Sunday, February 26, 2017

The Impending Super-Cycle In The All-Glass Optical Internet

Technology lagged the Trump rally at first, but now appears to be coming to life. There is one area of tech that may be contributing to that.

There is a Gigabahn in our information superhighway that has been compared to Germany's famous Autobahn, where there is no speed limit except for urbanized or congested stretches, and they typically fly at around 100 mph. Their trips are badly bottlenecked by the last few miles to their home or office. (click on images for full view)

The biggest problem facing the age of the web, other than perhaps security, has always been the immense difference in data speed as light waves vs electricity waves. They built the main trunks of our network with plenty of fiber optic overcapacity, and they have been doing the local switching into our computers by converting the lightwaves into electrical waves and pretending they are to be dealt with as in the party line era. With a lot of cleverness, that has worked OK, but with net traffic doubling every year, we have now come to the limits of that optical denial.

Local area networking with optics is an inevitable revolution of the future. Fortunately, smart minds have been developing light wave switching to do what the very old school electricity switching has always done. If the experts are right, we are about to enter a quantum leap closer to the lighted "cathedrals of glass" in George Gilder's all glass view of the future. There are several public companies involved in "last mile" fiber optics that are the can openers that let the 100 gigabit light speed shining across the globe into our neighborhoods. A lot of this is done with the simple prism effect. There is a super cycle in this can opener function coming as explained in the Investor's Business Daily article "Fiber Optic Super Cycle: Raising The Speed Limit On The Gigabahn's Last Mile."

Clearfield, Inc. (CLFD) Oclaro Inc. (OCLR) EXFO Inc. (EXFO) Ciena Corp. (CIEN) and Acacia Communications (ACIA) are some of the small pure play investments in this anticipated super cycle. The experts think this super cycle is just beginning.

I have traded Clearfield over the years twice for 44% and 130% profits, but I would have done many times better had I just held it after my first buy. The growth, despite all the cycles, has been that massive.

What's all the crazy explosion about? You could say it's a leftover (or maybe "hangover" would be a better term) from the internet bust of over 10 years ago. If you will recall, everything went nuts over the future growth of traffic on the 'net. The telcos frantically laid a fiber-optic backbone to replace the hopelessly slow copper, and mindful of the high cost of continual upgrades, laid in a pretty vast oversupply of capacity. This is commonly referred to as "dark fiber" until end-user demand eventually lights it up.

There is still dark fiber and excess capacity left in the backbones. But the internet traffic growth has been catching up to the last mile distribution systems. This part of the network was sped up with many different half measures using on-demand switching strategies that worked well on the old copper or copper/fiber combos. The basic problem with all the combos is that information on light travels at the speed of light, which is umpteen thousand times the speed of electricity in copper, or silver or any conductor. It has always been cheaper and a little more convenient to speed up the combo switching than tearing up the neighborhood and installing pure optic to the premises.

Is there really a "super-cycle" coming in optical last mile?  Metro traffic jams and upgrades do tend to run in erratic cycles. But they say a super-cycle happens about every 10-15 years, and a big new one is emerging being fueled greatly by the "Web 2.0" phenomenon - Netflix, Facebook, and the whole FANG revolution. They have started buying their own net equipment to squeeze product to their customers. This isn't your teenage daughter's web anymore. Now we have presidential campaigns being waged on Twitter, and our new president has a full blown audience before his first old fashioned press conference. The IBD piece above states:
During "every prior optical cycle, there has been one major driver and a couple of minor drivers," said Needham analyst Alex Henderson, who as argued that the uptrend could be a telecom industry "super cycle." "This cycle looks quite different than that in the sense that there are at least three major drivers and a couple of smaller drivers as well." As a result, the fiber-optics group ranked No. 3 on Thursday among the 197 industries tracked by IBD, up from a No. 95 ranking six months ago
The huge new FANG imposed driver is being complemented by another major driver - a pressing upgrade cycle in China. As they try to emerge as a superpower, their net speed is ranked #82 in the world, and they are seriously attacking that.

Local systems, foreign and domestic are being strained to keep up with traffic on the internet.  According to Wikipedia, the traffic pouring out of the fiber backbones is doubling every 9 months. According to the above IBD article:
The problem is, we've now got these fibers that have 90 channels of 100-gig traffic on them, and they're dumping into a metro core that's 10, 15, 20 years old that can accommodate 10 to 40 gigs of traffic, Henderson said. "That's like trying to drink out of a fire hydrant with a sippy straw.
FTTH stands for Fiber To The Home, which is about the same as FTTP (Fiber To The Premises) and other FTTs that are sometimes collectively called FTTx - or simply last mile fiber. There is now a sharpening need for all-fiber last mile installations.
The disruptive switch-over from copper hybrids to an all fiber last-mile in urban areas means an expensive mess. So ever increasing speeds have dictated a next generation of DSL on the very old copper that's already there. This new standard known as G.fast is being deployed. It is a nice speed improvement, and often compared to broadband fiber based on lab results, but will suffer the usual shortcomings of anything imposed on copper. The fade with distance is very bad relative to fiber.  Also, there is a vast difference in the normal existing copper and what they are using in the lab for these tests. I refer you to Jim Wegat for this. He was an optics engineer for Terabeam, and when I asked him about this supposed equality of G.fast to fiber, he said it:
:... is confusing the wiring in homes and neighborhoods with the high quality wire used in the study by Alcatel-Lucent. It is a bit like saying that a specially designed car broke the land speed record and since most Americans have cars they should be able to break the land speed record with their cars too.
Still, it's just been too much mess to upgrade all the copper we've built into our advanced countries. Thus, if you look at a global fiber penetration chart, it's the emerging markets where they are enjoying higher FTTP rates. The United Arab Emirates already has all fiber to 75% of premises with South Korea at 68% and Hong Kong at 57%. At the other end of the scale, there is France with 3%, and both the US and China with a mere 9%. In the copper infested nations, the switchover is a complicated mess. What we need is simplification!

This is Clearfield's heart and soul. In a story, on this company, Bloomberg starts off with the statement "Clearfield Simplifies Fiber Distribution." Putting all the moving parts together in a last mile light-wave distribution system is an incredibly complicated, disruptive, and costly thing. In the Bloomberg article, they give us the general description of Clearfield's offerings. I am no tech geek, but the thing that stands out to me is the phrase "delivering the industry's only fiber management platform that is built upon a single architecture. Scaling from 12 to 1728 ports". This ease of assembly and a gradual scaling of a customer's build out (so as not to bankrupt the premises) all on one platform (no big interfaces and what have you) is exactly what you see stressed at Clearfield's website. If you go there, the first thing plastered on your screen in big letters is "Scalable Fiber Management". They say:
Based on the patented Clearview™ Cassette, our unique single-architected, modular fiber management platform is designed to lower the cost of broadband deployment and maintenance while enabling our customers to scale their operations as their subscriber revenues increase...No matter which product line you are using….or whether you're networking in the Inside Plant, Outside Plant or In-Building, everything we do is based on the architecture of the Clearview Cassette. While it's designed to fulfill all the "musts" of fiber management, the cool thing is that your technicians learn it once and then the cassette becomes like using legos….simply rearranging the pieces, depending on your fiber network solution!

Clearfield's selling point is that they are the least painful way to get the limitless bandwidth and "future proof" nature of light waves sent to your screen. Clearfield has the kind of insider ownership I like. With CLFD, you actually have more insider ownership (18%) than mutual fund ownership (11%) and almost as much as institutional ownership (23%). It is somewhat undiscovered, except by the insiders. Despite all the cycling fiber has done, this company has produced steady, strong growth over the years:

If these are the results just before a super cycle, there's a chance they may get even better! This, of course applies to, EXFO, CIEN, CSCO, LPTH, and the whole local optical group.