To: SlowThinker who wrote (42663 ) 2/2/1999 11:56:00 PM From: Earlie Read Replies (2) | Respond to of 53903
S.T. Good to see you state your assumptions. Many do not do this. A few comments on your assumptions if I might. -PC revenues will be negative year-over-year for 1998. This is obvious even if one accepts IDC's 92 million unit estimate (they are always optimistic), given the falling ASPs. The IDC figure includes a big channel stuff at the end of the year, and a rapidly decelerating corporate sales environment. If I recall accurately, the lowest year-over-year revenue growth in the past was approximately 10% (positive) back at the turn of the decade. It is difficult to suggest a stronger year in global PC sales in 1999, given the multi-year declining trend, and the loss of many countries as viable PC markets. The most likely scenario is for further depressed PC revenues, especially if the global deflation continues to spread, or North American earnings continue to fall. Both appear more than likely - MU sells 72% of its product to the PC environment. This percentage won't vary too much for a year or two yet. - PC sales up 15%? Perhaps the referenced quote referred only to consumer unit sales (not revenues, and certainly not corporate sales). Also, there is a big difference between shipments and sell through. The channel experienced degraded shipments last year at this time as a result of channel stuffing over the preceding holiday period. This year looks worse. - PC ASPs continue to fall. This will intensify the price pressures on memory. - PC base memory requirements are largely driven by applications. No new applications suggests that the current efforts by some suppliers (especially the local cloners) to stuff the memory slots may not be a long term trend. - Behind the scenes in Korea one finds decisions at the very top levels that mandate increased chip shipments as a means to assist Korea to export its way back to some form of stability. There is frenetic and intense yield improvement work in process. Excess capacity is rampant across all of Asia. The talk of 15% increases is but short term opportunism with respect to the Korean union/amalgamation problem. - I agree that digital cameras will help, but not in the near term. This Christmas, video cameras (which do a great job of creating stills as well as video) sold well while few stores were as yet subject to strong digital camera demand. HWP sources are a bit glum about the slowness of the public's acceptance of the digital cameras.. - Fab cancellations relate to massive over-capacity and continuing intense yield improvement. Fab closings relate to plants that are in need of upgrades. Unfortunately, poor market demand and pricing provide no justification for the required big expenditures, so they are shut down. We still have far too much global memory manufacturing capacity. This fact is common knowledge to all involved players. - Intel's "investments" in memory producers have much to do with that company's attempt to head off competing memory technologies and with attempting to ensure viable "second sources". Intel still dreams of the old monopoly days, but Rambus isn't going to do it for them according to many players in the industry. Rambus is also a bit late, so dollar resources are being dumped on the problem. Micron has plenty of problems. -Too much debt. Check out the interest bill that must be paid. The company won't be able to so do from cash flow unless dram prices rise a great deal. - Cash is committed. Most if not all will go to upgrading the TXN plants (check out the 10-Q). - MU has a "must-buy" contractual arrangement with the TXN plants. This is a continuing disaster, given the well-behind-the-current technology employed in the TXN plants. This engenders an extremely ugly choice,...spend all the remaining dough to upgrade the plants so that even more product can be dumped into an already saturated market, or keep the cash and watch the huge losses-per- shipped-chip grow even larger. - The company has a nasty test problem. Low yields at test, send chips (with many manufacturing dollars already invested), either to the dumpster or to expensive re-work. - The company appears to have gone out of its way to carefully craft a lie to cover up its burgeoning problems as far as its recent public earnings report is concerned. The report didn't even hint that there might be serious difficulties or problems . The addition of the TXN plants should have seen mbit shipments rise very significantly. Instead they actually dropped. The 10% plus in mbit shipments indicated to the public, was actually 10% minus. Any analyst with even a modicum of understanding of the game would have experienced heart palpitations if this fact had been known. The company appears to have made a decision to try to cover this up, at least until the test problems and the TXN plant messes could be cleaned up. Further motivation to mask the problem may be apparent in the the massive secondary just announced. It is comprised largely of insider stock, which will be offered to the lambs before they realize MU is in serious trouble. The insiders have sold heavily for many months. Now they want to sell even the stock that is available through their options. One can understand the desire to grab all valuables and head for the lifeboats (given the mess) but the apparent cover-up of the problems to provide time for an insider exodus is a bit much, even in these mad times. I sincerely hope the SEC both investigates this glaring set of highly questionable circumstances and leans heavily on these callous rascals, if the evidence at hand is corroborated. As this is written, most analysts, except Tom Kurlak, have a "buy" or "strong buy" on the stock. Most have projected numbers that would shame a Kindergarten child as far as any use of reasonable assumptions or acceptable arithmetic is concerned. Niles in particular just seems to insert random (but large) numbers into his "model". It will be interesting to see if he in any way "revises" his silly predictions, given his use of what are now publicly recognized as flawed assumptions. To do otherwise tells much about the quality of his analytical work. Kurlak is to be congratulated for "damning with faint praise" as a means of protesting his recent enforced recantation of the bearish faith. A "neutral" on MU says it all. Best, Earlie