they may become “good enough” to cap the TAM for Adesto and limit their ability to reach a scale where they can generate attractive margins.