Are financial analysts AMD believers? Well, maybe.
Financial analysts supportive but skeptical; "They have to sell them as well as make them". They worry that AMD may have missed the window of opportunity as Intel girds to compete with it.
AMD is definitely on the hot seat. The company is said to have its best--and maybe only--shot at the Intel microprocessor juggernaut. It has the subtle support of the industry, uncomfortable with an Intel monopoly. The company claims that 10 of the top 20 PC makers will use the K6--if they can get it. The K6 is aimed squarely at the lower end of the PC market, just where Intel was caught unawares. But AMD has been unable to make enough K6s to meet demand, and to make a profit. It planned to make 15 million K6's in 1997; plagued by poor yields, it actually made only 1.5 million good MPUs.
But as Intel races to catch up in the low end MPU market, AMD must make amends to customers bypassed during the period of unexpectedly limited K6 production. At the time, it stressed K6 commitments to giants Compaq and IBM. Some lesser-sized companies may have been ruffled, admitted CEO Jerry Sanders to analysts at a phone conference last week. But he claimed the situation "is not irreparable."
At Bear Stearns, Nimal Vallipuram said: "They have given a 12 million-unit target for this year. I have calculated it at 11 million. They clearly seemed to have improved their yields but the question is rate of the ramp. The ramp seems to be pretty aggressive. Their past problems are a concern."
At C.S. First Boston, Jack Geraghty said: "The earnings situation was worse than most people, including myself, thought it would be, but it was due more to other segments of AMD's business; for instance, networking, memory and programmable logic devices. The MPU business did what AMD said it would do."
He added, "I didn't have an aggressive forecast this year for the K6 ramp so I am not necessarily disappointed about that. AMD is doing a little better than I would have thought at this stage, but nothing dramatic. AMD's past history on production isn't as good as they would have liked. I have a slightly lower forecast for the year, a little over 10 million K6s. The only variable is the other businesses; but that reflects the overall semiconductor industry."
Mr. Geraghty predicts a loss this year of about 65 cents per share, and a gain of $1.31 per share for 1999. He has a "neutral" on the stock.
At Josephthal & Co., Larry Borgman said: "I believe AMD's statements in the sense that it is what they want to do; whether they accomplish it is another story. It's a pretty steep ramp in the second half of the year given that they are only talking 2 million (units) in the second quarter. I believe their real goal in the quarter is 2.5 million units, although they set a 2 million floor."
The company has met the 2.5 million-unit goal before on the 486. "They have the running capacity to do it. They have to add some more equipment, but they have a pretty strong cap spending budget this year."
He noted: "I would be a little bit surprised if they actually make it. And they have to sell them in addition to making them. Unfortunately, they have always had a moving target. The market always changes, so that what you want to make today, and sell today, is one thing, but then Intel all of a sudden moves the bar and you discover that in order to sell it, you have to produce 330MHz instead of 300MHz or 266MHz. Then you have to up all of your parameters, and the question is, can you make it under the more stringent conditions."
Mr Borgman has doubts about an AMD-IBM financial relationship."I was expecting exactly what AMD said. People were making way too much out of it," he said. "IBM was not interested in investing in a semiconductor company like AMD. It made no sense for them, I believe. IBM has a pretty good foundry operation and they want to make some money at it."
Mr. Borgman has AMD shares on "hold" as well. "There was a pretty nice move already based on their getting the yields up and IBM speculation. We have them at breakeven for the year, but there are a lot of assumptions in there. They will have to go a long way toward achieving the goals they talked about." He has a "wild guess" figure of 75 cents per share profit in 1998.
"There are so many variables with this company," he said. "AMD is pursuing a very high risk strategy. Focusing on the PC and a way to second source Intel's product. Intel gets a good price. The question is whether AMD can slip in under their umbrella and sell for less than them. It isn't as easy as it looks. When you are pricing chips in the hundreds of dollars, it looks very attractive."
Mr. Borgman is unhappy about AMD's strategy. "Jerry Sanders has bet the company," he says. I wasn't happy when he bought NexGen. I thought that with AMD's base of business in the other areas he could have pursued a much less risky approach. They have some talent at AMD, and good production facilities. They had to emphasize other areas. It is tough going up against Intel in that market. If you look at AMD's history, there are only a few times that they have made much money and only for short periods. Mr. Sanders could have pursued a less risky approach with the company and he probably would have had a smaller return but a profitable company."
At Piper Jaffray, Ashok Kumar quipped: "AMD is ready for a take-off but the question is, does it have a flight plan. In the past, they clearly had more difficulties than they would have liked. It's a question of guilty until proven innocent."
On K6 yields, Mr. Kumar said AMD has the manufacturing ability and should be able to ramp up to 12 million units, with 2.5 million, ramping to 4.5 million in the fourth. But he warned, "This assumes Intel is going to roll over and let AMD add another incremental million units a quarter for each of the next three quarters."
He argues, "Intel is like a shark. You have the shoals of fish that feed off the shark, which really needs those fish to keep it healthy and keep the SEC off its back. But then it has a threshhold of pain. When it sees the symbiotic relationship has changed, the story is over. Maybe 10-12 million is a theoretical limit that AMD can address. But they go from a supply constrained situation to a demand constrained situation by as early as the third quarter. From a PC OEM perspective, nobody in the history of technology has made money off lagging-edge products. Perception is everything. Pentium II is considered leading-edge; Pentium is considered lagging-edge, thanks to Intel marketing. AMD rolled out a 300MHz K6. It was priced at $1,499. For the same price point, you can get a Pentium II 300.
"In terms of valuation, I can see them realistically doing no more than $1 (per share) in earnings for next year with the stock trading where it is at 27 or 28. It is trading on a hefty multiple to the market. Even Intel isn't trading at that multiple. Mr. Kumar predicts 4 cents per share profit in 1998.
The analyst has another take on the relationship with IBM. "If I read between the lines, AMD doesn't need additional K6 capacity at all. The earliest that IBM can contribute K6 units is in 1Q99, which is too late for the K6. By then the market is top to bottom Pentium II. AMD is trying to have IBM as backup for the K7 because if AMD can use IBM as a manufacturing partner, they can jump through a legal loophole and be able to graft a Pentium II interface into the K7. IBM will be crucial in helping AMD develop the 0.18-micron technology."
The K7 is key, agreed David Wu of ABN AMRO. "AMD will be profitable in the second half. The key after that is when the K7 will be available. The sooner the better. I think they need the K7 in 1Q99 so that if there is any problem with the market moving to slot 1 from socket 7, AMD will be viable. Otherwise, they will have problems in the rest of 1999."
Meanwhile, Standard & Poor's last week lowered its bank loan, senior secured debt, and corporate credit ratings on AMD to single-'B' from double-'B'-minus. S&P's also lowered its ratings on the company's $1 billion universal shelf registration to preliminary single-'B'/triple-'C'-plus from double-'B'- minus/single-'B'. The outlook is negative. "Ratings reflect the company's continuing shortfalls in executing its business plan in a very competitive market, resulting in ongoing losses and substantially negative free cash flows," the agency said.
Intel Could Kill the Recovery
AMD's manufacturing problems have been widely aired. Inability to produce the K-6 in volumes required have held back its financial performance. AMD has recently achieved a good market share in the thinly profitable low-priced PC market segment.
"But Intel's first entry-level product is expected to be available soon, exacerbating already aggressive price declines. In addition, AMD will have to upgrade its manufacturing processes again to make the successor K7 next year. Intel remains the biggest threat because of its rapidly proliferating product line, aggressive pricing policies, and substantially greater balance sheet strength, in addition to R&D, marketing, and manufacturing prowess."
Standard & Poor's worries about AMD's financial strength. The agency reported, "continued net losses remain likely over the intermediate term. Capital expenditures have been high, averaging 28 percent of sales." |
