The Futile Exercise of Trying to Put a Rational Value on Rambus thestreet.com
Tuesday Thud:
Just for fun: Yesterday Rambus (RMBS:Nasdaq) closed in on the 12-month targets of some analysts when it topped 100. No big deal in a market where many, if not most, Internet valuations are considered way out of whack. Or in a market that has helped inflate Micron Technology (MU:NYSE), despite every suggestion that the darling of DRAM's valuation at almost any point in recent years was more than out of whack. (Yep, the same Micron that has made yours truly look like a fool more times than I care to remember.)
So where does Rambus stand in the world of the out-of-whacked? One argument, posed by a longtime anti-Rambus rabble-rouser, goes like this: If Rambus gets 100% of the memory chip market over in the next year or two, and gets a royalty of 2% on each chip sold -- the high end of expectations -- you'd be looking at 2% (again, on the high end) of an estimated $10 billion market. That would be $200 million, or roughly one-tenth its current valuation.
Even if you applied Intel's (INTC:Nasdaq) current multiple of seven-times-sales you'd still come up short, and that's before deducting expenses (albeit low) and taxes (like any other company). "You can't justify that unless you think the whole DRAM market will explode, like some kind of nuclear strike in Korea," our skeptic says. "How is 1% or 2% worth 2 to 3 billion dollars" in market valuation?
Simple: market sentiment and momentum. Just as they drove Micron to unrealistic and unsustainable levels, so will they drive Rambus. "It's not like people are making rational judgements of Internet stocks, either," says SG Cowen's outspoken Drew Peck, who is just as likely to be negative on a company as he is positive. (In this case, he has a buy rating on the company; Cowen did not underwrite Rambus' 1997 public offering.)
In a nutshell, Peck sees it like this: The DRAM market is currently at the low point of its pricing cycle, and annualized, the DRAM market is twice the size of the number used by our skeptic (who based his estimate on Micron's sales). Peck also is looking for a DRAM shortage starting late next year, and he believes Rambus will be in the right place at the right time with the right (256-meg) chip. He also says critics don't take into account the fact that Rambus' logic chips that are bought by video game makers, which he figures add another 50% onto the projected sales figures.
He also believes Rambus could earn as much as $7 to $8 per share in 2001 or 2002, compared with 28 cents last year. He says his analysis assumes that Intel provides "unwavering" support for Rambus. Intel's chips, using Rambus technology, are expected to ship in September. (See my colleague Marcy Burstiner's excellent coverage of Rambus for more detail.) If Intel backs away, he adds, "all bets are off."
In the meantime, our Rambus rabble-rouser argues, so is rational analysis. |