I suspected that my dueling partner Anders Bylund would say Cisco Systems (NASDAQ:CSCO) looked cheap after its shares recently plummeted. To make that claim, he's offering up two critical assumptions with which I simply must disagree.

First, he's assuming that using relative valuation is the right way to invest. It may look like a historically cheap multiple for Cisco's shares, and Cisco may look inexpensive compared with its peers. But that's the same sort of thinking that led to the dot-com bubble in the first place. Cisco's bubble peak of around 200 times earnings looked cheap, in the terms of the era. It was, after all, being compared with companies that grew all sorts of obscure metrics but never profited.

Besides, 23 times earnings looks decent only when compared against Nortel's (NYSE:NT) loss, Juniper's (NASDAQ:JNPR)'s 62 P/E ratio, or Ciena's (NASDAQ:CIEN) 53 P/E. The overall market, however, as measured by the index-tracking SPDRs (AMEX:SPY), looks significantly cheaper, with a P/E of less than 16.

The second assumption Anders had to make to justify Cisco's valuation is that it can still continue to grow in spite of clear cyclical troubles in its primary market -- a market, I might add, that Cisco already dominates. For a scary parallel on how poorly that may wind up, check back a little over a year ago to a duel on Select Comfort (NASDAQ:SCSS). A certain growling bear warned of the dangers of investing in cyclicals that looked cheap on a P/E basis. That bear got soundly defeated at the ballot box, but won in the market. Those shares currently fetch less than half of what they did then.

No company can outgrow its market forever. Not even Cisco. Until its shares reflect that reality, my cash will stay away.

You're not done with this duel yet! Read the other three arguments, sound off at Motley Fool CAPS, and vote for a winner.