Fool To Buy TDFX
January 05, 1998

**This trade is being made under the regular portfolio policy, namely, once The Fool announces an intention to trade, that trade will be made within the next WEEK, as opposed to the next day. For more detail, please read the "New Trades" section of the Fool Portfolio.**

Buying

3Dfx Interactive, Inc.
Nasdaq: TDFX
Type: Small-Cap, High Growth
Closing prices: $20 3/4 bid, $20 13/16 ask
http://www.3dfx.com

Fool.com discussion board: click here!

Market Cap: $259.1 million
Trailing 12-month sales: $26.2 million
Price-to-sales: 9.9

Recent cash and equiv.: $33.8 million, or $2.79/share
Long-term debt: $0
Trailing 9-month '97 earnings per share (eps): ($0.31)
Next quarter reported: Q4 '97 (Dec.), around Jan. 20th
Consensus EPS for fourth quarter: $0.08e
Consensus EPS for fiscal year '97 ended Dec: ($0.25e)
Consensus EPS for fiscal year '98 ended Dec: $0.50e
Fool Ratio:N/A

Trade: Buying $11,000 worth of shares

OVERVIEW
We are gamers, we are early adopters, and we have bought and loved this company's product.

What do you get when you mix together an avid gamer, the latest great computer game on the market, and $200? The answer increasingly these days: a new 3D graphics add-on card... the hot thing in the personal computer (PC) home retail market in 1998.

3D add-on cards now bring video-arcadelike graphics and speed to home PCs so effectively that they sell themselves.

Fool HQ houses a bunch of avid gamers, and the latest great computer game on the market can generally be found on numerous hard drives inside our humble offices. (And by the way, the founders of The Motley Fool encourage that; in an age where so many employees feel they have to hide their Web surfing from their boss, we actively encourage our employees to blow off steam by using that Intel microprocessor to have some fun; don't you want to work for us?) And while not all of us have $200 to spare, we've by now read enough Internet newsgroup postings to know that lots of people are still managing to save up the extra pennies to pick up a 3D card. The latest buy for the Fool Portfolio is the company that is both creating and benefiting from that consumer interest.

Our best investments have traditionally involved our proclivity toward being "early adopters" of good consumer technology items. Take the America Online service, for starters, or Iomega's Zip drives, or most recently, our addiction to Amazon.com (its website, and its 49% four-month return). Another similarity that each of those stocks shared was that each had already risen a frustrating amount (at least double) before we finally knuckled under and bought. The case will be the same this week for 3Dfx (Nasdaq:TDFX), which came public last June at $11, dropped below $9, and then stormed to a recent new high of $23 1/8. Today's market action saw the stock lose $15/16 to close at $20 13/16.

Our investment, debatably the riskiest made to date for the Fool Portfolio (and thus suitable only for experienced aggressive investors), is predicated on the expectation that we are soon to be reaching the next destination on our trail toward the ultimate multimedia PC. It began, in some senses, with color monitors 15 years ago, continued with sound cards 5 years ago, and today we reach this latest translation of the "real world" into the PC world: three "dimensions." The 3D graphics card.

These are the guys developing the best technology, and these are the guys establishing the brand name. They are one of those small companies (with all the attendant risks) that has managed to "change the rules." It is for these reasons and the ones that follow that we'll be adding this stock to our baker's dozen of Fool Portfolio holdings.

THE BIZ
A NEXT GENERATION survey showed that 3Dfx had over 55% of the new 3D accelerator market, with the next major competitors split at 13% of the market, 7%, 4%, and 1%.

3Dfx is the world's leading name-brand developer of 3D graphics accelerator technology, currently one of the fastest growing segments of the personal computer (PC) industry. UBS Securities put out a report a year ago showing sales of 8 million accelerators in 1996, 27 million predicted for 1997, and 110 million units by the turn of the century.

What exactly is this segment, though? At present, it's largely based on the growth of computer games; many of the best new PC games being released today are "3D-enhanced" (best played with 3D technology). That's because 3Dfx cards make the games much more lifelike, including enhanced graphical detail, realism, and speed (speed, baby, speed). It is 3D cards that consumers are installing in their machines, via a link to their existing graphics card, that makes the games that much more appealing and fun. As one industry publication noted recently, 3Dfx has put new technology on the PC that only ran on $40,000 machines as recently as two years ago.

3Dfx actually produces the powerful accelerator chipsets that are sold to retail consumers in the add-on-card market. Thus, over 70% of the company's revenue comes from selling its product to multimedia companies that then implement it on their own branded cards which they sell direct to the consumer, and to PC manufacturers. A good example of this, and probably 3Dfx's best customer, is Diamond Multimedia (Nasdaq:DIMD), with its bestselling Diamond Monster 3D card (retails for about $200) -- but there are numerous others, including Canopus, Hercules, Orchid, and Quantum3D. 3Dfx's focus on the retail add-on-card market will remain the case with the introduction at the end of this quarter of its new Voodoo 2 chipset (Voodoo 2 cards will initially retail for about $300).

By the way, you gotta love the names these companies stamp on their 3D cards using 3Dfx technology: Monster, Righteous, Magic, Intense, Flash, Emotion. (The current full list is in the Voodoo for Sale section of the sometimes "bizarrely-graphicked" -- to coin a phrase that should never have been coined -- 3Dfx website.) But the real brand awareness that we see here is not ultimately any one of those names, but the 3Dfx name and logo, which is included prominently on the boxes of many of today's bestselling games. Not only that, but a dynamic spinning version of the logo shows up within the games themselves, as they start up. The company is working its way into the consumer conscious in much the same way that Creative Labs did with its Soundblaster audio cards.

Back to the biz: The chipset is not now, in any mass-market way, present on the home user's PC motherboard (the "brain" of the computer). This is both good and bad for everyone involved. On the liability side of the ledger, the sales markup for the consumer is more significant than if the card were placed directly on the motherboard. But on the asset side, technology across the PC world continues to improve rapidly -- especially with 3D -- so that the cards' add-on format brings superior new technology right to the desktop of new buyers. No frustration about your motherboard's old locked-in chipset getting bypassed by the latest stuff! That's why we like the add-on format at present, though it does leave this company more vulnerable to someone else's latest gizmo; that is to say, consumer loyalty isn't guaranteed. (At over $200+ for a 3D card, these things aren't exactly cheap.) Either way, the company that has the best technology and the most computer gaming endorsements will most likely see the strongest sales. Currently, 3Dfx has created a buzz around its powerful Voodoo chipsets and has won the support of game makers, product reviewers, and consumers -- making it number one in the industry. Voodoo 3D is currently worked into 400 games that are either already available or in development -- casting a wide shadow over this industry.

Yep, 3Dfx's Voodoo Graphics and Voodoo Rush products have won numerous awards and are commonly thought to provide the best and quickest in 3D technology. So much so that demand has been backlogged significantly -- meaning, the company hasn't yet been able to produce product in quantity enough to meet surging demand (uh oh, RED FLAG!). (No investment is perfect; nothing truly good comes without risk.) In fact, all of 3Dfx's products are made by TSMC in Taiwan, which means that the company is constrained by TSMC's production capacity. Recently, though, the supply constraint improved for the current fourth quarter and is believed to be of even less consequence in the first quarter of 1998, as the market develops. With anticipation being high for quite some time in the 3D market, and with the games and technology now finally available to quench that anticipation, demand has taken off and supply is working to catch up. Due to this, 3Dfx recently topped sales forecasts and announced on December 5, 1997, that it foresees beating revenue estimates by a very considerable amount again in the current quarter, and in fiscal 1998 as well.

As mentioned earlier, the company will continue its technological lead with the release of Voodoo 2 at the end of March. A best of Comdex finalist, Voodoo 2 has received glowing praise in PC Gamer magazine, and is already the graphical standard used by the bestselling new Quake 2 by id Software [sic], the latest sequel to the original Doom. Voodoo 3D is currently worked into 400 games that are either already available or in development -- casting a wide shadow over this industry. The new Voodoo 2 chipset should almost certainly establish the performance standard in the PC graphics market, as it can run at over 100 frames per second on some games, while rival products produce screens at two to three times slower (and the average non-3D PC game runs at a fraction of that). The company pushed another 2D/3D product, named Banshee, back by one quarter in 1998 in order to bring Voodoo 2 to market early. This made sense, since management believes that Voodoo 2's early release can enable the company to outpace current revenue projections for fiscal 1998 by 25% to 30%.

To close the section, this technology isn't used only for PC gaming. 3Dfx also markets to video arcade and home game systems such as SEGA. The kicker could be the long run, though. 3Dfx eventually hopes to use its leading position in the game industry to make headway into the business graphics market. It stands to reason that one day, better graphics in the form of 3D will makes their way into Powerpoint demonstrations and Excel charts, and our new FoolPort holding is positioning itself for this success. To recap, you have a company:

(1) providing unsurpassed 3D technology to game producers, with whom they communicate regularly and well,
(2) keeping the programming standard open (nothing proprietary here, meaning that 3Dfx isn't taking any sides in the battle between Microsoft, Silicon Graphics, and other software companies to create a 3D programming standard -- 3Dfx cards support them all), and
(3) building strong consumer brand recognition with its increasingly ubiquitous logo.

3Dfx aims to capture mind share and market share of both the consumer and the game development communities. In its admittedly brief history, 3Dfx and chairman Gordon Campbell (Silicon Valley type also on 3Com's board, among others) have been successful in doing all of these.

FINANCIALS
Sales growth has topped estimates, while profits are first expected in the current quarter and in all of fiscal '98.

There isn't a plethora of historical numbers to crunch here, as the company was incorporated in August of 1994 and went public six months ago. The numbers aren't very difficult to understand, though, as the company's expenses are mainly constituted of research and development costs and selling and general administrative costs, with all of its production being contracted to Taiwan. The performance numbers that are available for study begin in late 1996, which is when the company first moved from a product development-stage firm to a revenue-producing industry leader.

After reporting record revenue of $10 million on October 23, 1997, the company has '97 revenue of $21.7 million for the first nine months, losing $3.7 million (or 31 cents per share) in the process. On December 5, 1997, the company announced that the projected 40% sequential increase in fourth quarter revenue was about 100% too low, meaning that it expects fourth quarter revenue to grow by about 80% over its previous quarter, to $18 million. That quarter is now of course finished. Assuming the investor relations and PR department over there weren't speaking idly, that would put 1997 sales at $40 million (putting today's stock price at 6.5 times sales).

As mentioned in the Biz section of this report, due to the early release of Voodoo 2 chipsets, 3Dfx anticipates that current '98 revenue projections are 25% to 30% below what it will achieve. For fiscal '98, the company is now expected to earn $0.50 per share, up from the consensus of 40 cents per share from a month ago. Earnings per share of $0.08 are now expected in the current quarter -- first-time-ever profitability -- on now anticipated revenue of $18 million.

With back-of-the-envelope possible revenue near $100 million in fiscal '98 -- as the $0.50 in earnings per share estimate intimates -- the company's stock is trading at 2.5 times that sales projection, and 41 times fiscal '98 estimates of $0.50.

This is decidely expensive for any chip-related technology producer, as the chips are seen as commodities... especially in the add-on arena, where one company's technology can suddenly leapfrog another's and even make the old obsolete. However, the company is starting from a small base, similar again to Creative Labs (Nasdaq:CREAF). As a true small cap, you can expect higher market valuation based on higher expected growth rates. Creative today trades at 8 times earnings, but it has revenues of $1.2 billion and insignificant top-line (sales) growth. 3Dfx can grow its numbers at near triple-digit rates for several more quarters. That's the reason for the valuation, and the risk inherent should the company stub its toe (or break its nose!). For 3Dfx to really become a viable market-beating investment, we'll want the company to equal or beat its 1998 estimates and then nearly double those numbers again in 1999. We think it can do it, but the risk is clear and obvious.

In short, we're expecting fireworks and we think Wall Street is too. Much like America Online in the early days (heck, even now), or Amazon.com, the stock is presently riding high on industry dominance and on the resulting future earnings growth that is hoped for. But hey, unless you have a well-balanced portfolio that has an open slot for a new small cap, you shouldn't consider investing in such volatile and risky propositions as 3Dfx. It could be a hairy ride (the ride's a lot more fun in 3D, of course).

FOOLISH EIGHT CHECKLIST
Running 3Dfx through the Foolish paces -- they serve to protect and guide, but not to be the overbearing law.

Derived from The Motley Fool Investment Guide, this eight point checklist that we run on small-caps is for guidance only. Few stocks meet all eight requirements, but the list is a fine way to get to know your companies better as you consider buying them -- to see how they stack up. Let's take a look at 3Dfx.

1. Company Sales: $200 million or less. With $21.7 million in trailing nine-month sales, 3Dfx is well below this limit. 1997 sales should total about $39 million. We like a small-cap with lower sales because sales can then -- at least theoretically -- improve more quickly percentage-wise.

2. Daily Dollar Volume: Around $3 million. 3Dfx has daily dollar trading volume of just above $3 million, which is perfect for our small cap stocks. It isn't so thinly traded that we could never sell the thing, and it also isn't extremely well known or volatile, or followed by a large handful of analysts. 3Dfx has four analysts giving estimates on the stock, in fact, which is a nice amount. That's enough to get some context and comparison, but not too many. If the business succeeds, many more analysts should come along later and many more firms will be buying, long after us.

3. Low Share Price: Between $5 and $20. We're not sticklers about this point, but again 3Dfx is favorable here, at just over $20. We only have this criteria listed because a less expensive stock can sometimes appreciate more quickly than a more expensive stock, if the business merits the rise.

4. Net profit margin: 10 percent or more. As with America Online, Iomega, and Amazon.com when we purchased the stocks, there is no net profit margin yet at 3Dfx. The company is expected to turn profitable this current quarter, with $0.08 per share expected on about $18 million in revenue. With 12.4 million shares outstanding, this means that $960,000 in profit is expected, which on $18 million in sales would mean profit margins of 5.3%. If the trend of improving gross margins continues, net profit margins should continue to improve as well -- also, they should continue to improve as business improves. Even so, we're taking a chance and breaking our guidelines here.

5. Relative Strength: 90 or higher. The relative strength is not very meaningful because the stock just came public in June, and relative strength is best measured over a five year or longer time period. That said, having first traded around $14 and now trading above $20, the relative strength is currently 88, meaning the stock has outpaced 88% of all stocks since June.

6. Annual Earnings and Sales Growth: 25 percent or greater. The company was a developmental firm until recently, so any sales now represent considerable sales growth. Quarter over quarter this year, the company grew sales 53% in the third quarter from the second, and is expected to grow sales another 80% this quarter from last. In 1998 earnings per share should increase by an impossible number to figure: from a negative number, to positive half a buck a share. First Call shows the increase as being 300%, going from negative $0.25 to positive $0.50. Um. Sure. We'll go with that. Whatever it is, it tops 25% by a long shot.

7. Insider Holdings: At least 15 percent. 18 percent. 'Nuff said.

8. Cash Flow from Operations: A positive number. 3Dfx managed to turn cash-flow positive last quarter, even in advance of profits. The company generated $300,000 positive cash flow from operations, reversing the negative $2.9 million of the first six months. This is the first time in the company's history, of course, as in 1996, 3Dfx had negative cash flow of $7 million on a few million in sales. With the initial development stage over and the company actually selling product, we're hoping for good things ahead in Cash Flow Land.

THE RISKS
Competition abounds in this cutthroat, largely commoditized and quickly evolving industry. Intel?

The company sells primarily to PC and graphics board manufacturers, and sales are highly concentrated to only a few companies. Orchid, Diamond Multimedia, and Williams account for 44%, 33% and 11% of sales, respectively, in 1996; and Diamond and Williams accounted for 59% and 15% of revenues during the first quarter of 1997.

Also, 3Dfx is fabless (fab being the jargon for "factory" in the world of semiconductors), and so the company relies on semiconductor fabs in Asia for production. 3Dfx is a technology developer, and doesn't have the infrastructure to produce its stuff (for better, and for worse). Revenue in the current quarter is going to be much higher than previous estimates due to increased yield (better output at these fabs). But as 3Dfx still can't meet demand (another worry), there are no guarantees that supply from the fabs in Asia will continue to be this high or will continue to rise in the near-term.

Also, being fabless, shareholders' equity isn't based so much on assets as it on employees and proprietary technology (which can be valued differently at the drop of a hat).

Just as Microsoft now competes with virtually all software and Internet media enterprises, so too does Intel remain the primary competitive threat to so much of the computer hardware industry. At present, Intel does not perceive the 3D market to be big enough to devote much resources to competing in this field. And 3Dfx has made its technology scaleable with the improvements in Intel processors, complementing Intel's efforts. Good move. Whether and if Intel winds up trying to compete here is a key question, and an ever-present risk. One thing's for sure, though: Intel stands to make much more money by focusing on ever bettering its own platform, the microprocessor, rather than worrying much about 3D add-on cards.

Finally, as with any company just moving into profitability, and any small cap investment, all the risks are too numerous to mention. If you don't know of the many risks involved in investing in small caps, you should please head over to the 13 Steps to Investing Foolishly, and stay away from such investments until you're well prepared -- Foolishly!

CONCLUSION
At the very least, this will be fun.
Volatile, and fun.

We hope by now we've made our reasoning clear. We like this company as it conforms to much of our classic Foolish small cap model, with some notable exceptions. On the grand scale of risk, this is a risky investment. That's OK with us, because at present the Fool Portfolio is overweighted in large caps and megacaps. We have an open slot for a dynamic mosquito grower that could sting or get stung.

This investment is suitable only for aggressive, Foolishly diversified, knowledgeable investors, preferably those who are interested in and closely follow the PC games market.

It is our eventual hope that 3Dfx can continue to change the rules, and wind up cracking the business graphics market. That represents a huge untapped source of business not included in any estimates or projections, but it's a long-term hope at this point.

To return to the beginning, as gamers ourselves we appreciate finding a company in this business sphere that we consider an attractive investment. We've followed the industry for years, but most entertainment software companies haven't had much resilience, and many have been downright awful. We'll be gaming and watching our way to profits, we hope. Heck, even if this investment doesn't work out we'll enjoy the lesson of it. This is a fun purchase.

Finally, we do expect the stock to open up tomorrow, as some people insist on following our portfolio. Nobody finds this more frustrating than we, as we derive no benefit from that, and always wind up having to pay more for our stocks because of the bandwagon effect. Hey, we understand this; we revolutionized the investment world with our own set of self-penalizing rules. How Foolish! It's called total accountability: we let you know BEFOREHAND what we're doing, in a Wall Street environment where the professionals are always putting out their reports AFTER they have acted. We are Fools -- you already know that. Anyway, we do urge you only to invest in this company if you're prepared to do so, and if you wait a few days, weeks, or months, even better... we won't have to pay up for this darn stock at market open!

Please join us for more discussion on the 3Dfx Interactive board, particularly the superb one at www.fool.com, which is where we do most of our posting these days. Also, a great FAQ is available on 3Dfx's Web page.

Fool on!