re: TXN:
Richard Palm, on the AMAT thread, asked me to comment on your post.
Disclosure: I've never held a LT position in the stock. I've range-traded it, about 8 times, in both directions, all this year, mostly trading the 30-38 range. Bought a bunch (long) in the September panic-selling, and have recently been adding (small increments, from 31 on down). I plan to buy LEAPs (04C35s) if the stock goes below 20, and then sell my higher-cost lots on the next rally into the mid-to-high 30s, and hold my lowest-cost shares LT.
12 months ago, the analyst consensus EPS estimates for TXN, were about $1.50 in 2001, and $2.00 in 2002. Since then, both those estimates have fallen, and are approaching zero. Analyst estimates, IMO, tell you what current sentiment is. They tell you nothing about the future.
In a ratio, like P/E, as the denominator approaches zero, the ratio approaches infinity, and the ratio becomes absurd. That's what's happening here. Every tech company, TXN included, found themselves suddenly oversized in 2001. In a capital-intensive industry like chips, profits are going to disappear in conditions like we're currently seeing (sharpest decline in capacity utilization, ever). Actually, if you use GAAP earnings, you can't calculate a PE, as EPS has gone negative.
When conditions come back to normal, and the SentimentPendulum swings the other way, the analysts will start ratcheting their EPS estimates up, as fast as they have recently been bringing them down this year. When I say "normal", I mean: no recession, average supply/demand conditions, and no bubble either.
Management has given clear indications they see business conditions bottoming. Sales, margins, profits, the news gets better from here on out. How fast that happens, depends on macro conditions, which is a big unknown. But it's a question of when, not if. Consumers are not going to stop buying gadgets with DSPs in them; wireless is about the only part of telecom where capex is growing.
So, what's a reasonable valuation for TXN? According the S&P, the 5-year PE range for TXN is 20-94. I'd consider a reasonable PE range to be 20-40. When the recession is over, EPS will rebound to the LT trendline (20-25% growth). If TXN earns $1.30 in 2003, then a PE of 20 gives a stock price of 26. That's a reasonable price. Again: using trough earnings from 2001 is using the wrong yardstick. And 2002 earnings could be anywhere from $1.30/share to zero, depending on your guess of macro conditions.
The analogy to Xerox and Polaroid is wrong. DSPs are not a mature industry, TXN's balance sheet doesn't look anything like theirs. DRAM makers will not be able to transition quickly into TXN's markets, as TXN's products are not commodities. It would take a decade, starting from scratch, for a company like Intel to build up it's patent portfolio and relationships with customers, to be able to challenge TXN in DSPs. A company like Micron has zero chance of doing it, not now, not ever.
In the recent panic selling, the price bottomed at 20. I doubt sentiment gets worse than it was in late September, that had the feel of a capitulation low. At least a medium-term low that will hold well into 2002. We could retest it, in the current dip, but I doubt we go lower. I've make money short and long TXN this year, but I wouldn't short it here. Covering a short, and/or going long, in the low 20s sounds, like a good plan. $6/share, P/S of 1, that'll never happen. |
