Silicom (NASDAQ:SILC) is an undiscovered gem that produces networking hardware for many well-known OEMs such as Cisco, F5, Citrix, Check Point, etc. The company’s main product line is network interface cards (NICs) that provide connectivity and performance enhancement to networking devices, such as firewalls, intrusion prevention boxes, application delivery controllers, storage devices, network management boxes, etc. Its cards go into empty slots in the network devices and provide flexible connectivity to the network as well performance boosts to the CPU. The great thing about Silicom’s business model is that the company grows as its customers grow. It then enhances this growth by adding more customers and selling incremental new products. Another positive feature of the business model is that its relationships are very sticky. The company has more than 120 customers and has never lost a customer since inception. Design cycles are long but once Silicom gets designed into a product, that product usually has a 3-5 year life cycle. The company is intensely focused on customer service. Silicom engineers are in close contact with its customers’ engineers and as a result, they learn what the customers want, resulting in a high hit rate for Silicom’s new products.
SILC has experienced waves of growth over its history driven by product cycles at its end users. There was a wave of growth in 2005-2007 driven by the emergence of WAN-Optimization boxes, which basically allowed end users to send more data down existing data connections and thus save money on bandwidth. Silicom had cards that went into all of the major WAN-Optimization players’ boxes so the company was able to piggyback on this growth trend. Silicom experienced another wave of growth from 2011-2014 as its cybersecurity customers showed strong growth driven by the need of their customers to protect their networks.
Our thesis is that Silicom is poised for another wave of growth today driven by three product trends. The first is the trend toward cloud computing. As more and more data and computation is done in the cloud, the cloud players need solutions such as Silicom’s for connectivity and CPU offload. Taking some of the load off of the main CPU onto Silicom’s NICs allows the CPU to do its intended job more effectively and quickly. Silicom is working on business with large cloud providers to garner increasing sizable design wins. In the spring of 2017 Silicom announced a $25 million design win with IBM which was upsized to $75 million that fall. This was for an innovative design where data would be switched on each server rather than on separate switches in the data center. Unfortunately for Silicom IBM got cold feet at the last moment and pulled the product right before launch. That explains the decline in Silicom’s share price from $75 to $35 in 2018. This does not mean that Silicom’s cloud effort is dead. It is working on sizable business with many other cloud vendors currently and will have announced design wins as time goes by.
The second revenue driver that Silicom is experiencing is for a product called Universal Customer Premise Equipment or uCPE for short. This is a type of box that goes into branch offices and functions as a router, a firewall, and an SD-WAN device. SD-WAN is a technology that simplifies and cheapens enterprise communication between branches and into the Internet. This is currently a very hot area where all of the telcos are preparing offerings if they don’t have them already. Silicom is winning the lion’s share of these designs, with the large telcos such as AT&T and Verizon, as well all the software OEMS such as Versa, Velocloud, Viptela, and Silver Peak. The company already won $30 million plus in design wins between AT&T and Verizon and recently added a $15 million design win with one of the major SD-WAN OEMs.
The final current growth avenue for Silicom is FPGAs (Field Programmable Gate Arrays). These are types of semiconductors that have greater firepower than normal CPUs and are thus well positioned for uses in artificial intelligence and driverless cars, as well as other instances of CPU offload. Silicom purchased a company three years ago that focused on FPGA cards and has furthered developed its offerings so the company is very well positioned for the increasing uses of FPGAs. It announced an initial $4 million FPGA design win in January for a cloud operator who will use the product for application acceleration globally. Recently the company announced a potentially giant win with a leading Autonomous Vehicle company who will be manufacturing the FPGA cards itself but will pay a 100% gross margin royalty to Silicom on each card. Silicom is agnostic to the FPGA manufacturer, working equally with Xilinx and Altera (Intel).
Silicom currently has a market value of $230 million and $89 million in net cash. It is on pace to do $108 million in revenue in 2019 and $13 million in operating income. We are forecasting $127 million in revenue in 2020, which represents 18% growth off of 2019 numbers, and $18 million in operating income. We think revenue can accelerate to 25% growth in 2021 as the FPGA part of the company’s business really begins to kick in.
On a valuation basis, Silicom trades on an EV / Revenue basis of 1.3x 2019E revenue, 1.1x 2020E revenue, and 0.9x 2021E revenue. It trades at P/E multiples of 12.2x 2019, 8.8x 2020, and 5.8x 2021 when its $11.79 per share in cash is excluded from the calculation. We feel these multiples are extremely depressed given the prospective growth rate of the business. We feel P/E targets of 20x would be more reasonable. Applying these would result in price targets of $55 (78% upside) and $76 (150% upside) for 2020 and 2021.
Why do we think SILC is currently so undervalued? Two reasons. First, there is still a hangover effect from the canceled IBM deal last year. Many investors were in the stock only because of that deal and fled when it ended. Second, Silicom is having a down year in 2019 (revenue down 9% ex-the IBM deal revenue) and the market is still waiting for the SD-WAN and FPGA revenue to emerge. We have seen this movie dozens of times where the market waits for the results to hit the P&L and will not take a stab beforehand, no matter how rosy the picture looks. We make the bulk of our money getting in early to these situations and profiting when the revenue shows up.
In terms of liquidity, Silicom does not trade that well when it is out of favor but once investors find it, its liquidity can rise dramatically (see the period during the IBM design win).
This article was written by
Disclosure: I am/we are long SILC. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.