3Dfx Bear's Den
Jeff Fischer (TMF Jeff)

When it comes to discussing the strengths of a particular equity, invested parties are always willing to wax poetic about how "their" company is going to revolutionize its industry. Though Fools aspire to be optimistic, having too much optimism that isn't grounded by a healthy and researched skepticism is not Foolish. As the co-author of the Fool Portfolio's 3Dfx buy report, it seems appropriate that I present this bearish argument for the shares as well.

Why?

Knowing the risks facing a company helps an investor place a fair value on the stock, as well as analyze unfolding situations correctly. It behooves all Foolish investors to create bearish scenarios for their investments at the same time that they outline a bullish story. On the message boards of popular stocks the bearish posts should be the ones most cherished. When everybody already knows the "great" side of a particular story, the next question should be, "What are the risks?"

On to 3Dfx.

I'm sure that Louis presented a more than adequate bull argument (I'm not allowed to read it until after writing this). He probably restated that fourth quarter sales soared, resulting in first-time profitability. He probably mentioned the 150 games created for 3Dfx technology. He probably mentioned branding. (Branding is only really valuable, in my mind, if the company can enjoy the pricing power and increased sales that a brand can create over decades of time. Near-term branding might not be a big deal -- at all. Reebok had a great brand once. So did Atari. And most people have never heard of 3Dfx.)

So, the bull story is well known.

But even in the short time since David Gardner and I wrote the 3Dfx buy report, the story has changed. The primary risks that I see now are:

1. The size of the market...

...in relation to

2. The ramp-up of Intel's new 3D technology chips

...as well as

3. Intel's alliance on 3D, which is with another company

...and finally

4. The valuation of 3Dfx.

With all equity investments, there are business risks and valuation risks, and 3Dfx sports a higher-than-normal quantity of both. On the business side, there are 38 companies listed on Hoovers as "competing with 3Dfx." Over half of those companies don't compete directly, but many do on some level. In the Nasdaq stock listings alone, there are five companies quoted beside 3Dfx that have "3D" as part of their corporate name.

I've looked at these relevant companies, though, and the risks don't seem serious. 3Dfx has a large technology lead -- at least one generation -- so an investor can discount many of the small players and move directly to the next risk, the big gun, and that's Intel.

Intel's alliance on 3D technology has been drawn and unfortunately it is not with 3Dfx. There are a few reasons for this (including that 3Dfx products are for high-end users, while Intel's technology is aimed at mainstream PC buyers so it might not compare performance-wise), but let's think in terms of years rather than just right now.

On February 12, 1998, Intel announced its new Intel 740 3D graphics accelerator chip, which was engineered by Real 3D, a company owned jointly by Lockheed Martin <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: LMT)") else Response.Write("(NYSE: LMT)") end if %> and Intel. Intel recently took a 20% stake in Real 3D while at the same time selling all of its shares of 3Dfx. The investment constitutes a partnership, and Lockheed Martin (owning 80% of Real 3D) is no slouch of a business partner. Lockheed's Real 3D company holds 40 key patents in computer imaging and 3D graphics.

Meantime, the industry has already adopted Intel's Accelerated Graphics Port (AGP) interface, and though competing graphic chipsets can of course run with it, Intel's 3D chip can be sold with Intel motherboards less expensively -- and most conveniently.

This brings up market size.

The majority of computer users will be very happy with Intel's technology. Only a small handful -- the hardcore gamers -- will want to spend the extra hundreds of dollars to go to Voodoo technology from 3Dfx. And that's right now. But consider two years from now. Intel's technology will continually improve (economies of scale will help) until any performance difference between its technology and more expensive technology is less noticeable. Then even fewer users will want (or need) to pay up for 3Dfx.

The slogans of Intel and Microsoft <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: MSFT)") else Response.Write("(Nasdaq: MSFT)") end if %> should be, "What your company can do better, ours can do much bigger."

As we all know from the Apple Computer <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: AAPL)") else Response.Write("(Nasdaq: AAPL)") end if %> story, or even the IBM <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: IBM)") else Response.Write("(NYSE: IBM)") end if %> OS/2 story, the best system doesn't by default win the market. It's the best-marketed and most accessible system -- to everyone -- that usually wins the consumer (and then forms the standard). Sure, 3Dfx has carved a niche for itself and should see tremendous sales growth this year and probably next, but the size of its market in the long-term is a concern -- especially when much less expensive technology is being pumped out by the computer industry leader. And Intel's technology should suffice for more and more users as it advances. Meanwhile, as far as 3Dfx's possible attempt to get into the business graphics market, Intel already has the business alliances (and ownership of the motherboard) that make it much easier for companies to work off of Intel technology.

It's a sad fact, but due to switching costs and "peer pressure," it makes the most business sense to use the technology that everyone else is using. Companies such as Intel and Microsoft, whose products set the trend, always grow stronger in every direction. Games will be optimized for use on Intel technology as well as with 3Dfx's -- you can count on it. Most people will just use Intel.

Valuation is the next risk. 3Dfx's earnings estimates doubled from $0.50 per share to over $1.00 per share in January, but the stock is down since then. If a stock can't sustain a rise in price after its earnings estimates are doubled, then arguably the market already factored the upside into the current price and the stock could be fully valued. So any slip might not be pretty (look at peer stock valuations). Most likely more earnings surprises are now needed to get more stock price appreciation. That could happen this year, but the risks are high. The coming 14% share dilution with a secondary offering doesn't help earnings per share power at all.

Next: The Bull Responds