For the 10th anniversary of The Motley Fool, we're looking back at some of our best columns. My colleagues honored me on Monday with my debunking of Qualcomm's (NASDAQ:QCOM) valuation in early January 2000. In response, I'd like to present the worst column I have written in four years here at the Fool. I do this for two reasons: First, my Protestant modesty won't allow this kindness to go unpunished. Second, it's when we make our biggest mistakes that we learn the most. PSINet chastened me. Big time.

Oh, sure, I write hundreds of columns per year, and there are plenty of instances where my interpretations are incorrect. But in this case, I stared a dying company square in the face and said, "Get him back in the game, boys. He's fine." Wrong. Within five months, PSINet filed for bankruptcy and was sold off for scrap. The problems were as plain as the nose on my face, and I ignored them.

As you read on, you'll see that I was fairly diligent about disclaimers and warnings. Didn't matter much -- the stock rocketed up to as high as $4.45 in the next week, buoyed by reports that The Motley Fool had initiated a strong buy.'s Jim Cramer had a field day, hollering on CNBC that he'd love to be on the other side of that trade. Jim had many of his facts wrong -- including his claim that I "recommended that people go long PSINet," but he was still spot on -- PSINet was doomed, and my analysis was lousy.

What are the lessons here? Lesson #1: When you find yourself making qualification after qualification ("If this, this and this happens, then it might succeed"), you are likely taking on much more risk than you believe. In this case, PSINet had to sell some assets; the problem was that everyone KNEW they had to sell some assets. Get full value? Not on your life.

Lesson #2: Debt kills. Lesson #3: Do not expect that a textured argument made will be received the same way. "This is a fine company with problems" became "Strong Buy!!!!!!!" in translation. Both, of course, were wrong, but one substantially more wrong than the other.

Lesson #4: Companies under this type of threat are in no position to negotiate. Lesson #5: Irony on the page can fail miserably. I would never, ever say " I'm hot on PSINet" with a straight face. Can you see my face? No? Failed humor.

Herewith, then, is a multiyear low for me, offered with the greatest humility.

P.S. Notice how NONE of the e-consulting businesses listed in this article are in business anymore? What carnage. Here goes:

I think PSINet's going to succeed.

Yep, you heard right. I'm hot on PSINet. In fact, I'm late to the show. Already in the New Year, PSINet is up more than 80%!

To $1.44 per stub. Oh, never mind.

PSINet was beyond a shadow of a doubt one of the big horror shows of last year, a company that quickly beat a retreat from market darling to complete goat. Last year at this time, no one could say enough good about the company, its strategy, or its management team (fronted by CEO Bill Schrader). By the end of the year, PSINet was the gang that couldn't shoot straight, beset by class action lawsuits, a stock that had declined 98%, and the unimaginable horror of having Schrader's stock holdings liquidated by a creditor in a margin call for a personal $25 million loan.

2000 was not a year to write home about for PSINet. It was, however, a year to remember. Regardless, I still really like the fundamentals of this company.

(Stop right here. Near the end of the article, I am going to list some huge risks related to PSINet. If you cannot commit to reading all of the next 700 or so words, read none of them. It is rare for The Motley Fool to cover a company that has penny stock status. I feel strongly about this one from a qualitative perspective, but there is a really good likelihood that PSINet will fail or languish in penny stock hell for eternity until it is taken out in a mercy killing by the likes of famed raider T. Boone Pickens. OK, back to the show.)

Through all of the tough times that PSINet has endured, it is still signing up customers left and right for its Web hosting and Internet Protocol services. General Electric (NYSE:GE) signed on, as did Hewlett-Packard (NYSE:HPQ). So has Sallie Mae (NYSE:SLM).

And the company has continued to invest in additional facilities. It started service on additional bandwidth, four OC-192s running from Atlanta to New York to Toronto came online in December, and additional wideband links are scheduled to come online this month connecting New York to Boston and Atlanta to Miami. These facilities represent enormous investments in capacity and equipment, which the company is purchasing primarily from Juniper (NASDAQ:JNPR), Nortel (NYSE:NT), and Cisco (NASDAQ:CSCO).

It also has brought up three of its Web-hosting centers, in Dallas, Toronto, and Amsterdam, representing 300,000 square feet of state-of-the-art space at a cost of $200 million. This is something that has remained deep in my mind: If the company is, as it would seem by looking at the stock price, in danger of failing, why on earth would PSINet be spending all of this money?

The answer, I am convinced, is that PSINet's core business is much better off than the market thinks it is. Since a fateful day in November when a bad earnings report caused the company's stock to plunge by more than 50%, it has been put forward as a poster child of an organization with a crisis of leadership, as well as one that got caught up with the excesses of the seemingly endless streams of money that were being thrown at bandwidth companies in the past few years.

But I'll tell you what I hear from people in the business. Bill Schrader is about as smart and driven a person as there is. A local wag recently called him "the mind that MicroStrategy's (NASDAQ:MSTR) Michael Saylor fancies himself to be." Schrader made a horrendous mistake when he margined his shares in PSINet, but there really wasn't anything stopping him from selling a few into the froth in 1999 -- yet he did not.

Schrader has also been quite open about the misjudgment the company made when it bought Metamor Worldwide -- along with its 80% stake in Xpedior (NASDAQ:XPDR) -- last March. In November, PSINet declared its entire investment in Xpedior to be "impaired," meaning it was written off as an asset and any revenues it had could not be included in PSINet's total. That was a cruel turnaround from a few months before.

But Xpedior's dive has been consistent with the other companies in the e-consulting space, among them marchFIRST, iXL, and Razorfish (NASDAQ:RAZF). In other words, PSINet made a determination that it needed an asset represented by the holdings in Metamor, paid a price consistent with the going rate, then watched in horror as the entire sector collapsed along with the change in the economy.

Ah, the economy. You knew I'd bring it up eventually. In hindsight, PSINet's big mistake -- mirrored by most carriers -- was assuming that the economics of the Internet and bandwidth would not change. The expectation for return on investment in 1999 was dramatically different than it is now, and companies that have racked up enormous amounts of debt and have continued capital requirements got caught in a pinch. PSINet swaggeringly announced acquisition after acquisition to build out its network. This is the primary reason why the Competitive Local Exchange Carriers (CLECs) have all but shriveled up and blown away, and this is what happened to PSINet. It still needs a significant amount of cash to operate until the investments it has made in Web hosting and infrastructure pay off.

But even with the changing environment, the annual revenues for data-hosting centers still come out to $2,500 per square foot. That's a big opportunity for PSINet. In fact, it's a gold mine. Through all of this, PSINet still receives the highest rating by customers of any of the national or international Internet service providers (ISPs). In an industry where there really is no such thing as a happy customer, if PSINet is able to maintain or even improve its service ratings, it's got one of the cheaper sources of additional revenues: existing customers.

The question is how they get there from here. PSINet needs additional money to operate, and after a bad earnings report in November, in this market environment any cash it needs will be expensive. PSINet has hired Goldman Sachs (NYSE:GS) to look for strategic alternatives, including the sale of some of its non-core assets, such as Xpedior (which would be for pennies on the dollar, if that), its program, or even of its Web-hosting centers to a real estate firm that would lease them back to the company. None of these paths will be easy, or cheap, for PSINet.

Nor is it certain that broadband economics will not somehow get worse. How will the company get the cash to survive? Man, that's the $2.5 billion question. As for whether a surviving PSINet can expect a stable environment for its services, consider this: At the end of this shakeout, there will be far fewer international ISPs standing. Those that remain and provide top-tier end-to-end service should expect broadband demand to continue to skyrocket.

Fool on!
Bill Mann, TMFOtter on the Fool discussion boards

The Motley Fool is celebrating its 10th anniversary with 10 Ways to Make More Money Now .

Bill Mann owns none of the companies in this article, especially the ones that no longer exist. He swears that the night terrors subsided after a year or so. The Motley Fool is investors writing for other investors.