I have owned EZ since 1999 and have done countless hours of research which I have posted on the Gilder Board. I just thought I would share my reasons why I oppose the deal with Mellanox along with most on the Gilder Board.
First, Mellanox is an excellent company and Waldman is an excellent CEO and I have 15% of my stock portfolio in it. They have some new Ethernet switching products out that are better than anything out there and are the leaders in infiniband. They high end computing and datacenter market they target is growing fast.
I also agree that the combined companies of EZ and Mellanox would be much more formidable together than separate and the reasons stated like sales channels, larger is better, product synergies, and others are valid. The merger is a good one if you want a more powerful Israeli tech company.
But I completely oppose the deal at the present time at these prices and have voted no on my proxy. For one thing, stock prices rarely reflect the cash so taking out EZs cash they are only paying 19-20$ a share for the company. With this years earnings probably over 1.30 that is only a 15 PE for a tech company. It does not reflect future products and all the money and promise in them.
The timing is also terrible for EZ shareholders. EZ sells almost 90% to the edge router market and that has gone though a 9-12 month down cycle across all companies. The down cycle is ending now and a new significant up cycle is projected to run for a number of years. EZ also announced the potential loss of the ASR 9000 which sent the stock price way down.
The 100G upgrades are projected to skyrocket starting next year in the metro and data center where EZ will sell products. The 4G buildouts that took so much capital are ending and now telecoms are doing capacity upgrades and network densification to handle the surging mobile video usage and the upcoming internet of things. EZ will do much better in the next couple years in its core market. You do not sell at the end of a long downturn when you are weaker.
The NP5 is also ramping now with a 50% higher price than the NP4 and more design wins and more 100G deployments. The NP5 also brings in the first significant data center wins with switches and 100G router upgrades at the hyperscalers with the ASR 9000. Right now the time is a transition from the NP4 to NP5 which causes some orders to wait for the NP5. The NP5 designs are all locked in after years of development and are not going to be lost. They will drive an increased positive earning and cash flows for the next 3 years at least. Third party reports and others state that the ASR 9000 is not yet lost for EZ and even if it is it will not be noticed until 2019.
EZ also has also only had one product to one market with 3 years plus development times between new generation chips. Now that has changed. The NPS has taped out and already has 3 tier one design wins and multiple tier 2/3 wins before tapeout. It will get many more. The NPS opens up a huge new TAM targeting the data center and SDN/NFV. There is no comparable chip and this was 6 years and well over a hundred million in development. The latest briefs on it show the power of the chip and the huge new markets it hits. The NPS can handle every networking function and with c programmability and 960Gbs output nothing comes close.This will also start ramping along with the NP5 in 2017.
EZ then has the TileGX and the TileMX coming combining EZ tech with Tilera. This chip will enter the arm world and add control plane, ability to execute virtual functions, target nfv servers, and be highly differentiated and superior to the competition. If you read Mellanox's talks about the merger they seem more excited about this than the NPS. It also will start ramping in 2018 so EZ will have 3 products ramping at the same time when it has never had more than one.
The SDN/NFV movement will be the biggest telecom/datacenter networking has ever seen and is inevitable. It will have a huge growth rate near triple digits for years and the EZ chips are perfect for this movement. The TAM is expanded to 2.2 billion with the new chips and that number will rise much higher as SDN/NFV really kick in from 2017-2025. New markets and selling directly to hyperscalers and service providers will reduce dependency on Cisco and provide diversification.
EZ is also fabless, small number of employees, zero tax rate which leads to high margins and profits. Mellanox has 2400 employees, lower margins, higher costs, higher taxers, and more competition. EZ will also only be 20% of the combined revenues so the effect of successful products will be much less than an independent company. EZ also has no merchant network processor competition left while mellanox has intel with omnipath coming online in its main market and many competive switch companies. Mellanox is still superior but omnipath is baked into the intel chips and will be good enough for many applications and could impact Mellanox growth and possibly commoditize the market.
I would prefer the deal to be canceled because I do not want EZ selling at a time of greatest weakness when the future looks much better. I think EZs earnings will go up substantially and the PE multiple will also go up with 3 chips selling instead of one,diversification out of the edge router market and with new products in hot areas like data centers and sdn/nfv.
I would prefer an alliance with Mellanox like marvell where they use EZs NPS and TileMX in their products and work together on custom SOCs combining their tech that mellanox would sell. This would be beneficial for both companies and EZ will stay independent.
I think the price is way too low, the timing is wrong, and cash only is a bad deal for long term investors. |
