Why Vitesse leads the 4 horsemen of chip stocks?
In this stock picker's market, careful research is the key. In the communications chip sector, Applied Micro Circuits, Broadcom and PMC-Sierra all have long-term potential, but Vitesse looks like the one to buy now.
By Jim Jubak
1/31/2001
What do we mean when we say that some markets are "stock picker's" markets? And why, right now, does everybody again seem to be stressing the importance of careful research into all the stocks you own and all the stocks you're thinking of buying?
As any investor who has owned Wal-Mart Stores (WMT) instead of Kmart (KM) (or vice-versa) over the last decade knows, picking the right security from a group of similar stocks in a sector or an industry always counts. It's just that in some markets, stock-picking counts more than in others. And the current market is one of those where it counts a lot.
A transitional market -- like the current one that is making the chaotic turn from bear to bull -- exaggerates the differences between even very similar stocks in a sector or industry. And on the basis of those easily overlooked differences, it rewards some stocks extravagantly and punishes others severely.
Examples in chips
Let me show you what I mean by taking a look at the shares of the four horsemen of the communications chip revolution -- Applied Micro Circuits (AMCC), Broadcom (BRCM), PMC-Sierra (PMCS) and Vitesse Semiconductor (VTSS).
In strong bull markets -- like that of 1999, for example -- the key to making money was being in the right sector -- technology -- and in the right industries within that sector. During that year, you'd certainly have preferred to pick Applied Micro Circuits over the rest of these stocks, but even the worst performer -- Vitesse -- crushed the 86% return posted by the Nasdaq Composite index ($COMPX). Applied Micro gained 649% in 1999, Broadcom 351%, PMC-Sierra 408%, and Vitesse 130%.
In a strongly down market, like that which ruled the Nasdaq from March 9, 2000 to the end of December, the group also moved roughly in concert. Applied Micro was up 2% during the period, Broadcom fell 65%, PMC-Sierra declined 68%, and Vitesse tumbled 35%, compared to a 39% drop in the Nasdaq. Again, a stock picker who backed Applied Micro Circuits reaped significantly better relative returns than one who chose any other stock in the group.
But so far this year, returns for the group have scattered. Broadcom and Vitesse are up 31% each as I write this, while Applied Micro is up just 4% and PMC-Sierra is down 6%. Score that two stocks out of four that have raced ahead of the Nasdaq's 15% year-to-date gain, one that has seriously lagged and one that has declined in absolute terms.
In my opinion, those numbers indicate that when the market looks at these four stocks, it is now seeing differences that are large enough not just to determine relative over- or underperformance among the four, but that are of a size to send the performance of some members of the group in opposite directions. If that's true, understanding what these differences are could, during the current transitional phase of the market, determine which stocks will show a profit for an investor and which show a loss. And if we later see a return to even a modest bull market for technology stocks, the differences are likely to determine the relative over- or underperformance of the individual stocks in the group.
Two key factors
So what are the differences that the market is emphasizing right now? There are certainly a lot to pick from. But I'd cite two that I think far outweigh the others: the positioning of the company's product mix either at the edge or at the core of the network, and the company's relative position in the current race to own the fastest chip.
Let me give you my own quick explanation of those two factors, and then I'll try to handicap the four communications chip makers accordingly.
First, the edge of the network versus the core. In general, it's still true that communications-chip company product lines tend to be dominated either by edge or core products. To simplify drastically, edge products connect users to the communications network. Examples are set-top cable TV boxes, modems for PCs and cable TVs, access concentrators that bundle signals from end-users, Ethernet cards and networks in business or homes and the equipment that connects those networks to the core network, and switches and routers that direct traffic in workplace or home networks. Core products are the building blocks of the central network that connects all these edge networks. Examples are the switches in telecommunications networks, the network processors that help direct traffic on the Internet and the optical components that guide data along fiber-optic networks.
The reason this edge/core distinction is important right now is that revenues from core products are, in general, growing faster than those from edge products. For example, Vitesse's core OC-192 optical revenue grew by 23% in the December quarter over the September period. That's in marked contrast to the 3% quarter-to-quarter growth that Broadcom saw in its edge set-top and cable modem business with Motorola (MOT).
Second, speed sells -- and it sells for more. It's no secret that speed adds to the value of a semiconductor. That's certainly true for PC chips. When Intel introduced its new 1.5-gigahertz Pentium 4 in late November, it priced the chip at $819. Before recent price cuts, the 1-gigahertz Pentium 3 chip -- theoretically two-thirds as fast -- was selling for $465, or only a bit more than half as much.
The same is even truer in communications networks, where chips that can push data through a network at higher and higher speeds can increase the value of the entire network, not just a single PC. And chips that can switch and route data at ever higher speeds and over more and more complex networks are more valuable yet.
Not only do these higher-speed chips earn higher margins, but in the current market, they're selling better than chips that run at slower speeds. You can see the revenue differences clearly at Vitesse. In December, the company's OC-48 optical revenue grew by 18% over the September quarter. Revenue for the faster OC-192 optical products grew by 23% in the same period. And a big reason for the recent cuts in revenue projections at PMC-Sierra is the company's lack of an OC-192 product. Until PMC-Sierra can get the faster gear to market -- sometime in the first half of 2001 -- it is stuck selling OC-12 and slower chips, and the slower equipment is at the heart of the current slowdown in revenue growth.
How the quartet looks
Ok, so how do the four companies shape up?
Applied Micro Circuits. In my opinion, Applied Micro Circuits offers the best combination of concentration in the core and high-speed products of any of the four communication chip makers. According to Robertson Stephens, about two-thirds of Applied Micro Circuits application-specific chip business is concentrated in the optical core of the network -- and that's the fastest growing part of the network. Applied Micro Circuits offers both OC-48 and OC-192 products, and in the December quarter, the company's OC-192 revenue grew by 79% from the September quarter. That's even faster than the impressive 23% sequential growth recorded by Vitesse in this product line. The combination of concentration in the core and on high speeds gave Applied Micro Circuits the highest gross profit margin, at 76%, of any of these four chip makers in the fourth quarter of 2000. The huge increase in profitability from 70% in September 1999, to 76% in December 2000, was a major factor in powering the stock's extraordinary gains in 1999 and in buffering shares from last year's technology bear market. The only question remaining for Applied Micro Circuits is what the stock does for an encore. It is currently the most expensive in the group, trading at 111 times estimated earnings for the next 12 months. Earnings growth for that period is an estimated 62% to give me a price-to-earnings-to-growth ratio (PEG) of 1.79.
Vitesse Semiconductor. Vitesse has lagged Applied Micro Circuits in concentrating its business on the core of the network -- it has a sizeable edge business -- and in penetrating the OC-192 market. But it may, therefore, have more room for improvement than Applied Micro Circuits. For example, the company is growing its revenue from higher-level, more-intelligent chips and from faster switches at rates that are quickly shifting its product mix. In the December quarter, these two areas made up just 13% of the company's total revenue. But revenue from higher-level processors grew by 326% over the September quarter, and revenue from faster switches grew by 100%. Vitesse has scored a total of 95 design wins over the last 12 months in these two areas. While analysts are projecting that Applied Micro Circuits' gross profit margin will remain roughly steady at 75% over the next year, they're projecting that Vitesse will grow gross margins to 70% from the current 68%. If the company can deliver on current projections, the stock is the cheapest of the group, trading at just 57 times projected earnings per share for the next 12 months. Earnings growth for the period is an estimated 68%, to give me a PEG ratio of .84.
PMC-Sierra. This company is definitely facing tough times in the quarters ahead. Its products are concentrated at the slower OC-8 and OC-12 speeds -- or less -- and it won't have an OC-192 product on the market to match Vitesse or Applied Micro Circuits until later this year. Its higher-level chips and faster switches, the drivers of future growth at Vitesse, are so new that they can't make up for declining revenues from older products. The PMC-Sierra story is a classic tale of product transition -- the goodies are on the way, but until they arrive, much that it has to sell seems slightly stale. The company has led analysts to expect a roughly 30% decline in revenue in the first quarter of 2001 from the fourth quarter of 2000. And analysts are now also projecting that revenue for the second and third quarters of 2001 will fall below the December quarter's $232 million. Year-to-year revenue growth will be a paltry 12%, according to analyst estimates, while the stock is now trading at 96 times projected 12-month earnings. I think that's a high enough multiple to limit the stock's bounce off the recent sell-off until investors know whether the company will be able to meet even its drastically lowered forecast for the first quarter. The company will pull off the product transition that's now under way, but I think this stock is essentially dead money until investors start to see an increase in orders -- instead of cancellations. I'm going to hold onto these shares in Jubak's Picks for a few days to get the benefit of any post-Federal Reserve interest rate cut rally. But I'll be looking to sell them before Cisco Systems (CSCO) -- a major PMC-Sierra customer that might announce bad news -- reports on Feb. 6.
Broadcom. Broadcom is the real wild card in this group. It is the most edge-oriented of these companies and therefore has the most exposure to slowing growth in these markets. The company has seen orders cancelled in December and January from Cisco Systems, Lucent Technologies (LU), Alcatel (ALA), Juniper Networks (JNPR) and Nortel Networks (NT). The company only made its December 2000 quarter thanks to a huge 100% quarter-to-quarter surge in revenue from 3Com (COMS) that made up for huge declines at Motorola and Cisco. That kind of pick-me-up from 3Com is unlikely in the next quarter. So, will Broadcom blow up next quarter? Will the company manage to meet current projections? I think it's simply impossible to tell. But given this uncertainty, the stock feels expensive to me at its current P/E of 74 times projected 12-month earnings. With estimated earnings growth at 53% for the next 12 months, the stock has a PEG ratio of 1.4. I think we'll know a lot more about Broadcom after Cisco reports; I think it's worth waiting for a better idea of how bad the growth news is at one of Broadcom's key customers.
Over a longer time period, all four of these stocks are likely to be growth machines. If I had a limited number of investing dollars, I might prioritize them this way: Vitesse first, on the strength of the pickup in its current business; Broadcom next, after the Cisco shoe has dropped (and maybe not until after the company reports in April); PMC-Sierra late in 2001 when the product transition is closer to completion; and Applied Micro Circuits whenever events create a temporary dip in this very strong but very expensive stock. Or when you've run out of other great technology stocks to buy at more attractive prices. |
