Friday File: “Cognitive Sight” examined, plus some updates

by Travis Johnson, Stock Gumshoe | May 13, 2016 4:31 pm

Looking at a Michael Robinson teaser... plus, what's up with Input Capital and Crius Energy?

Happy Friday! I have a couple updates and notes for you today, but I thought I’d start with a look at a new teaser many of you have been asking about from Michael Robinson[1], it’s headlined “Cognitive Sight” and it’s really about artificial intelligence and it’s ability to make the blind see again (among many other things)… and, it turns out, it’s an extremely timely teaser (more on that in a moment).

Robinson’s ad features him wearing a pair of normal eyeglasses with some extra little thingy at the temple, and talks about how this “Cognitive Sight” ddevice can folks with nearsightedness, macular degeneration or other vision problems to identify Mount Rushmore when all they can “see” is a blur. All because of that little thing on his glasses, so it must be some cool new gadget that turns you into a half-man, half-robot, right?

Well, not exactly. That’s some of the future possibility he’s hinting at — but, as we’ve learned to expect from Robinson, he’s selling us on a fantastic dream of huge advancements that are quite possible but then actually recommending a stock that’s just a small part of the background of making that technology possible.

So it was when he pitched the “end of disease” thanks to the small sensors that can be used to more closely monitor our health or alert us to potential problems, and was actually teasing not some innovative new sensor maker with a new way of doing things or a brilliant medical discovery, but big ol’ STMicrolectronics (STM), the diversified semiconductor giant that’s also the biggest player in the MEMS chip sector.

And when he was promising that a new “neural imprinting” technology would rid the world of pain, he was really pitching virtual reality as a broad concept… and NVIDIA (NVDA) as his stock pick because their graphics chips will help make high-end virtual reality systems work better and faster.

Often we have to look past the headlines of his ads and get into the nitty gritty before we can make sense of what he’s actually talking about… so what is it this time?

Here’s a little sample from the intro:

“I’m wearing what The Financial Times reports could be a $20 trillion technology…

“Cognitive Sight….

“Right now, the biggest players in Silicon Valley are betting it could be the breakthrough 324 million people who suffer from macular degeneration, cataracts, glaucoma, even blindness have been searching for.

“And remarkably, stakes in the 7,300 patents driving this multi-trillion-dollar invention have been put up for grabs!”

Wow! Thousands of patents backing a cure for vision problems? That sounds awesome! We’ll be rich!

Oh, but wait, we need to look at the details… right?

What he’s really talking about is the combination of visual sensing and what data scientists seem to now be calling “deep learning” — which can mean something as simple as a little glasses-mounted camera being able to understand the images in front of you and describe them to the wearer using words, or more dramatic enhancements to the visual field made possible with “augmented reality” glasses like Google Glass or Microsoft’s Hololens.

And then we get into a bit more about why he’s picking one particular “secret” company…

“Now here’s the big secret. And it’s a really big one.

“It doesn’t matter whether it’s Microsoft or IBM or any of the veritable “who’s who” of tech behind this massive $8.5 billion funding wave…

“The lion’s share of this cash is all set to land in one place…

“One tiny company controls an astronomical amount of patents that are driving this Cognitive Sight revolution.

“7,300 patents to be exact.

“A truly rare situation has presented itself.

Are you getting our free Daily Update
"reveal" emails? If not,
just click here...


“You have a chance right now to actually secure ‘patent stakes’ in this tiny firm’s technology.

“And be at the center of it all, possibly becoming a millionaire in the process.”

This is, of course, what’s wrong with these kinds of ads — the “possibly becoming a millionaire” bit. Many of us will be millionaires by the time we retire, one hopes, because it’s hard to generate enough income without a large nest egg, but the promise of “becoming a millionaire” based on one stock idea is what leads individual investors to chase rewards and take large and risky “bets” on a specific outcome for a specific company.

And don’t get too hung up on the idea of “securing a patent stake” on this technology — that’s just his way of being more exciting and “insider-y” than if he were to say something more pedestrian like “buy some stock in the company.” The owners of the corporation, which would be stockholders like you or I, are owners of everything the company owns, including their patents, so it’s not necessarily untrue that you’re getting an indirect “patent stake”… but it’s misleading.

Most of the quotes that Robinson pulled in the ad, the pieces that are supposed to jump out at you because they’re from recognized names like consulting firms or business publications, are about either artificial intelligence or “deep learning”, a more specific concept of AI, combined with gadgetry that seems to be based on the augmented reality stuff (like Google Glass), so perhaps he’s just broadly piecing that together into a single revolution that benefits his particular company.

More hype:

“…. with veritable pocket change you can get a small piece of every device these companies roll out using this firm’s patented technology.
Because of the deals they are negotiating, their market share on some of their boldest applications could hit 82%.

“They’ve established contracts with:

“Even three of the biggest early backers of Cognitive Vision:

“We’re at the beginning of what could be a wild and exciting 6,642% burst in revenue.”

So that’s all sticking in your subconscious brain — the only references to what you might earn are huge percentages or “millionaire” status, the only references to what you might have to invest are “pocket change.” It’s subtle, but it subconsciously drives many of us to huge expectations.

He then goes on to describe the “cognitive sight” device he’s wearing and using as an example, which essentially is an extended depth of field camera that somehow transmits data to his brain, and he says then that the “Artificial Intelligence Processor” is the key enabler….

“None of this would be possible without the most important component of this technology.

“Look at it as an “Artificial Intelligence Processor”…

“Which I’ll just abbreviate as A.I.P….

“It’s what makes it possible for this device to use artificial intelligence to interpret all of the images being captured by the lens.

“It’s what makes it efficient enough to run all day, processing any kind of writing.

“It’s what allows it to tell time, see colors, and identify faces of friends and family…

“Not only recognizing them in the moment, but storing them in its advanced memory to access later, just like the human brain can.”

And he compares this “AIP” to a more conventional CPU, like the central processor in your computer:

“Your typical CPU consists of 8 to 10 cores…

“However, the A.I.P. can hold up to 15 billion microscopic transistors….

“This A.I.P. is at least 3,200X more powerful than the CPU you most likely have in your computer right now.”

So, in case you haven’t jumped ahead yet, we’re starting to understand that he’s taking this whole artificial intelligence, deep learning, augmented vision world and pin it all on the processor that makes it possible. And it’s the owner of those “A.I.P.” patents that he’s excited about…

“… the A.I.P. is what gives Cognitive Sight devices the power necessary to augment or even replace damaged human sight….”

The specific “tiny device” that Robinson is wearing in the picture, and which he references over and over again, is almost certainly the MyEye from Orcam[2], an Israeli company founded and controlled by the same guys who built MobilEye (MBLY) — Orcam is still private, though they’ve got $15 billion in funding from Intel a couple years ago and they may go public at some point fairly soon. They also have a clip-on “Portable A.I.” device that’s camera based called MyMe[3], but unlike the MyEye that’s not a commercial product yet. It’s pretty cool-sounding stuff, but you can’t invest in it directly at the moment and the specific little device is more of an example of some of what this artificial intelligence can do in different settings than it is a universal “cognitive sight” device. You can see a video of the MyEye in use by a journalist here[4], it’s been available to a limited extent for a year or two and they’re trying to roll out to more markets.

The connection with MobilEye makes sense, since their technology for processing visual data is what makes the MobilEye devices so powerful, able to process things like pedestrians, speed limit signs, lines on the road, and the speed of approaching cars or obstacles — it seems logical that the same kind of visual processing would be useful for a blind person, too, reading them words from a page, recognizing the face of someone to identify them, or otherwise describing the physical environment.

But, of course, it’s early days for that product — and it’s not Orcam we’re being teased about, it’s the company with those “A.I.P.” chips.

Robinson goes on to describe tons of ailments for which this kind of technology would be useful, whether it’s helping Alzheimer’s patients remember faces or alerting Parkins’s patients to when they’re about to have a “freeze,” or help autistic people recognize facial expressions and emotions, and then implies (as these ads always do) that this means the potential is like the several-thousand percent returns generated by pharmaceutical companies that developed treatments for these diseases. Which is a stretch, of course, markets aren’t static or one-sided — there’s no overseer who says, “this year we’re spending $10 billion on Alzheimer’s Disease … whoever comes up with a treatment gets 11% of that and is guaranteed a 1,200% stock price jump!”

Robinson ties it back in specifically with Google Glass as well, since that’s the other major example that’s gotten a lot of press coverage…

“By now, most people have heard of Google Glass.

“An early version of it debuted in April of 2013, just to give people a taste of its potential.

“IBM, Microsoft, Facebook, Tesla…everybody is racing to tap into Cognitive Sight.

“$8.5 billion in venture capital money poured into this technology… just last year.

“Because the coming windfalls are going to be nearly incomprehensible.

“Everybody wants to position themselves to tap into this $20 trillion in wealth creation.

“And this is where our enormous opportunity lies.

“Because each device these companies roll out requires the same critical component.

“That Artificial Intelligence Processor – the A.I.P.

“Popular Science has described the A.I.P. as ‘the workhorse of modern A.I.'”

This is somewhat conceptual, I suspect — not so much a literal claim that “each of these devices must use the same chip” in their current iterations, because Google Glass and Orcam have no components or chips in common that I’m aware of. Google Glass got its computing power from the Android phone with which it was paired, and Orcam’s MyEye uses the technology developed by MobilEye’s founders but also uses chips from Freescale Semiconductor[5] (now part of NXP (NXPI)).

So does that mean he’s pitching Freescale? Probably not — they’re not primarily focused on this product, not even close, and they don’t have a “deep learning” or A.I. chip with 15 billion transistors. I don’t know how many patents NXPI has, but they say they’re “supported by 9,000 patent families” so it’s a heckuva lot — Freescale has sold thousands of patents over the last ten years or so, and was part of Motorola before that so I imagine that, like most big tech companies, they’ve got bucketloads of patents that won’t ever be worth anything.

And then, finally, we get down to one more batch of clues about this “A.I.P.” company…

“The ‘A.I.P. Prospectus’ is going to take you deep inside the company with the patented technology driving this Cognitive Sight revolution.

“It has attracted some of Silicon Valley’s top C-Level Executives.

“The founder and CEO was a microprocessor designer at AMD, the $1.7 billion chipmaker.

“And his co-founder was an engineer for both Hewlett Packard and Sun Microsystems.

“The ‘A.I.P. Prospectus’ will explore the distribution deals they’ve been reaching at breakneck speed.

“Consider this – in 2013, this firm supplied A.I.P.s to just 100 companies employing Cognitive Sight technology.

“In 2014, that number jumped to 1,549.

“And last year, it spiked again to 3,409.”

So who is it? Well, believe it or not, this is again a tease about our old friend NVIDIA (NVDA).

Yes, really.

NVIDIA is primarily a maker of graphics processing cards/units (GPUs), and their biggest market (by far) is gaming — they make the high-power graphics processing chips that enable faster, clearer images in video games, and for the high-end gaming computers those GPUs alone can easily cost 500-700 dollars and, at the very top level, several thousand dollars. That’s still where most of NVIDIA’s revenue and profits come from, but it’s also a fairly low-growth business and sometimes, when gaming trends are plateauing for whatever reason, it doesn’t grow at all. Most of the rapid growth for NVIDIA comes from their newer lines of business, which are essentially using the high power of their graphics processing chip designs to create much faster processors for “deep learning” or for processing multiple inputs for applications like self-driving cars, or even speeding up processing in data centers[6].

That “A.I.P.” referred to with the 15 billion transistors is NVIDIA’s Tesla P100[7], which is designed using their Pascal architecture to dramatically speed up “deep learning” and processing for data centers and also powers the “supercomputer in a box”[8] that NVIDIA is rolling out this year. Beyond those “deep learning” and A.I. applications, NVIDIA is also seeing growth from its Tegra processors, which are used in mobile devices and in automobiles for both cockpit electronics and autonomous driving “thinking” (including in Teslas).

And, in case you’re double-checking the clues, those specific numbers about the companies using these “A.I.P.” chips for “cognitive sight technology” match up exactly with the number of companies “engaged with NVIDIA on deep learning.”[9]

Robinson also gets very specific about near-term growth projections, saying that the “market for cognitive sight applications” is projected to hit $109 million this year, and he estimates the “total economic impact” could reach $175 billion next year. There are so many estimates of the impact of virtual reality and augmented reality, which apparently are related to his made up term of “cognitive sight,” that you can really take pretty much any number you want and justify it somehow — Goldman Sachs published a good report on virtual and augmented reality[10] earlier this year and put an estimate of $80 billion on that market in ten years, but they’re just guessing too… and their estimates drop down to $23 billion if uptake is delayed, the products disappoint in terms of power and speed, and they remain just the province of videogames and really disrupt only the video game console market.

And — this is why I said above that it’s a timely tease today — NVIDIA just reported earnings last night, did better than analysts projected and issued guidance that investors were happy with, and saw its stock bump up by about 13% to a new all-time high just above $40 a share. As I’ve noted in the past, I continue to have a long-term LEAP[11] call option position in NVDA, which has now done nicely, but what’s the situation now — is the stock worth a nibble, either because of your reality-based analysis of their business or because of Robinson’s fantastical “Cognitive Sight” projections? (And in case you’re wondering, no, there’s no indication I’ve seen that NVIDIA’s chips are actually in the Orcam device, and they’re certainly not in Google Glass).

Well, I can’t tell you whether Robinson will be right about NVIDIA being the linchpin of a $20 trillion business and swimming in patent royalties — but it seems a bit of a stretch. All growing fields attract competition, and NVDA already has plenty of competitors — it’s far too easy to read exciting analysis about the prospects for a huge and growing field that’s really just starting to explode, look at one well-known company that is doing a lot of work that might end up being a strong supplier to that market, and assume that somehow they’ll have that market to themselves or will maintain their nascent leadership in that field.

And he does, of course, do quite a bit of exaggerating — including implying that the “early A.I.P.” sold for $2,800 apiece a couple years ago but is now available in a “basic model” for $27.50. It’s true that the low end GPU cards NVIDIA sells bottom out around $27.50 (or a bit less for the chips that are built into computers and sold wholesale), but those are, of course, not the “deep learning” 15-billion-transistor chips that he spent most of the ad talking about.

That’s not the only projection, either:

“2,591,753,000 computers, tablets and smartphones will be sold this year. Each will contain at least a base model A.I.P.

“At a minimum, that’s $71.3 billion dollars.

“And the tiny company at the center of it all has nearly cornered the market.

“Without anyone noticing, they have positioned themselves to become more important to the future of computing than Intel, the largest computer chip company in the world.”

That’s drawing lines between several points that are all quite a ways away from each other, and though all computers and smartphones have at least some sort of graphics processing capability (often built into their main processing chipset, and often not an NVIDIA chip), that doesn’t mean that each of those chips costs $27.50 or more. Intel and Qualcomm and the other biggies build GPU capabilities into their core chips, and have for years, it’s not something that has to be “added on” for basic applications.

So I’d say it’s a little crazy to look at the patents and current business NVIDIA has, project a huge surge in demand for their processors (including both the GPUs primarily used for PC gaming as well as the advanced Tegra or Pascal/Tesla chipsets for deep learning, data centers or self-driving cars), and somehow conclude that it’s reasonable to expect 6,642% growth in three years, which is Robinson’s assertion (that’s turning $15,000 into over a million dollars in three years).

And if the market is really $71.3 billion today, with chips in every product as Robinson asserts, then NVIDIA, with annual revenue of about $5 billion, has only about 7% of the market — so “cornering” is a bit of a strong word.

As I’ve noted, I’ve been a semi-skeptical believer in NVIDIA’s potential (that’s why I dabbled in a small options[12] position last year instead of risking more capital) — mostly because of the fast growth of their data center and automotive segments and the stability of the gaming segment to this point… so we can look at the reality of NVIDIA’s current situation and the expectations about their near-term prospects, particularly since we have some nice, fresh data from their earnings release last night.

The primary concern most folks seemed to have about NVIDIA before earnings was that they were still reliant on their core gaming GeForce business to a huge degree, and that business might not grow… and might be facing more competition from AMD, the only other major GPU chipmaker that has been capable of being a genuine competitor for NVDA — that’s pretty well summed up in this analyst note published by Barron’s[13].

But NVIDIA’s results continued to beat expectations, even though the “beat” was not terribly dramatic, and they raised expectations slightly for the current quarter, which investors always love.

Growth on the top line was decent, with revenue up 13% over last year, and improving margins and some stock buybacks meant that earnings per share growth was quite a bit more substantial — earnings per share grew by 38%. And that’s with actual GAAP accounting, not the sexier non-GAAP stuff that analysts usually accept (non-GAAP for tech companies usually means that they’re ignoring stock-based compensation and pretending it’s not an expense to shareholders). They’re not exhibiting crazy, breakneck growth, the business is still seasonal to some degree and the earnings were down a little bit from last quarter, but year over year growth for a mature company is what I usually look at.

Here’s what CEO Jen-Hsun Huang said about their “deep learning” and artificial intelligence prospects:

“We are enjoying growth in all of our platforms — gaming, professional visualization, datacenter and auto. Accelerating our growth is deep learning, a new computing model that uses the GPU’s massive computing power to learn artificial intelligence algorithms. Its adoption is sweeping one industry after another, driving demand for our GPUs.

“Our new Pascal GPU architecture will give a giant boost to deep learning, gaming and VR. We are excited to bring a new wave of innovations to the markets we serve. Pascal processors are in full production and will be available later this month.”

So that’s the growth excitement — new markets enabled by “deep learning” capability on top of their more advanced GPU architectures, which are designed to for much faster and intensive processing than standard CPUs, even the multi-core CPUs that power most higher-end computers today.

And, perhaps as importantly, they’re expecting the next quarter to have continued revenue growth, continued margin expansion as costs are reduced slightly, and they’re planning to return a billion dollars to shareholders through a combination of continued dividends[14] (the dividend yield is only about 1%) and a large buyback program… which they can easily afford, since they have about $5 billion in cash and short-term investments (market cap is about $22 billion, so it’s a fairly large chipmaker).

But what’s keeping my expectations a little bit in check, and making me a little bit cautious, is the concern that the gaming GeForce business might not grow enough to keep the overall earnings growth near this breakneck pace… and the other, faster-growing businesses, are much smaller than GeForce so it will take a lot for them to make up for any softness or “maturity” in the gaming business. That was not a problem this quarter, with the gaming business continuing to grow (down from Q4, as you’d expect for a consumer-focused business with a holiday push, but up 17% over the year-ago quarter as they keep rolling out new high-end GPUs)… here’s the segment breakdown for the quarter from the CFO’s commentary[15]:

nvdaq1

That gives me some confidence that they’ll be able to keep growing the gaming business, which will remain an important foundation for them for the foreseeable future — even though AMD may be offering stiffer competition now than they have in the past. And for those counting on patent royalties to roll in rapidly, you can see that’s not happening yet — they combine IP revenue with the OEM business, which is their legacy, lower-margin business of selling built-in GPUs mostly for desktop and laptop computers, and that segment is the one area of the business that’s materially shrinking. I don’t know if patents are currently growing within that bucket or not, but they’re not the driving force for NVIDIA’s financials right now and seem unlikely, given the relative lack of commentary about patent royalties from the company, to be a huge factor in the foreseeable future.

And though I’d probably argue in favor of dividends or reinvestment in the business instead of buybacks at a time like this, when the shares are priced at a relative premium to the market like they are now, it’s true that their continuing aggressive buyback will probably help bolster their earnings per share growth — and that is, after all, what most investors care about each quarter (and, frankly, they don’t have much reinvestment potential — they don’t need much capital spending, so the only reasonable way to use a lot of that cash for the business would be to acquire someone). And they’ve been consistent, they’ve bought back about $2.5 billion worth of shares over the past five years, reducing the share count by about 10% since 2012.

Analysts are still a bit disjointed about NVDA, probably partly because they want to be cautious about the gaming business and the higher-growth segments are harder to predict (self-driving cars and A.I. and virtual reality aren’t really established enough to have much hope for modeling them accurately). Depending on which published average of analysts you use, the long-term (five years) growth is expected to be either 9% or 20%… and the expectations for this year are for earnings growth of about 40% (their year ends on January 31), but next year is expected to be flat. So really, it’s very hard to guess at what the future will hold… but NVDA does have strong and growing businesses in some “hot” areas that are at least potentially going to keep growing rapidly for a long time, in automotive and data centers and “deep learning” high end chipsets, and there is some real potential that virtual reality could give high-end gaming another kick in the pants and spur some real demand growth for NVIDIA’s core GPU products to fuel those resource-intensive immersive virtual reality gaming experiences.

So I’m still slightly skeptical, but I’m getting a little more comfortable with NVDA — and it’s too bad I had to wait for another solid earnings report (and a 15% higher share price) before I felt that comfort level improving. My call options are doing just fine, but it’s still a small position and still slightly out of the money.

Given the continuing rapid growth in their smaller segments and the continuing growth in gaming GPUs, I’d now be willing to put on a real equity position instead of just speculating that the virtual reality, autonomous cars and other “story” pressure would drive the shares up in the near term. I haven’t done so yet, and won’t for at least a few days because of our trading rules, but once the stock calms down from this post-earnings frenzy I’d think about opening a small equity position and hoping for some dips in the coming quarters to add to that position (assuming, of course, that the core business potential remains).

The shares are not cheap — for earnings growth expectations in the 10-20% neighborhood I’d be reluctant to commit much capital at a PE much above 20, and NVDA is now trading at about 27 times current-year and next year earnings… but I’d be willing to nibble here, partly because I think the analysts are probably being too conservative about next year IF autonomous cars, virtual reality, and the data center/deep learning businesses continue to show signs of breakout growth.

But no, NVIDIA is almost certainly not going to make you 6,000% returns because of augmented reality or artificial intelligence investments over the next few years… that would make NVDA a $1.3 trillion company in 2019, and that’s almost inconceivable (it’s a $22 billion company today, with grow expectations of 10-20% a year according to the boring old Wall Street insiders… that would make them about as big as Apple, Facebook and Google combined today).

So there you have it — lots of excitement about “Cognitive Sight” conjured up by Michael Robinson, but it’s all really about the same stock he was touting for its “Neural Imprinting”[16] potential starting six months ago. The overpromising is crazy, but the company is real and the potential is impressive (as long as you can get those crazy 6,000% numbers out of your head).

In other news…

Crius Energy Trust (KWH-UN.TO, CRIUF), the energy marketer which I own and have written about a few times since Keith Schaefer[17] started teasing the stock as a favorite last year, also reported earnings yesterday (press release here[18]).

My general sense is that the company grew a little bit, with a consistent (and small) level of organic customer growth (as they often do, they also increased their customer base through acquisitions — but they rely on pretty strong selling just to maintain their customer count, since many customers have annual renewals, so slow organic growth is pretty good). They didn’t get much in the way of economies of scale from the growth, from what I can tell, so the net impact has been quite slow growth in cash flow and distributable cash flow, consistent with the 2% boost in the distribution/dividend that they announced earlier in the year.

The outlying growth prospect has been that they could substantially bump up the value of each customer by cross-selling, and that it could be seriously impactful for upside “surprise” earnings if they manage to sell a lot of solar[19] systems in any given quarter (since they get big up-front commissions for solar sales, whereas their standard electricity sales contracts pay out over time), but solar system revenue was actually down slightly year over year so that particular catalyst hasn’t hit. It still might at some point, I imagine, particularly if they continue to get access to more of Comcast’s customer lists if that cross-marketing partnership continues to be acceptable to both companies, but no real growth yet.

So that leaves me with a feeling of “meh” about Crius — there’s nothing identifiably wrong or disappointing, they’re good marketers and the company seems to be well-run, they are currently growing at a slow pace and they pay a solid dividend that’s balanced by the risk that they have no real “assets” and have to re-recruit or replace most of their customers on a rolling basis, but the potential spike in growth from either solar or the Comcast partnership (or both) has not materialized in a significant way. And “meh,” particularly if they’re still paying a solid dividend and show no signs of having trouble covering that dividend or facing any balance sheet problems, means “might as well hold and continue to collect that dividend” — the business is getting slightly better, not slightly worse, and an 8% yield is nothing to sneeze at.

And while we’re in Canada[20]…

It’s always hard to know what will be the catalyst for a beaten-down stock to recover — in the world of somewhat unusual “financier” investments where I like to dabble from time to time, we’re starting to learn that writedowns and troubled customers often have an outsize impact on the share price. I wrote about Grenville Strategic Royalty[21] last week, and they continue to be troubled (the stock is down more this week on no news, perhaps just investors digesting the bad news)… and this week we had news of a company digging out of a defaulted contract with a troubled “borrower.”

Input Capital (INP.V, INPCF) is the canola “streaming” company I own and have written to you about a few times, and they crashed last Summer because of an announcement that three of their streaming contracts were essentially defaulting and being canceled. That made everyone fear for the safety of the business model, and for the drop in near-term cash flow those cancellations would bring, and the stock fell straight down in a wild overreaction, going from C$2.80 to C$1.60. It had been overvalued before, and that brought it down to being fairly valued in my book, so I bought some shares. Time passed and not much happened… and they announced perfectly solid quarterly results, and told shareholders that the contract cancellations would be “made whole” because of their solid contracts. Investors remained suspicious and worried after that drop in price and the “test” of their business model, and the stock remained in the same low trading range from C$1.50-1.75.

Then they announced this week that they have resolved the largest of the three contracts that were canceled by taking possession of 5,000 acres and a bunch of equipment that’s being sold at auction, and that, again, they expect to be made whole. This is essentially the same thing they said back when the stock fell, but back then they hadn’t actually gotten the cash yet — now they’ve got the land and some of the cash for one of the three contracts (the biggest one), so perhaps that was the little bit of reassurance that investors needed, and the stock shot up by 20%. It’s now back at about the maximum price I’d ever consider paying for Input given their current financials, just under $2, but in all fundamental respects the company has not changed, nor has it done anything different than what it clearly stated it would do over the last six months. Investors are fickle.

And, of course, investors also often fail to pay attention to the things that matter most — it is important to see proof that Input Capital has good terms in their contracts to ensure compliance, and I suppose we have more of that proof now (though there could certainly be other similar problems in the future, particularly in a bad growing season), but what no one seems to be talking about is the largest driver that’s outside of Input’s control, and the thing that will have the biggest impact on whether or not Input becomes more profitable: Canola prices.

Input’s streaming deals are structured to give them the canola at a set price that’s far enough below the market price at the time the contract is signed to give them a nice profit if all goes well (and a great profit if yield improves), so if you combine the up-front capital and the per-tonne payments they’ve been paying, on average, about C$309/tonne for canola from their partner farmers, and last year they solid it at market prices for about $483/tonne. That margin is where their cash flow and margins come from, so the potential for upside comes from higher realized prices that have the potential to bring Input to a net profit earlier in the relationship (normally their investments would really generate profit above the original capital investment only in the last couple years of the five or six-year contract, since most of the cash is paid out at the beginning of the contract).

Near-term canola futures have bumped up a bit, over C$500/tonne, so there’s possibility for some improvement — prices have ranged from the low-$300s to over $700 over the past decade, so there’s certainly also the possibility of a drop… a good reminder that we are, indeed, talking about something whose profitability depends on commodity prices. They continue to diversify as well, with 94 “streams” now active, and, importantly, they have done all of this with equity — they have no debt, so they are able to quickly compound any positive cash flow into new streaming contracts to grow. The real risk is a terrible growing season in the Canadian prairie — and that could happen at any time (farmers in Alberta[22] are suffering from dry conditions right now, for example, and many of them are no doubt impacted by the wildfires that were encouraged by those same dry conditions — I don’t know if any of Input’s 14 Alberta farms are specifically impacted).

So no news really at all, no surprise, and the stock climbs 20%…. and it took most of a day for investors to get excited about that settlement and drive the shares up after the press release, so maybe there was some other external impetus for the recovery (like a newsletter recommendation, for example, or a mention on Canadian investment TV).

I’ll take it, but the company is reporting their next quarter on May 31, so I’m not in a hurry to add to my position at these higher prices — we’ll see what they say about the growing conditions and whether they make any forecasts when they report in a couple weeks. If the model keeps working, it’s still pretty easy to buy in the C$1.60-1.90 range, that’s paying up to a small premium over the actual capital they’ve invested (but not a crazy premium like when the shares were at C$2.80), but, like I said, it’s at the top end of that range and is not currently cheap, so I’d prefer to wait and see what the actual financials look like in a couple weeks — the “we’ve done part of what we said we would do” news doesn’t merit a 20% jump by itself, in my book, but that’s a reasonable recover from the fact that the stock fell too far based on the first announcement. It’s always hard to guess at what will tip the balance for individual investors and give them confidence.

I’d say, in fact, to wrap us back to that Grenville Strategic Royalty comment I made above, is that what has made Input’s business seem a bit more sustainable than Grenville’s, even though Input is far less diversified and is based on a single commodity produced by farmers in a single geographical area, is that Grenville made the mistake, I think, of paying an aggressively high dividend almost from the beginning, and then of raising that dividend again during the year last year to keep investors excited. I do still hold a bit over half of my Grenville position after stopping out of part of the holding last week, and I think we’ll likely conclude, in retrospect, that the post-earnings selling has been overdone… but they were skating on thinner ice than I expected, and a small financier company that makes investors nervous, particularly with a default that makes us question the business plan or the credit screening being done, can take a long time to rebuild confidence (as we’ve seen with Input Capital, which took almost a year to get to this week’s partial recovery). I think Grenville’s big dividend made expectations too high, and that the dividend hike was, in retrospect, a mistake that will take them some time to correct. A solid balance sheet would probably have done more than a barely-sustainable dividend can to keep the stock relatively stable, and able to grow more quickly and absorb currency fluctuations and the ups and downs that their royalty investees face.

And that’s all I’ve got for you on this lucky Friday the 13th. Enjoy your weekend!

Disclosure: I personally own shares or call options on Grenville Strategic Royalty, Input Capital, Crius Energy Trust, Apple, Facebook, Alphabet/Google and NVIDIA. I don’t have direct exposure to any other stocks mentioned above, and won’t trade in any stock covered for at least three days per Stock Gumshoe’s trading rules.

Endnotes:
  1. Michael Robinson: https://www.stockgumshoe.com/tag/michael-robinson/
  2. MyEye from Orcam: http://www.orcam.com/
  3. MyMe: http://lp.orcam.com/myme/
  4. video of the MyEye in use by a journalist here: http://www.bloomberg.com/news/videos/b/5574166f-f2f2-48c3-8741-30d267387113
  5. chips from Freescale Semiconductor: http://cache.nxp.com/files/industrial/doc/brochure/ORCAMIMXCS.pdf
  6. data centers: https://www.stockgumshoe.com/tag/data-centers/
  7. NVIDIA’s Tesla P100: http://www.nvidia.com/object/tesla-p100.html
  8. “supercomputer in a box”: http://www.nvidia.com/object/deep-learning-system.html
  9. number of companies “engaged with NVIDIA on deep learning.”: http://images.nvidia.com/content/technologies/deep-learning/pdf/NVI-02-DeepLearning-Infographic-v10.pdf
  10. Goldman Sachs published a good report on virtual and augmented reality: http://www.goldmansachs.com/our-thinking/pages/technology-driving-innovation-folder/virtual-and-augmented-reality/report.pdf
  11. LEAP: https://www.stockgumshoe.com/tag/leap/
  12. options: https://www.stockgumshoe.com/tag/options/
  13. this analyst note published by Barron’s: http://blogs.barrons.com/techtraderdaily/2016/05/12/amd-taking-gpu-share-from-nvidia-says-wells-all-eyes-on-polaris/
  14. dividends: https://www.stockgumshoe.com/tag/dividends/
  15. the CFO’s commentary: http://files.shareholder.com/downloads/AMDA-1XAJD4/2058315882x0x891766/878F4C1B-2D21-4C3C-8882-43D78A363F58/Q117_CFO_Commentary.pdf
  16. “Neural Imprinting”: http://www.stockgumshoe.com/reviews/radical-technology-profits/how-neural-imprinting-will-wipe-out-pain-and-generate-2-86-trillion-in-new-wealth/
  17. Keith Schaefer: https://www.stockgumshoe.com/tag/keith-schaefer/
  18. press release here: http://s2.q4cdn.com/717426720/files/doc_financials/2016/Q1/CET-Q1-2016-Earnings-Press-Release-2016-05-12-FINAL.pdf
  19. solar: https://www.stockgumshoe.com/tag/solar/
  20. Canada: https://www.stockgumshoe.com/tag/canada/
  21. wrote about Grenville Strategic Royalty: http://www.stockgumshoe.com/2016/05/friday-file-gold-grenville-and-some-buys-and-sells/
  22. Alberta: https://www.stockgumshoe.com/tag/alberta/

Source URL: https://www.stockgumshoe.com/reviews/radical-technology-profits/friday-file-cognitive-sight-examined-plus-some-updates/


16 responses to “Friday File: “Cognitive Sight” examined, plus some updates”

  1. justice51 says:

    Travis…. Once again a big “THANK YOU” for all the money you save us irregulars !!! I get countless E-mails from these financial newsletters and all their off-shoots (brothers, sisters, uncles) financial hacks requesting my $49, $99, $149, $500, $2500 for the next “big” thing that will make me a million…. All I have to do is wait 2-4 days, open Stock Gumshoe and I have my answer, with plenty of research and input from fellow irregulars…..You provide an invaluable service, keep up the good work…..

  2. bluesharpbob says:

    Been a Morningstar subscriber for a while, their analysts estimate the fair value of NVDA as $28 a share, & voice concerns over the rapid product cycles & heavy competition the company faces in its markets.

  3. consultnick says:

    Travis….I completely “ditto” justice51’s comment.

    I continue to be amazed at your broad band of knowledge, your ability to put theses together, as well as your fine-tuned “gumshoe skills”….all necessary to sleuthing out the truths from the fusillade of BS some of these financial newsletters throw as us. You could easily flip to the dark side, but your moral and ethical compass stay solidly locked on the pursuit of truth! Thank you for saving me endless amounts of time…..and serious investment mistakes!

  4. Normally Dubious says:

    NVDA – the one good recommendation I got in a year subsciption to M. Fool….

  5. maddogdne says:

    When you see the nanotech-engineered silicon-neural interface connected directly to the occipital cortex on the back side and streaming data from the glasses-based optical system on the front side, Wi-Fi enabled AI chip included, with some bonus radiations over the convexity to the frontal lobes, THAT’s the company you’ll probably want to buy some stock in…but for now, Travis, thank you for the essential cold water that brings us to our senses – at least briefly. Even after reading Kahneman’s “Thinking, fast and slow” it amazes me that I am so easily manipulated by the newsletter writers!

  6. News from Crius today — the Trust will be buying out the rest of the company that it doesn’t currently own (currently the trust owns about 43% after also bumping up the ownership stake last year).

    That doesn’t necessarily mean anything significant for the financials, though it will make the balance sheet look a bit cleaner — it shouldn’t be dilutive to earnings or impact the results, though it will, I think, remove a little bit of the potential risk because the Trust will now directly control the entire company. I don’t think the ownership structure was a meaningful concern before, but this does at least simplify things — I thought at first the market agreed and was bumping up the price a few percent, but it appears the reverse has happened and it’s taking a bite out of the shares.

    Doesn’t make sense to me that the company should be less valuable as a 100% controlling owner than as a minority owner of the same asset… so I’m holding. The only risk that comes to mind is that the insiders could be trying to get out of their share of the rest of the company for some nefarious reason, but I don’t see any indication of that in their results. We’ll see, could just be a knee-jerk reaction in the stock.

  7. Timothy Ludolph says:

    I know I am a little late with this comment, but I have been looking at a couple of teaser articles the last week or so about a tiny $6 ($7 in one) stock that sounds a lot like Robinson’s. Money Map Press and Stansbury and agora are all interconnected, not in business but they all have a very close relationship. What I mean is that Money map I think came from Agora and they were joined to Stansbury a few years ago. And they moved down the street from them in Baltimore. My point they all push the same stuff. Remember KLH that was touted last year as the LIving Metal. all 3 firms touted it fairly close to each other and you figured out the co. The come on from Robinson was that it was to be listed on Nasdaq. ( it finally was on Nasdaq and flew up to $10 from $1.60. Anyway this tiny co at $6-7 was put forward by all the Baltimore Boys. So “tiny” and a $6 stock doesn’t sound like NIVIDIA. In fact I came to your site Travis to see what stock the Thinkolater came up with. I am probably wrong, but I am sure those companies have a hotline between them in Oriole land to share the Ideas so that they can synchronise them and not spill all the beans. Anyway I am gonna take a closer look myself tomorrow to find this tiny co. OH I forgot the tease I saw tonight said that Microsoft & Google put a couple Billion each into this technology and that it would be the smartphone killer. So we will give up our phones to put on some cool glasses or something. They did mention the co. Google bought and Microsoft bought-that you also mentioned in your piece so I don’t know I am more confused now than when I started typing so maybe I do need some artificial Intelligence. AU Revoir

  8. Stuart Uffner says:

    I love reading your analysis on these “touts”. I really try not to buy into the hype of these sales pitches but it is hard to find the right places to invest, when your just a novice trying to preserve and grow savings so we don’t outlive our money.
    I wanted to inquire about this new pitch, which seems familiar to me, since these guys always recycle old pitches, anyway, it goes like this: “A $6 company holds critical patents that could help unlock a projected $37 trillion fortune” and it goes on from there.
    Any thoughts on this pitch?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.