PS: From 1990 to now the Nasdaq beat the Dow.
If you bought it in 1990, the correct investment decision would have been to sell it in late 1999-mid 2000 when it seemed to plateau, and possibly buy back in to it later. However you look at it, staying long in a bubble is always irrational. Even if the company performed miraculously before the bubble and will perform miraculously after it, you can only lose by holding it in the bubble.
(Of course, this predicates knowing when the sector is in a bubble. Which, suffice to say, is hard.)
Or you can look at the bubble from the point where explosive growth started around 1998-1990 and died in 2000.
PS: Looking only at winners distorts the picture, so while MSFT was still undervalued (relative to its current dividend adjusted price) in 1996 there were also plenty of overvalued stocks in 1996 that tanked.
No you wouldn't. MSFT sits now at about a 20% discount to what it was in 2000, while CSCO is trading at about a third of the price it had in early 2000.
This is the defining feature of a bubble -- not only do the bad companies fetch exorbitant prices (there are plenty of those in any economy), but the good companies with valid business models also trade at prices that cannot be justified, meaning that no matter how well you pick the winners, you are still going to get burnt when the bubble bursts.
There were losers in the british railroad bubble in 1830's, but the ones who bought into the winners, did get out ahead. Meaning, the bubble never burst.