The following is an unofficial and edited transcript from Thursday's "Mad Money."
Today was a triumph of the technicals over the technological. Today was a day we touched the Dow 8,000 level -- down 20% from where Cramer last said to sell. When you hit that level, you catch buys. If you're using Cramer's strategy of buying stocks with bountiful dividends like Caterpillar at 4.5% or Nucor at 4%, you caught a great price earlier in the day.
Now you should be done buying and, as the high-yielders rally, it's time to start the selling. You can't buy again until the stock takes out your last low price and the yield's even bigger. That's the only strategy that's worked consistently in this crazy market -- stocks that bounce most have the biggest yields.
The NASDAQ closed up 6.5%, but from peak to trough, it was up 11.8%. This is undeserving after Wednesday's shocking news from Intel -- it preannounced to the downside. Early today, that caused us to take a beating. But then we touched the Dow 8,000 level and caught a rally sending tech stocks higher -- when they should have gone lower. Technicals have produced endless trading rallies like this one, and the thing to do every time is to sell tech. Don't be seduced by the notion that we've bottomed for real. This was a trading bottom on the idea that tech is, of all things, a safe haven.
The market isn't done discounting the bad news we got from Intel. That's garbage. The truth is, the fundamentals of tech are bad, as Intel's preannouncement revealed, and if you own tech, you were saved from huge losses by a late-day rally based only on technicals -- NOT fundamentals.
Remember Rule 22 from "Real Money: Sane Investing in an Insane World": when a company preannounces an earnings shortfall, wait at least a month before you buy it even if you think there's a bottom, because no company would ever preannounce if it thought things would get better.
All this talk about a tech bottom is fanciful. Cisco stated the recession really got going in October, so we're one month in. We're conditioned to buy tech halfway through a recession, but if you think this recession is only going to last two, or even four, months, you're fooling yourself. Don't believe today's rally is an indication of a real bottom, a turn for the better. The only turn you should expect is one for the worse. When people tell you this is the time to buy, because we're "knee-deep into a recession" that's lunacy. We're not even ankle-deep.
Take a look at history, which is repeating itself: we know how a tech bottom looks because we got one in late 2002 and 2003, but that was after the NASDAQ composite fell 78% from its peak in 2000. We're talking two years, and while tech stocks were inflated in terms of valuation going into 2000, does anyone doubt that the current recession is much, much worse than that baby recession -- really just a slowdown -- from 2000 to 2002? At that time, the NASDAQ composite bottomed at 1,108. It's currently at 1,596, more than 480 points higher. We may be closer to the bottom than the top right now -- but we're headed down, not up.
What we need to see for a real bottom in tech is repeated estimate cuts from analysts, not repeated upgrades of semiconductor and semiconductor equipment stocks. These companies need time to work off inventory, and more importantly, the darned economy needs to start improving before tech can bottom.
Does anyone think the economy's improving?
We just got signs from Intel and Dell that things are getting worse. The last time tech bottomed, Intel fell 82.7% from peak to trough, Dell fell 71.4%, Microsoft went down 66.4% and Oracle dropped 84.4%. Those stocks got killed and kept getting killed.
Bottom Line: Don't be seduced by the sirens of a "tech bottom." Remember 2000 to 2003 -- we've been there and we need to heed that decline, not ignore it. This isn't the bottom and it isn't the time to buy tech, but thanks to today's rally, it's not too late to sell.
Questions for Cramer? email@example.com
Questions, comments, suggestions for the Mad Money website? firstname.lastname@example.org