My first reaction to seeing that Infinera filed an S-1 was the emotion of relief. I felt this emotion because Infinera has been the subject of much speculation over the past few years and the subject of many Google searches. I am amazed when I look at the entrance URLs to my blog and see how many times people are searching on the terms “Infinera IPO” or “Infinera S-1” or “get rich from Infinera IPO.”
If you think I am being sarcastic about the last search string, see the image below. I captured this image of a visitor to my blog on Feb 8. The person was located at what appears to be the Starbucks Coffee Company headquarters and had Googled the terms “get rich from Infinera IPO.”
I spent part of yesterday morning reading emails and blogs on the Infinera S-1 and instead of just repeating what has already been written, or repeating what is in the S-1, I would rather provide some observations about the investment thesis from a venture capital perspective.
I posted on Level 3 Communications Tuesday and then went to meet a venture capitalist at Battery Ventures in the Boston area. I came away with two observations from the visit with Battery. The first was that VCs are looking for networking and system plays that do not require a $300M funding plan. The second was is it possible to build a public company in the networking space on $40-60M in venture capital?
When I left the meeting I found that I had a large amount of emails from irate Level(3) investors and traders and even a retiree of Level(3) who were not happy with my post on the company. If you take the time to pursue the L3 message boards at Investor Village, you can read in detail how much these people despise me. Between the L3 post and answering emails from people who accused me of working for Akamai, I did not have time to read the Infinera S-1 until this morning. It should be noted that one person accused me of being a consultant to Akamai (NASDAQ:AKAM) and launching a guerrilla marketing campaign to discredit the L3 CDN network because Akamai sees it as a looming threat to their business. For the record, I am not an Akamai investor and nor have I worked for Akamai. I know people from Akamai read my blog and should they deisre to hire me as a consultant or full time employee I would enjoy conservation on the subject – but I am not on the Akamai payroll. As I am typing this blog, Kevin O’Hara is presenting at the Merrill Lynch Communications Forum and I am getting emails from the L3 boosters promoting his presentation.
After reading the Infinera S-1, answering emails on my L3 post, and thinking about the Battery meeting, the confluence of several data points and anecdotal data points emerged. Here they are in no particular order:
Infinera has closed $315M in venture capital funding. Force10 has reportedly closed $400M in venture capital funding. Raising venture capital at these levels must involve marathon investor meetings. They are really fantastic sums of money and I think they are indication of how hard it is build a new technology company in the optical and Ethernet switching markets. These markets are difficult because they are mature with suppliers who have control of market share and in the case of the service provider market the number of actual buying entities is declining in the US market. Is the VC model broken if companies require a $300M funding plan and six years to go IPO? There are number of smaller public companies in the networking space. Is it a better investment thesis to consider buying one or two or three of these companies (PE-buyout model) to secure the sales channel and then find or fund innovation to be part of the company? In the past, VCs would start a company from the product/innovation side. Is the future VC model to start with the sales/market share side first and then fund innovation, which is a reverse model? On page 98 of the Infinera S-1, it states “On April 11, 2005, we entered into a long-term supply agreement with BTE Equipment, LLC, an affiliate of Level 3, for the supply of our DTN System and support services. On May 19, 2005, August 20, 2005 and November 15, 2006, we amended the terms of our supply agreement with Level 3. Certain of these amendments were entered into when Level 3 owned more than 5% of our outstanding shares. Pursuant to the agreement, we recognized $1.0 million of revenue in 2005 and $35.2 million of revenue in 2006.” Of customers listed in S-1 on page 59, Level (3) is the largest of these customers. In order to win the L3 business back in 2005, L3 invested in Infinera as part of the supplier agreement. This is not surprising, but it does provide an indication of how difficult it is for small companies and startups to win business with large service providers. The net effect of this agreement is that Infinera and L3 are dependent upon each other. Infinera states that L3 is a 60% customer and at one time L3 was a 5% owner. To get a return on their investment, L3 needs Infinera to go public and meet quarterly numbers. That means L3 must continue to order product from Infinera. The downside risk for L3 is if Infinera does not raise capital via the public markets, then L3 could be left with a network full of equipment from a defunct supplier. I think VCs will become averse to funding business plans in which success requires the establishment of a symbiotic relationship between the customer and the supplier through private investment. The deferred revenue of $110m on page 36 of the S-1 is another indication of how difficult it is to sell products into the service provider market in a SOX world. Back in 90’s, it was standard practice to sell on futures. The best PowerPoint slides showing the future often won the day. In the SOX world, selling on futures affects revenue. If product enhancements are part of supplier contract, the revenue recognition will be deferred until the contract obligations are met. There are all sorts of conditions that affect revenue recognition. Public companies are hypersensitive to these rules and I think startups are still learning the new rules of the game. Over a two year period, Infinera has generated $173M in sales on a cost of sales of $166M. I think this is a reflection of the competitiveness of the optical market from a cost perspective. I think the sales plan has been to sell slightly above cost to acquire footprint and then as capacity requirements emerge, line cards sold into the customer base will be at a higher margin. Side note to all the L3 people who have been sending me links to media articles expounding the “explosion of video” that is overwhelming the internet. I just heard from Bernie Ebbers via email, he said the “internet is doubling every 100 days.” One last S-1 observation is that Infinera is burning through cash at a rate of $5.5M per month. They need to raise capital via the equity markets because I think it is increasingly difficult for the VCs and later stage investors to commit another $100M to the company. The plan has to be to secure the proceeds from the IPO, fix the product margins and then close a secondary offering similar to Riverbed. If they can raise $14OM from the IPO and then do a secondary by the end of the year, it is plausible that they could have $200M in the bank by this time next year.
As always, thoughts and comments welcome, whether private or public.
Full Disclosure: Of the companies listed in this post, I only own shares of Google.