What Now for Rambus?

With its technology required for the first version of Intel's Pentium 4 and designed into several hundred products in under two years, Rambus is a success -- but the stock's off a cliff. Lawsuits and competition haven't helped, but the real culprit is crashing computer memory selling prices. Rambus is lower risk today, but with a lower ceiling too.

Format for Printing

Format for printing

Request Reprints


By Tom Jacobs (TMF Tom9)
February 11, 2002

Remember Rambus (Nasdaq: RMBS), the stock that zoomed from $16.86 on New Year's Eve 1999 to $111.27 in two and a half months? The company burst with promise, proud that Intel (Nasdaq: INTC) chose only its RDRAM (Rambus dynamic random access memory) computer memory technology to work with the first Pentium 4. Memory makers signed up to pay royalties, and companies began to design Rambus technology into hundreds of products. 

How bright the future looked, but how quickly 2001 brought shades. After a long decline, last Friday the stock closed at $6.47, below its 1997 IPO first-day split-adjusted close of $7.56. Ouch.

Neither competition nor lawsuits     
Critics incorrectly cite the cost of RDRAM and the company's numerous court cases for Rambus's descent. True, Intel followed the first Pentium 4 with versions compatible with old-fashioned SDRAM (synchronous DRAM) and then RDRAM's competitor DDR DRAM (double data rate DRAM), but today RDRAM is cost-comparable with DDR and more techie sites rate it better performing.

And yes, litigious Rambus believed that the old standards SDRAM and the emerging competitor DDR infringed on its patents, and it's embroiled in lawsuits in the U.S. and Europe with second, third, and fourth largest memory makers Micron Technology (NYSE: MU), almost-dead Hynix Semiconductor, and Infineon (NYSE: IFX). I'll grant it didn't help when a U.S. court slammed Rambus last year, either.

But something else hurt the company, something it could not control. It depends heavily on a few memory maker customers because there are only a few, and that semiconductor industry is notoriously cyclical. Just as things looked dandy, 2001 introduced oversupply to anemic demand. The bottom fell out of the DRAM market, and average selling prices for computer memory plummeted 90% and more. Estimates of $38 billion for DRAM sales in 2001 and up to $76 billion in five more years proved laughable. The year dragged on and that $38 billion slid to the twenties, and finally to the low teens.

Suddenly I -- the guy who swore to a colleague that you'll "never see shares below $40 again" -- learned that telling fortunes was not my forte. A shareholder, I had ridden Rambus all the way up and rode it all the way down. With not nearly as much fun as Space Mountain. 

Down but not out
Yet Rambus's business did not feel the same pain of the memory makers or computer hardware or networking businesses in 2001, because RDRAM volumes increased even though prices crashed. The company stayed profitable and cash flow positive, though at lower levels:

       Revs.     Cash Flow      EPS
Q1 02 $ 24.9 mil. $ 5.4 mil. $ 0.06 
TTM    107.2       13.7        0.23 
2001   117.2       17.0        0.30
2000    72.3       29.4       (1.55)
1999    43.4        0.8        0.09
1998    37.9       14.1        0.07
1997    26.0       28.0        0.02

Feb. 8 closing price: $6.47
Price/Earnings                   28
Enterprise Value/Free Cash Flow  41
  (w/o $7 mil. in legal fees)    28 

The worst is likely over. Year 2001 cash flow slumped due to increased legal fees, with less expensive European cases and U.S. appeals slated for this year. DRAM prices appear to have bottomed out and are on a slow, welcome rise, with RDRAM no longer more expensive than competitor DDR. Not only that, but Intel has signed a new blanket royalty agreement paying Rambus a fixed $40 million a year for five years, the company secured its first licensee for its RaSer technology in communications and networking, and Sony's (NYSE: SNE) Rambus-ized PlayStation2 video game console sold strongly.    

But just as the stock presents remarkably low downside risk at current prices, its upside is lower too. Through no fault of its own, the cyclical crash of DRAM prices means that Rambus has royalty cash it can never recover. The business is simply worth less than when many of us were dreaming of retirement by this high-margin cash green river. To remember that time, enjoy our August 2000 Duel on Rambus.

Profits foregone
From the start, Rambus's technology and patents gave it a competitive advantage for an undefined period -- however long before another company speeds up from behind with the next advance. Rambus and any other company toiling in the fast-moving world of computer chip technology has only so long to exploit a lead, and through no fault of its own, it just lost a huge amount of that sole determinant of its future. It will have less cash sooner to fuel R&D or buy competing technology -- to do anything to extend its growth and its competitive advantage period. Absent that cash, its prospects are limited no matter how good its inventions or its intentions.

Think of it this way. Let's say your product is gaining market share in a $100 billion market, and you have a five-year lead on the competition. The market collapses to a quarter, or $25 billion, but you increase your lead. The market may slowly inch back, but all things equal, you'll never recover your share of that lost $75 billion, and the clock is ticking on your ability to fund new product development or buy someone else's.

Less attractive, but no slouch
Rambus still flaunts an attractive risk-reward profile for the five years of the Intel agreement. It has a debt-free balance sheet with $110 million in cash and short-term investments. Its high-margin royalties will grow rapidly with the slightest uptick in DRAM prices and with even modest penetration into non-personal computer memory and communications networking markets.

Those assumptions lead reasonably to free cash flow of $60 million or so on revenues of over $200 million in three to five years, which at an enterprise value to free cash flow multiple of 25 would return a little less than 2.5 times from current prices.

As for dilution, it's currently negative because Rambus is smartly buying back undervalued shares, but even a couple percent annually wouldn't tarnish those returns unduly. A potential 250% gain in five years with Rambus's more limited downside today would normally be sweet from any stock, but it's far from the retire-early dream some of us believed in 1999 and early 2000. What a difference two years can make.

Clean Rambus
In this era of Enronitis, Rambus does offer a few other benefits. It's got a clear business model. It works with companies to design-in its technology and earns royalties when those companies purchase the tools from Rambus's licensees. You can measure success clearly in financial reports that are excellent and easy to follow, with no off-balance sheet partnerships (that I could find -- and you should triple-check what anyone writes when you do your own research) and straightforward explanations of all business operations (even the risks are drawn in far better detail than the boilerplate stuff you almost always see).

Rambus recognizes revenue conservatively and late. It's even easy to follow the impact of stock option grants -- though I'm really just learning about this accounting quagmire. Finally, considering that lawyers -- I used to be one, so I can take this swipe --  probably wrote the sections on lawsuits, they are remarkably clear and a must-read for any potential or current shareholder.

While there are never any guarantees, Rambus appears to be one company whose business, not its books, carry the risks.            

Fun with Rambus
I just may be crazy, though, and you can say so on our Rambus discussion board, which is consistently one of Fooldom's busiest and easily one of the most exciting. Join industry insiders and you-and-me-type investors to debate the company and its competitors. You can speak up or just lurk, but it's free only through Wednesday night, so don't dally and please become a Charter Member of our Community today to get a bonus free year. Think of it as a Valentine Day's present for yourself, and as a gift for a Fool you love.

Have a most Foolish week!

Tom Jacobs's (TMF Tom9) favorite chip is chocolate. He owns shares of Rambus, which you would see on his profile, rigorously prepared in compliance with The Motley's Fool disclosure policy. (Suck up.)