Betty,
Thank you for your kind post.
In my post last Thursday night, I said that some shorts have argued that RMBS's stock is trading at 100 times 1999 EPS estimates and, that on this basis, is grossly overvalued. I said that it all depends on what the earnings are after that period. So, I decided it would be helpful to summarize what I could find concerning the public projections of RMBS royalties that have been made by consulting and brokerage firms.
What makes projecting RMBS earnings so difficult is that it is unclear how large the future DRAM market will be and what level of penetration Rambus technology will achieve. Compounding this uncertainty is the fact that logic circuit companies will also pay royalties and have a different level of market penetration for their chips. What is certain, however, is that the sooner RMBS chips are in full scale mass production by a number of major manufacturers, the more likely that competing standards like SLDRAM are dead. So we need a lot more production announcements like the one from LG Semicon Co. on November 12, 1997.
Now, let's go to the projections. The first one is a "sort of" projection. It was in an interview of Gary Harmon, RMBS's CFO, by Darren Chervitz of "DBC IPO First Words". It was dated June 18, 1997 and says the following:
"DBC: Sales of Rambus-enabled chips to $477 million in 1996 from $13 million the year before. A Robertson Stephens analyst said that the DRAM market will be $50 billion by the year 2000; your own prospectus said $62 billion. The total market you'll compete in, including logic chips, will be much higher than that. What kind of percentage of the market do you see Rambus generating by the year 2000?
Harmon: I wouldn't be able to project a number because we try to stay away from that. But I think that the reason Rambus has a high total available market, is if you just took that $50 billion number, for example, and said that if we only get 1 percent royalty on DRAMS, that would be $500 million total available market to Rambus. And then we get a royalty from the other side - the logic chips - which is probably comparable. So it says that basically that at that $50 billion DRAM market estimate there is probably a total available market for Rambus of something over $1 billion, even with a very low royalty rate."
What I find important about Harmon's answer is the sentence about "the logic chips - which is probably comparable (in royalties to DRAM)." That's because some of the projections out there only seem to focus on DRAM. And Harmon seems to be saying that DRAM is only about half of the future royalty stream.
To protect Rambus against shareholder lawsuits, Harmon said "I wouldn't be able to project a number." I think that management felt a lot less constrained about making projections prior to being a public company. And I think that shows up in an article in the February 1997 issue of OEM Magazine (see techweb.cmp.com. In that article, the author, David Lammers, provided management's estimates of their penetration in the Intel based desktops shipped in 1999, 2000 and 2001. It says ' Rambus expects about 15 percent of the Intel-based desktops shipped in the last year of the decade to use its RDRAM technology, rising to 35 percent in 2000 and about 65 percent in 2001'.
On the next to last page, the article also develops a projection of Rambus DRAM royalty. It says "But Rambus doesn't stand to remain small for long. A back-of-the-envelope calculation suggests that if the DRAM industry rebounds to $50 billion by 2001, and 60 percent of DRAMs go to Intel-based systems, and 65 percent of those DRAMs are Rambus types for which Rambus collects 1.5 percent, then Rambus earns about $300 million. In one year. With no factories to build. And that's not counting royalties from licensees such as Intel, surviving chip-set makers, 3-D graphics-card vendors and non-PC vendors such as Nintendo and SGI."
Now maybe I'm wrong, but I believe that David Lammers got a lot of pre-IPO guidance from management and that they internally believe that royalties of $300 million from DRAM are possible in 2001 as well as $300 million from logic ICs for a total of $600 million. As an aside, this article is one of the best I've seen on RMBS. It's a must read.
The next projection I found has been mentioned in previous posts on this thread. It's the special technology issue of Worth Magazine dated September 1997. (see worth.com
That article recommended the stock at then lower prices. It says "Intel will begin using Rambus technology in 1999. This means any manufacturer that wants to make DRAM chips for Intel computers has to pay royalties to Rambus--roughly 2 percent of the value of the chips. DRAM sales are expected to reach $38 billion in 1999, and Rambus-based Intel DRAMs should account for about a third of that. Rambus could be looking at royalty revenues alone of $150 million by 1999 and $200 million by the year 2000." This projection is similar to the one in the February 8, 1997 issue of the magazine "The Economist." It says "Within a few years, Rambus could be collecting $100-$150 million a year in royalties just for coming up with the right DRAM architecture at the right time."
The next projection was in the June 9, 1997 research report on Rambus by Hambrecht & Quist. It has a section titled "High Growth and Large Market Potential." It says "Rambus-based ICs have already shown exceptional growth from $11 million in 1995 to over $444 million in 1996 led by the enormous success of the Nintendo 64 video game machine. Market researchers estimate that approximately 12 million Nintendo 64 machines will be shipped in 1997. However, we believe the long-term market opportunity for RDRAM growth is the PC main memory market.
Dataquest estimates that the DRAM market can grow from $26 billion in 1996 to over $62 billion in 2000. In addition, if you consider the logic/ASIC market, the total available market for Rambus-based ICs in the year 2000 is approximately $100 billion. PC main memory accounts for approximately 50% of the total DRAM market, which translates to an estimated $30 billion PC DRAM market in 2000. If RDRAMs become a de facto standard for main memory, and Rambus were to receive a conservative 1% average royalty, that would translate to a potential royalty revenue of $300 million in 2000 for Rambus. If one adds royalties from logic/ASIC, that revenue could double. Of course,
it would be very surprising if RDRAMs achieved 100% market share in the year 2000, but nevertheless, we believe Rambus has the potential to show enormous growth from the $14 million the company generated in calendar year 1996. Exhibit 22 illustrates the potential revenue stream from RDRAM royalties by market penetration of the DRAM market.
Exhibit 22
Rambus Revenue by DRAM Market Penetration in 2000
($ in million)
----Market-----PC DRAM(a)-----Total------RDRAM
Penetration----Market 2000----RDRAM(b)---Royalty
----------------------------Market 2000--Revenue
--------1%--------32,000-------320--------3.2
--------5%--------32,000-----1,600---------16
-------10%--------32,000-----3,200---------32
-------20%--------32,000-----6,400---------64
-------40%--------32,000----12,800--------128
-------60%--------32,000----19,200--------192
-------80%--------32,000----25,600--------256
------100%--------32,000----32,000--------320
(a) Assumes 50% of total DRAM market in 2000 will be used in PCs
(b) Assumes average royalty rate of 1%
Source: H&Q estimates"
Now if you are still awake, the last projection I found was in an article by Steven Przybylski titled "Intel's RDRAM Strategy a Sure Winner-CPU Maker Gains High-Bandwidth DRAM -One Way or Another". This article was part of the April 21, 1997 Microprocessor Report by MicroDesign Resources. Toward the end of the article it says "If hypothetically, by 2002 or 2003 Direct RDRAMs account for half of the PC industry's 70% share of an $80 billion DRAM marketplace, at an aggregate DRAM-plus-controller royalty rate of 1.5%, Rambus would gross $420 million in revenue with next to no cost of goods."
Now, this is from MicroDesign Resource's Microprocessor Report. If you haven't heard of this firm or their report, check out page 118 of the November 1997 Issue of Worth Magazine (see worth.com
The article says that this report is the one most read by the financial community. It suggests that its editor, Linley Gwennap, has the best understanding of Intel and the microprocessor industry of anybody outside of Intel.
As an aside, you might ask "what startup companies does Gwennap recommend for its future potential?" That's easy. In the most recent issue of Upside Magazine in the column titled "Sizzling Startups", Gwennap nominates Rambus (see upside.com)
Now, is Rambus overvalued? It depends on which of the earnings projections you use and which multiple you use to capitalize these earnings. After compiling all this stuff, I am partial to the OEM magazine that says Rambus expects to have 65% of the DRAM market in 2001 and $300 million from the DRAM side of the business. That means they might do $600 million in royalties that year if you count the IC logic royalties. Now, if you assume that the royalties go straight to operating profit, and put a 15 multiple on operating profit, you get a $9 Billion market capitalization. That's about $378 per share in four years.
In last Thursday's post, I mentioned that Cisco had returned a compounded 79% between 12/31/90 to yesterday's close. If Rambus gets to $378 per share on November 19, 2001, it will have compounded at 73% over the period. So if this happens, yes the stock is overvalued compared with Cisco's experience but otherwise no. But it all has to work out. We shall see.
Regards
David |
