eMagin (NYSE:EMAN ) can pack four million pixels, a 2048x2048 grid, in a display smaller than a postage stamp. At that dot density, ~800 would fit inside each pixel on my 20" 1680x1050 iMac display. And whereas my Mac's liquid crystal "LC" pixels merely filter beams emitted from a backlight, eMagin's are self-lit, no backlight required. The enabling OLED architecture is hard but intrinsically smaller, lighter, less wasteful of energy, and sharper than the backlit incumbents. eMagin, which was founded in 1994 and produced the world's first OLED microdisplays in 2001, is OLED's small form factor pioneer, both technically and commercially.
Generally speaking, microdisplays are used where larger ones would be intolerably bulky. Cellphone displays are just 4-6", ~100 grams, but microdisplays shrink to less than two grams. Whether the difference is salient depends on the application. All "near-eye" military systems use microdisplays: they're mounted to soldiers' heads, or guns, or embedded in handheld viewers to deliver functions like night and thermal vision, or graphics overlaid on background imagery to coordinate troops. But some gaming goggles stick, for now, with phone-sized displays to cut costs.
Camera viewfinders are the only other mature microdisplay application, and that, combined with defense, sums up to less than 1m displays and $100m per year, I think. The microdisplay market is small.
But it might be ~40 times larger soon.
When in 2012 Google (GOOG, GOOGL) launched Glass, and in 2014 Facebook (FB) bought Oculus Rift (for a shocking $2b), it was on. Into the fold rushed ~40 companies developing, prototyping, and preparing to launch mass-market Virtual and Augmented reality, convinced that this, finally, is not a false start. All the enabling techs - processors, sensors, software, and displays - have advanced, shrunk, and cheapened since past head fakes. And the proliferation of the smartphone bridges the gap between traditional and wearable computing, offering a taste of what's possible, while enabling developers to accumulate experience in adjacent and overlapping uses.
For several reasons, I won't obsess over estimating market size. Let's say HMD estimates are too rosy by 3/4's. So what? Then 10m HMDs ship in 2018, with two displays for most units, at ~$50 per display. That's a $1b pie, but will $70m eMagin get to eat? Share, not size, is the question, and since share-related questions are more idiosyncratic, it's where the insights are found anyway.
2) The Thesis: Military Margin of Safety, Plus Mass-Market Moonshot
My thesis has two pillars - one sturdy, one speculative - mirroring eMagin's strategy.
1) The Sturdy Pillar: High-end sales leadership, secured by best and rapidly progressing displays, justifies the valuation by itself.
eMagin's strategy has long been Defense first, Consumer second. Performance leadership now, cost reduction later. Given the company's high costs, it's not like it had a choice, but the military before mass market sequence is a sensible route that's worked for innumerable young techs.
eMagin's displays are recognized to be so good that it's hard to identify high-end systems (say, $5k and up) with anything but its OLED inside; the higher-end you look, the more recent the design win, the more likely it's eMagin. And the few elite markets the company doesn't yet own - like flight helmets - it ultimately will. eMagin has already eclipsed its competition on most display specs, and its pace ensures that real soon now it'll be all specs.
I think <1m microdisplays worth ~$100m shipped in 2014, which is what I get when I roll up the sales of a couple public companies and guess more blindly on a few private ones. (Though I think the figure was more like ~$200m a few years ago, when the military was stronger, and before smartphones ate into camera viewfinders.) Of that, ~$30-40m was (US and allies) military, probably a similar chunk was for cameras, and the rest was a motley mix of half a dozen other things. Let's say half of that $100m is high-end stuff - systems that cost $5k and up where $500+ displays are tolerated. So the 2014 high-end market was ~$50m on ~100,000 displays.
How big by 2018? Recall experts expect ~40m total units, which I've cut to 10m. Let's say 5%, or 500,000, of those are high-end; that's 1m high-end displays. Suppose the price per high-end display drops to $200. This series of pessimistic assumptions implies a doubling of the market between 2014 and 2018, a CAGR of 20%. (You won't find published estimates this low anywhere.)
That'll be blown away if just a couple applications take off. For example, there are ~7m blindness-disabled Americans. Let's say half of those have candidate conditions for eSight, a recently commercialized eMagin powered goggle that, through contrast enhancement, helps blind people see. Suppose 10% of those candidates purchase eSight type devices: that's 350k units, 700k displays right there, which might annualize to ~150k. There are 18k active surgeons in the US, and 100k+ in the developed world, who are targeted for see-through devices that label body parts and sequence processes. There are tens of thousands of engineers engaged in troubleshooting targeted for the same sort of thing. In short, there are many potential high-end infant applications, any of which alone could outstrip the currently dominant Defense market on which eMagin relies for 75% of its revenue.
If eMagin merely keeps gaining share in the $30-60m military space - which I think is a lock - then its $70m market cap will be justified. And if any other performance markets scale up, then eMagin's sales should easily double or triple off their $30m baseline, implying solid upside.
(A key point, which I'll detail later, is that OLED's difficulty and eMagin's head start should prevent the commonly overlooked situation where a growing market dilutes one's share and margins; entry here is very hard.)
2) The Moonshot: Cost reduction enables migration into exploding mass markets.
Once the high-performance island was fortified, eMagin intended - when the time was right - to invade the mainland. The time is right. The emphasis of the company's talk and action has shifted from technical breakthroughs to how it intends to profit from them. It has recently prototyped OLED VR goggles for gaming, which deliver an unprecedented combination of picture quality and compactness. And it has hired a well-known marketing vet to promote them. Meanwhile, the company is scenario-planning for that big hypothetical win, figuring out how to drive scale up and costs down 10+ times over the course of a hypothetical 1-3 year ramp phase.
eMagin's pitch is essentially: Our cost and scalability is fixable. But the competition's performance isn't.
So, from the safety of the sturdy pillar we can speculate-bet, for free, on the moonshot.
Rather than drown readers in an ocean of upfront technical detail, I've chosen to sprinkle it through the report, injecting it on a case basis to explain who is winning where and why, with links to more detailed info for those interested.
3) Why eMagin's OLED is Beating Kopin's LCD in the Military
A simplifying thing is that eMagin and its prime rival, Kopin (KOPN), collectively dominate defense, and since 75% of eMagin's revenue is defense, the sturdiness of the company's pillar essentially boils down to the outcome of this fight.
Kopin is mature, and so is its tech: the company has shipped 30 million liquid crystal microdisplays, mostly for cameras, plus 250,000 for defense systems. eMagin's cumulative tally is likely ~400,000 displays - a guess based on ~$200m cumulative sales at $500 per display - ~75% for defense. Kopin and LCD are old; eMagin and OLED are young.
Experience is good for scalability and cost: Kopin can make 300,000 displays per month, including models cheap enough for consumer applications.
Experience is bad for innovation, since the juiciest fruits are picked first: back in 2004, Kopin expected to double maximum LCD resolution soon, but instead, it hasn't budged in 11 years. Kopin stagnated at ~1200 horizontal dots, while eMagin leapfrogged from 850 to 2048. That's because Kopin's pixels got stuck at 15 microns as eMagin's shrunk a couple times to 9.
Resolution isn't everything, but it's a useful progress heuristic because it's objectively measured. It shows Kopin's LCD has stalled for a decade at a spec that limits potential: virtual reality needs a lush, wide field of view images beyond what Kopin is capable of. That's why the military paid eMagin to develop high-res displays for immersive simulation, training, and psychological therapy (this system was the precursor to the gaming VR prototype that's demoing now, discussed later).
But the primary reason eMagin first started stealing military deals was efficiency, not resolution. Self-emissive OLED is a much less wasteful tech than backlit liquid crystal.
To validate the power gap, you could check the specs, but it's simpler to observe who wins where and why. eMagin's CEO notes:
"... when I look at the LCD, which is what Kopin does, they have an LC [liquid crystal filter] then below them they have the liquid crystal backplane. They put the liquid crystal on the top, they have the backlights there. And so they are not as quite at the same disadvantage as Himax is [he's saying LCOS is physically too big]. So let's say they are small just like ours is small.
... the reason the military - let's forget about anything I say. The reason the military went to OLED displays was power. And I will give you just an example, a long time ago there was an LCD that a company [a systems company] was looking at and the company came to us and said, they had a power budget of 1,500 milliwatts and that the LCD display took a 1,000 of it, and this was SXGA.
And they took our - this was long time ago, then new SXGA, and they said, "200, 250 milliwatts, okay, you win." So, I think the power question - and you know the most famous glass eyewear you hear about is Google [Himax LCOS inside], I'm told they last two hours. So they need power. They need a much more power efficient display.
OLED's potential was obvious twenty years ago, but it couldn't be realized until recently. Before ~2006, eMagin's displays would dim too rapidly with use, so they couldn't achieve reasonable life (that's what's historically constrained Samsung (OTC:SSNLF) and LG (OTC:LGEAF) in OLED phones and TVs too). So Kopin/LCD ruled despite its problems, and eMagin lived off military research contracts.
In ~2006, eMagin blossomed. New materials enabled a major brightness and lifetime breakthrough, and its revenue, which had previously never topped $5m, surged.
EMAN vs. KOPN Military Revenue ex. research funds
Let's interpret. Yes, both grew fast between 2006 and 2009 as the military splurged, but Kopin entered the period as a proven display supplier (its aggregate sales were $90m in 2000). Conversely, eMagin was still formally a development stage business in 2003, so its rapid emergence in a sector with notoriously grueling design and test cycles was a shock.
From 2009 to 2013, the military pie shrank from ~$70m to $30m, but eMagin's share grew from 25% to 70%, directly attributable to contracts stolen from Kopin, like the big, high-profile Enhanced Night Vision program (the theft previously referenced by eMagin's CEO, I think).
"Market share loss" isn't in Kopin's vocabulary: the company always blames its weakness entirely on budgets.
Things reversed in 2014, as Kopin regained military revenue parity with eMagin, but on a fluke. What happened is that eMagin had to stop shipments on certain displays because a mundane wire bond issue was discovered, so Kopin, which had previously supplied the product, was able to fill the vacuum. That's not the stuff of a permanent resurgence. eMagin expects military growth in 2015, while Kopin speaks cautiously about the potential for piece-meal orders. The fluke will be revealed as such within a few quarters by the results; it's revealed today by how the military is behaving.
Military R&D funding is a big deal: it helps financially, and it signals that the military wants you to solve a problem for it. The problem the military is paying eMagin to solve is Kopin.
The military isn't funding Kopin R&D anymore, which, of course, the company blames on budgets. But in late 2014, eMagin was awarded $6.8m primarily to develop high-brightness displays for pilot helmets.
I believe Kopin is the avionics incumbent, or rather, "Fourth Dimension" LCOS (not LCD) which Kopin acquired in 2011 (if you see Kopin resolutions above ~1200, it's actually Fourth Dimension). The reason OLED isn't already in pilot helmets is that high brightness is LC's last remaining technical advantage; it's easier to ramp up the brightness of a backlight than four million OLEDs the size of dust mites.
(Nits measure brightness density. Your cell phone is ~450. Consumer see-through requires ~5,000, avionics ~10,000. A couple years ago, eMagin couldn't crack 1,000 nits reliably with color, but the company is hitting all the milestones the military is asking, and it has recently prototyped full color up to 7,000. The R&D contract specifies 10,000+.)
But pilots using LCOS helmets complain of "green glow" where their visors should be transparent, an artifact of backlight seeping through imperfect liquid crystal filters. A problem solved by a bright enough OLED pixel that simply shuts off when it ought to be dark.
Money talks: This is just the latest instance of military funding designed for eMagin to displace Kopin.
Kopin obscures the writing that's all over the wall by studiously ignoring its rival: eMagin doesn't even earn a reference in the company's 10-K! Kopin prefers buzzword babble about market size, wearables, and new paradigms, which is all fine and true, but who benefits? It must be annoying that eMagin never shuts up about its "principal competitor", how much better its displays are, what it has stolen thus far and what it's being paid to steal next. These firms behave as you'd expect if both understood that inertia is the only thing keeping Kopin relevant in defense.
4) Why Himax's LCOS Won Google Glass
Another simplifying aspect of microdisplays is that there are just three basic commercialized display architectures, each sponsored by a clear leader. We've discussed two, and now we're onto the third.
Google Glass generation #1 had Himax's (HIMX) LCOS displays inside. Cost and scalability kept OLED out, but what's LCOS got on LCD?
Structurally, LCOS and LCD have much in common. Backlights provide the general illumination. And a grid of voltage controlled liquid crystal filters - one for each color at each pixel - translate the backlight into a dynamic image.
However, LCOS filters act as controllable mirrors (reflecting or absorbing backlight), whereas LCD ones are windows (transmitting or absorbing backlight). Whereas clear filters lay simply over the top of backlight, reflective ones are arranged in a sort of complicated cube, wherein white light is color split by prisms, routed to color specific reflective filter panels, and then re-routed through the display into your eye.
LCOS looks fragile, which is probably why ground troops don't use it for anything, but it does some things well.
LCOS achieves pixels 1/5th the size LCD does. The Glass display was low-resolution, but much tinier than LCD could achieve.
Secondly, since reflective filters absorb less light (in the "on" mode, where you'd prefer zero absorption) than transmissive ones, LCOS can do the same brightness with a less powerful backlight, saving power.
So Google chose LCOS over LCD primarily on resolution and efficiency grounds.
It's said that displays can't cost more than $35 in consumer systems. I think both LC techs can do that on their lower-end options, but eMagin has probably never sold a display for less than $200 (the average price was ~$380 in 2007, which was the last time the company published that metric).
eMagin's pitch (or my imagined version of it) for Google Glass and augmented reality is essentially this: LCOS is a stopgap fix, or a solution for what's ultimately your low-end version. It's better than LCD, and Glass got pretty good reviews, but your users complained about battery life and heat generation; complained a little about contrast (even though your FOV was tiny!); and a little about bulk. This is OLED's bread and butter.
Yes, our costs look scary, but we're young and we've lived off nothing but several hundred unit military spec runs! If we're manufacturing a less cutting-edge display in 100 times greater volume with relaxed military specs, we'll fix our costs. And LCOS's limitations will remain.
(eMagin won't be in Glass #2, but it must be on the Augmented Reality radar, especially since, as the tech matures, versions aimed at different market segments - some of which won't require consumer price points - will begin to proliferate. Oh, and note that Kopin's losing in defense, and has lost out on the highest-profile consumer deal. This sort of looks like a two-horse race.)
5) Why isn't the Oculus Rift Using Microdisplays?
The Oculus Rift VR system embeds a 5.9-inch Samsung OLED display - not a microdisplay - which is optically sliced in half. Larger displays are acceptable - but not preferable - in couch potato applications. And they're cheap, since Samsung is producing hundreds of millions of OLED phone displays each year. (Samsung's phone displays are the first OLED to achieve mass-market status. They're approaching cost parity with LCD - see, it is possible - and users rave about the picture quality. It's not clear that LCD phones will survive OLED.)
But a phone-sized display in head gear makes for an awkward 1 lb protrusion: this isn't what Verner Vinge had in mind. To show what's possible with OLED microdisplays - and I'm guessing in utter shock at the $2b Facebook paid for Oculus - eMagin spent a few million bucks in 2014 secretly developing a competing prototype, an upgrade of its military/medical HMD.
eMagin's unnamed prototype achieves the same resolution per eye as Oculus (approaching the limits of human visual perception, which is 60 pixels per degree FOV) in a half-as-heavy, tighter-fitting, flip-up capable set of goggles. Its microdisplays solve the "screen door" problem that's a product of the dark space between the widely-spaced pixels of cellphone-sized displays. Initial testers have raved about the picture quality (a big VR problem is that it makes people nauseous, and it's not yet clear whether improved picture quality can solve that). And the VR stud who invented them at eMagin, now employed prominently in Google's HMD unit, calls them the "highest-performance VR HMD ever produced".
eMagin spent a few million bucks and produced a brilliant VR prototype that's at least achieved its basic purpose, which is to show what the company's displays enable - surely, high-end professional VR is eMagin's to lose.
But whether its gaming prototype attracts consumer interest depends not just on performance. The company's got to be convincing on the cost problem.
And it's got to do be convincing as soon as possible, while it's still the clear-cut technical leader.
6) The Durability of eMagin's Leadership Within Micro OLED
eMagin is the micro OLED technical pioneer, but the space is looking more vulnerable to competition.
Manufacturing even basic OLEDs is hard. It requires depositing on a chip an ultra-thin, uniform, pure layer of material that emits light when electrically shocked. The deposition process consists of a cooling a chip in a vacuum, so that the injected evaporated emissive material condenses on it. The other hard process is creating an almost perfect seal, because tiny external permeations will ruin the image.
This difficulty has repeatedly stumped LG in TVs, Samsung in TVs and phones, and eMagin in microdisplays. But difficulty is what makes patents and process know-how valuable; this is what experience curves are made of. So we've got some time, but the clock is ticking.
eMagin's setbacks are summarized by the ~$200m it lost before turning profitable for the first time in 2009, and the re-emergence of manufacturing problems in 2011.
eMagin's results peaked in 2010, then reversed. What happened is that the company ran out of capacity, and the subsequent expansion wasn't smooth.
It had been using the same deposition machine for 15 years - a low-throughput, labor-intensive R&D tool not intended for scalable production.
Pioneers can't simply buy capacity: off-the-shelf OLED deposition machines suitable for eMagin don't exist. So in 2009, the company engaged SNU Precision to build a custom tool for ~$7m, imagining that after an 18-36 month process of design, fabrication, receipt, testing, and fine-tuning, its capacity would expand 10-fold via 5x higher wafer throughput and 2x higher yields.
Leaders explode landmines. The gross margin decline from 60% in 2010 to ~25% in 2013-2014 reflects: a) simultaneous operation of both tools; b) new tool downtime, including a partial redesign; c) that the new tool's yields are just now finally approaching the old one's; and d) headcount growth from ~60 to ~100 to troubleshoot. Ironically, all of this was exacerbated by great technical progress - before 2008, one display comprised most sales - that broadened its mix.
Luckily, eMagin has exploded its worst landmines without having to sell shares to finance a prosthetic limb; it's got $7m in the bank, no debt, and is returning to profitability right now. The company's yields are trending up, and we won't have to cope with another major capacity expansion unless eMagin lands a whale of a deal.
The reason EMAN surged 25%+ on 1Q15 results is yield gain - finally! Product margins were 42%, up 10-15 points compared to the prior eight quarters, but still below its 50% medium-term goal and 60% prior peak. So the company broke even on just $6m in revenue. Since the $26-29m 2015 sales guidance implies $7-7.5m for the next three quarters, eMagin should return to positive profits and cash flow this year.
Anyway, the point is this: If eMagin intends to exploit its high-end dominance as a segue into mass markets, it needs to do it sooner rather than later, because its very real and valuable technical leadership will decay with time. And the best way to widen the gap between the company and its trailing competitors is to ramp up volumes ASAP.
7) eMagin Has Prototyped What Looks Like The Future
Flash-forward 10 years. Something like Google Glass is ubiquitous, every video game system comes equipped with VR goggles, and hundreds of millions of HMDs are sold each year. What sort of display is inside? Maybe something like what eMagin's already prototyped.
OLED TV's and microdisplays use a pixel structure called WRGB (White, Red, Green, Blue). At each pixel, there's a white emitter with red, green, and blue filters on top. The alternative called RGB (or 'direct emission') swaps each white emitter for tri-color ones, eliminating the filter layer entirely. With out filters efficiency and brightness leap ~4-5 times, widening OLED's already large advantage.
Though direct emission is standard in cell phone displays, LG wisely abandoned it in TV's on cost and lifetime grounds, problems that seemingly precluded direct patterning at microdisplay pixel densities indefinitely.
But in 2013 eMagin told investors they thought they could do it; they prototyped the world's first direct emission OLED microdisplays in June, 2014; and the company recently installed equipment to enhance yield and resolution.
The display won't be commercialized soon. But it's a brilliant showcase of eMagin's capabilities - another "first ever" for eMagin - and the sort of display you'd expect to find in HMDs once this market advances from the Palm Pilot to the iPhone stage of maturity, especially since direct emission's benefits are worth its costs in the smallest form factor, mobile applications.
The dictum to "think probabilistically" applies strongly here. We've got a defense-oriented baseline we can peg accurately enough, and it's enough to limit our downside. But much of eMagin's value is in the likelihood we assign to uncertain yet stupendous medium-term opportunities.
Our baseline OPEX run rate is $10m. Our sales baseline is ~$30m, which is what's expected in the next three quarters annualized. It requires no major new wins or design ramps.
Gross margin was 40% in 1Q15. The company "see(s) no reason" it can't maintain that for now, but it targets a return to 50%.
a) So our baseline near-term scenario is operating income of $2m on 28m fully diluted shares, ~$0.07 EPS (it's got big long-lived NOLs).
b) Shocks to the upside are more likely, but if we pessimistically model $27m revenue and 35% gross margins, then the company will lose $500k, or 2 pennies.
c) If, by the end of 2015, yields have kept rising and gross margins are 45-50%, then operating income would be $3.5m-5m, and EPS $0.12-0.18, on our baseline revenue.
If I had to guess, the "b" scenario means a $2-2.50 share price, "a" means $3-3.50, and "c" means $4-5.
d) If in 2016 the company delivers $40m sales, 50% gross, and $11.5m OPEX, it'll earn $8.5m, or $0.30 per share. The incremental $10m in revenue could be delivered by a partial rebound in military spending, or by the commencement of new deals. For example, eMagin is inside the next generation of Enhanced Night Vision goggles, which are supposed to start shipping in heavy (relative to eMagin's run rate) volume in 2016. Avionics is another candidate to ramp late that year. It's a real crapshoot outside the military, but as I mentioned, eMagin is in a couple promising young, recently commercialized infant techs that could emerge. So $40m, and something like ~$0.30 is plausible for 2016, even in the absence of mass-market adoption. The last time eMagin neared that level of growth and profitability, the stock shot to $9; and that was before mass-market HMDs were in the news.
e) Finally, the way I'd prefer to make my money in eMagin is some random Tuesday at 8 AM, with some news containing a combination of the words "eMagin", "Big Tech Firm X", "Licensing", "Strategic Investment", and "Consumer AR/VR". A concrete indication that eMagin has persuaded someone in the consumer space to begin developing on its tech would send the shares soaring to potentially inappropriate levels as psychotic momentum investors jump aboard. It's not that long a shot and that should petrify anyone crazy enough to short this stock.
That's a lot of optimistic speculation, so I'll conclude by reiterating the risks with the most bearish scenario I can envision, appropriately labeled "f".
f) eMagin's revenue stagnates at ~$25-30m through 2016 because the military pie fails to snapback, and Kopin holds onto designs where its displays are "good enough", while the avionics opportunity is delayed. eMagin's gross margins stagnate at 35-40%, so it roughly breaks even; and without any volume wins, the company is unable to prove it on costs and scalability. Meanwhile, mass-market HMDs are delayed too: they don't begin growing robustly until 2018, and in the interim, competition within OLED (which I'm convinced ultimately wins) matures, so that by the time mass-market HMD's take off, eMagin faces several competent OLED rivals.
There's much more that could be and is said on a daily basis in this rapidly evolving world. I hope this article gets folks up to the speed necessary to digest that newsflow. And I'm looking forward to collaborating on it in the comments, especially with the help of those few, poor, pathetically patient, shockingly informed EMAN bulls who await their ultimate vindication for the exabytes of insights they've loaded onto obscure stock forums over the years. I think it's coming.
Editor's Note: This article covers one or more microcap stocks. Please be aware of the risks associated with these stocks.
This article was written by
Disclosure: I am/we are long EMAN. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.