At least Carl Icahn is a billionaire. He can afford to blow a few million now and then -- just as he's doing now, by filing suit to force a sale of BEA Systems (NASDAQ:BEAS).

Here's the skinny for those who've missed the drama thus far. Oracle (NASDAQ:ORCL) revealed a $17-a-share bid for BEA on Oct. 12. Stirring speculation that a bidding war would ensue, shares of BEA soared to as high as $18.94.

Then, apparently expecting a counterproposal, BEA and Icahn said the bid was too low. Icahn went so far as to suggest an auction, whereby the company would be sold to the highest bidder. BEA never really responded. Instead, it opted to tell the world that it was worth $21 a share -- take it or leave it.

Oracle CEO Larry Ellison didn't bite, and neither has anyone else.

Cash me out, please
Now, Icahn is suing and promising a proxy fight, so as to presumably cash out his holdings as soon as possible, if this letter filed with the SEC is to be believed. Here's the most relevant excerpt:

BEA should not allow the stalking horse bid from Oracle to disappear (failure to take the Oracle bid as a stalking horse would be a grave dereliction of your fiduciary duty in my view). If a topping bid arises, then all the better. But if no topping bid arises it should be up to the BEA shareholders to decide whether to take the Oracle bid or remain as an independent Company -- not THIS Board ... .

Whether you know what a stalking horse is or not -- I don't, for the record -- there's a bit of flawed logic in Icahn's letter. You can't force an auction if no one else -- whether IBM (NYSE:IBM), Hewlett-Packard (NYSE:HPQ), or SAP (NYSE:SAP), which recently agreed to spend $6.8 billion for Business Objects (NASDAQ:BOBJ) -- wants to buy BEA.

Never buy for the buyout
But give Icahn some credit. He's right in that BEA shareholders should have the choice over whether to sell the company to the highest bidder. Last week, that was Oracle. Now its $17-per-share bid is off the table. For me, it perfectly illustrates why you shouldn't do as Icahn did.

That is, you should never buy on the hope that your stock will be acquired. You may not like the offer you get. And, if that's the case, there may be very little you can do about it.

Consider Applix (NASDAQ:APLX). Had you been a buyer simply for the acquisition possibilities -- I wasn't -- you'd be stuck with a $17.87-per-share offer from Cognos, which, on a cash flow basis, sharply undervalues the company.

Trouble was, I had little choice but to sell my shares and advise our Rule Breakers subscribers to do the same. Cognos was buying Applix via a tender offer. What that means is that if just 51% of investors agreed to sell for $17.87, Cognos would get the other 49% -- whether I wanted to sell or not.

Most buyout offers are like this -- Microsoft's bid for aQuantive, for example. Other deals are stock-swaps -- and require shareholder approval -- but Oracle is so cash-flow rich that, were a deal to get done, it's a good bet that cash and debt would be the financing source. Shareholder approval would likely be needed, but with Oracle's excellent track record and Icahn's 58 million BEA shares, getting that done isn't likely to be an issue.

If, that is, there's still a deal to be had.

When to buy a business in distress
For some people -- Icahn, for example -- buying businesses in distress is performance art. Not me. I'd rather eat spinach-flavored ice cream.

Nevertheless, turnarounds can be exceedingly profitable. What's required is that you buy a business with a reasonable chance of fixing itself. That's where I think investors went wrong with BEA. Once a dot-com superstar, it now badly lags in the market that it helped to create and, as a result, is winning fewer new customers each year.

Oracle may yet buy it -- there is a strategic fit here, after all -- but only if BEA comes cheap. Good idea. Too bad investors who bid this sucker up in a speculative frenzy failed to act as responsibly.