Skip to main content

HDTV: What's Wrong With This Picture?

SpatiaLight's projections look a tad optimistic if you examine the underlying numbers.

Shares in the high-definition television technology company



have more than tripled since late September.

But unlike the sharp pictures the company hopes to create on a grand scale, the logic behind the company's breathtaking run-up is fuzzier than the picture from a 1959 DuMont.

Late last year, SpatiaLight -- one of several companies developing a TV picture technology known as liquid crystal on silicon, or LCoS -- announced deals to sell LCoS products to three different China-based TV manufacturers. SpatiaLight Chairman Larry Matteson says these deals represent potential revenue of $70 million for the company in 2002 and a total of $400 million over the next three years.

The company appears to have hit some synergistic hot buttons, appealing to the investing public as a play on both HDTV, the next generation of television technology, and the colossal Chinese consumer market. SpatiaLight's stock jumped from $1.78 in early October as high as $8.15 earlier this month. But that's a modest climb to some: One sell-side analyst has anointed the stock with a one-year price target of $30 and a two-year of $100.

Despite a pullback Monday that sent the stock down $1.44, SpatiaLight was still generously priced Monday night, with its $5.79 closing price implying a market capitalization of $139 million. The stock traded Tuesday morning at $5.99, up 20 cents.

Not bad for a company with revenue totaling $167,400 since the beginning of 1998; with $66,000 cash on hand at the end of September; with losses of $7.2 million for the first nine months of 2001 alone; and, finally, blessed by a "going concern" letter from its accountants that casts doubt on the company's ability to stay afloat.

Moreover, the "potential revenue" numbers the company has issued seem more hypothetical than certain, given SpatiaLight's technical challenges, competition, and economic assumptions. For starters, the company expects that an LCoS-based HDTV set will sell in China for $3,500 -- 75% more than the maximum that most Americans pay for their sets, and more than three times China's per-capita annual income.

"This is just straight, bloody, venture capital," says an investor who has shorted SpatiaLight's stock, speaking on condition of anonymity. "It's pure concept. And it's not even a concept where you're selling into a proven marketplace."

Rear Projection

For an orderly look at the SpatiaLight story, let's work our way backwards from the end product: the $3,500 HDTV sets which SpatiaLight says companies plan to build and sell in China using SpatiaLight's LCoS panels.

The sets at issue are known as rear-projection sets. Like the more common glass-tube sets, rear-projection sets come in a self-contained box. But unlike typical TVs, which make a picture by shooting electrons onto a glass tube, RP sets use visible light, not electrons, to shoot a picture onto a plastic screen with the help of a sort of internal movie projector.

Traditionally, these visible-light projectors have used miniature picture tubes, but SpatiaLight and numerous other companies are developing new models based instead on miniature TV screens known as microdisplays.

Experts say a $3,500 set is too expensive for the consumer market. And that's in the U.S. Over the years, selling prices of television sets have crystallized around a number of prices, says Gerry Kaufhold, principal analyst in multimedia at the market research firm Cahners In-Stat Group. For RP sets, those prices are $1,200 for low-end models and $2,000 for HDTV-quality units; above $2,000, most sets are sold as part of custom-installation home theater packages, with the set representing only part of the cost, he says.

Those stickers would cause a lot more shock in China, where in 2000 the market intelligence firm Asia Pulse found that the market was dominated by sets selling for less than 2,000 yuan -- about $241 at today's exchange rates.

SpatiaLight's Matteson declines to question his manufacturing customers' expected retail price of $3,500 per set. "They feel confident it's the right price," he says. "I'm not in that business."

Light Sabers

After $3,500, the next puzzling number on the SpatiaLight tour is the average $1,600 in revenue that the company says it will gain when it sells each "light engine" -- a package comprising three of the company's proprietary LCoS microdisplay panels, associated optics and necessary electronics.

By one RP industry participant's rule of thumb, multiplying the light engine's price by 2.5 gives you the total bill of materials for the set -- in SpatiaLight's case, $4,000. By another participant's rough estimate, the wholesale price of an RP set should be three times the light engine's cost, marking it up to $4,800 for dealers. By either thumb, the $3,500 price -- as costly as it might be to the consumer -- seems optimistically low, given the light engine's cost.

The Chinese set manufacturers, says Matteson, "were satisfied with the economics of the deal we worked out together." SpatiaLight maintains that the Chinese government is making the development of high-definition TV broadcasting and manufacturing a key economic and technical priority. In fact, SpatiaLightcompetitor

Three-Five Systems


set up an engineering and manufacturingfacility in China four years ago, and is itself working with two differentgovernment-sponsored HDTV companies. Says SpatiaLight's Matteson, "Chinawill be in our view the first major market for HDTV sets, not the U.S."

Yet even if the $1,600 revenue-per-light-engine figure holds, investors may underestimate how much of that money is retained by SpatiaLight, not paid out to contractors like Fuji Photo Optical, which manufactures SpatiaLight's light engine. The only sell-side analyst who appears to be following SpatiaLight, Ray Dirks of Dirks & Co., says he reached operating earnings per share estimates of $2 for 2003 and $6 for 2004 by assuming a gross margin of 40%. But Dirks admits that figure may be optimistic.

"If some of that goes to Fuji, we have to reduce it," he says. Matteson won't talk details, but he says, "It turns out optics are a substantial part of the total cost." Dirks has a strong buy on the company; his firm hasn't done banking for SpatiaLight, and he has no holdings in the company.

Yet So Far Away

The final challenge facing SpatiaLight is whether it can even meet the production and sales targets it has publicized, given the competion and extraordinary execution risk it faces. LCoS competes with several other technologies in the nascent HDTV RP market. Other players --

Texas Instruments

(TXN) - Get Free Report

, for one -- are clearly better funded than SpatiaLight.

Other LCoS developers for RP applications appear to be closer to commercial production than SpatiaLight, or have shipped more microdisplays (though not necessarily for the RP market). Privately held DisplayTech, based in Colorado, says it shipped its millionth LCoS microdisplay in December. LCoS chips developed by Three-Five Systems were used in high-end TV sets that Thomson Multimedia (the marketer of RCA brand gear) and Samsung displayed at the recent Consumer Electronics Show in Las Vegas.

Asked about SpatiaLight's light-engine production volumes, Matteson says the company has a "limited number" of prototypes it has used for "serious customers."

As for execution, LCoS development -- which has translated into zero RP sets on the market today -- has proved devilishly difficult. That's a point worth remembering when one sees SpatiaLight's disclaimers that its sales forecasts are subject to "performance criteria" and "technical specifications."

"Where SpatiaLight is right now is really more of handshake to try to go forward," says Chris Chinnock, senior editor of the

Microdisplay Report

covering microdisplays and the projection industry. "There are so many milestones that they're going to have to hit between now and real production, and so many places to go wrong. ... Moving from a few prototypes to manufacturing is a huge step."

In other words, going from zero to $70 million in a year is highly indefinite.