I finally read through the Kerrisdale short report and began to draft a rebuttal. I got to page 12 before realizing this was not a good use of my time (or yours). In summary, the entirety of the report fits into two categories, disingenuous misdirection or blatant dishonesty.
Below is my review of pages 1 to 12. My responses are in bold:
"Company devotees, die-hard fans, fanatical retail investor base, naïve investors"
The author is very direct in their attempt to diminish the intelligence of LWLG long term shareholders. I personally met dozens of highly intelligent and thoughtful investors two weeks ago at the ASM. I've had the opportunity to meet some industry experts who were top in their field. Our investor base includes this caliber of people.
"Based on our exhaustive discussions with a variety of optical components experts, including former Lightwave scientists and materials engineers currently working on new polymer formulations, we don’t think much has changed in the arena of EO polymers."
There have been three notable former employees in the past five years. One of them moved back to Japan and started his own university polymer lab. Another works for an unrelated tech startup. The third one is working for a university spin-off that is trying to make a case for its technology in order to obtain seed funding to build a prototype. I hope that works out. Competition breeds innovation. None of them have worked for the company for the past 11 months. The LWLG 2022 ASM stated in no uncertain terms that “We have made unparalleled progress in the last 12 months”. The point is that only current LWLG employees are peering behind the curtain to understand the current state of the art in polymer photonics manufacturability. Former employees do not.
"Much of the skeptical investor discourse surrounding Lightwave has focused on the company’s history of paid stock promotion, management turnover, zero revenue, and its consistent track record of not delivering on its promises. All of those are true"
The last paid stock promotion I’m aware of was long before the uplist to NASDAQ. Management turnover is the most inconsistent argument. Prior to Michael Lebby running the company, we had Tom Zelibor. Tom is still chairman of the board. Prior to Tom, we had Jim Marcelli. Jim is still the company COO. Prior to Jim, we had Phil Smith. Phil is no longer with us. We have zero revenue because LWLG is a pre revenue startup. Go figure. Let's begin with Michael Lebby as CEO in order to quantify any missed promises. In the past five years, LWLG has only missed two stated goals. The CEO owned up to them and provided guidance that we will see those goals met and MUCH more in the coming year.
The problem for those fans is that, in terms of speed, Lightwave’s technology lags the most cutting-edge offerings from major optical players like Acacia, Infinera, and Ciena, and in terms of power, modulator power consumption is irrelevant
The author is conflating coherent optics with direct detection, or in other words phase modulation with amplitude modulation. LWLG modulators are capable of both.
The author is ignoring the fact that Acacia, Infinera, and Lumentum lack detail on size and power for their competing technology. Acacia, for example, needed to create an entirely new and non-standard form-factor pluggable module. That will be very difficult to see substantial market penetration with a custom module. The switch and router manufacturers would need to adapt that form factor in order to fit this custom module in their rack mounted switches. Acacia is now owned by Cisco, so we could perhaps see Cisco develop switches and routers to house these custom modules, but datacenters have been very vocal about wanting “fungibility”, a fancy economics term for interoperability. Standardization and vendor agnostic technologies will gain the lion’s share of the market. This is exactly where LWLG is positioning itself.
The modulator on its own is responsible for about 1% of total power consumption, though that might rise to almost 5% if we include the portion of driver power that’s used to drive the modulator. The overwhelming majority of power consumption in optical communication devices comes from the DSP circuitry and the laser, which are responsible for coherent modulation functionality and the light beam, respectively. Even bringing down the modulator’s power consumption to zero wouldn’t make much of a dent.
Again, conflating coherent with direct detect. Without the need for higher order modulation schemes, you do not need a high power DSP. This is the entire point of what LWLG is doing, which is purposefully being missed by the author. Lower power, high bandwidth modulators allow the removal of the driver chip as well as the use of a low power DSP. Comparing the power budget to that of a coherent transceiver is disingenuous or flat out dishonest. The power savings calculation for a low power, high bandwidth modulator should include a substantial reduction in DSP, laser, and driver to start. If one does want to use a coherent transceiver for comparison, those three components equal 60% of the coherent transceiver’s power budget according the example the author used from Infinera.
We should also note that optical experts with whom we consulted were unanimous in their view that because of the material’s reputation of continuous failure, even if someone got it to work – and nobody has yet come close – it would take years to earn broad acceptance. There is no foundry deal – and no revenues – anywhere on the horizon for Lightwave.
I agree with the part of the statement that it would take years to earn broad acceptance. The good news is LWLG committed to this in earnest starting in 2016 and culminating with polymers now being added to all of the global industry roadmaps starting in 2020. So yes, the author is correct that the new technology will be met with headwinds. However, those headwinds began six or seven years ago and concerns have been systematically addressed over a long period of time.
Even compared to 15 years ago (when Lightwave went public), data transmission speeds enabled by the latest generation of transceivers are 20x the 40Gbps that was standard then, and the devices are substantially smaller and consume 1/20th of the power per Gbps (see the bottom row on the table below). In that context, the 2-3x improvement in transmission speeds that EO polymers would enable just isn’t as revolutionary as it once seemed.
The table referenced shows a 90 – 100 Gbaud coherent device consumes .2W/G. This is 18 to 20 Watts per device. I’m not sure the author understands their own analysis.
This is where I stopped my review as it is clear as day the author is counting on the ignorance of the investor base in order for their arguement to have any legs to stand on. I give the LWLG investor base more credit and believe the Kerrisdale short report is either transparent or invisible to those that matter.
