When, exactly, will the Internet stock bubble burst? And I mean burst, not like last week, which was just a minor blip by comparison.
Well, I don't know. My assets are in a blind trust that routinely underperforms market averages.
And even if I did know, why would I tell anybody?
But the National Venture Capital Association (NVCA) recently invited me to address its annual meeting. And there weren't enough seats for all the venture capitalists (VCs) who could afford Boston's Ritz-Carlton.
According to the 1999 NVCA Yearbook (http://www.nvca.org), 1998 was the seventh straight record-breaking year for VC fundraising and investment. Exactly 547 companies raised $US25.3 billion, bringing their capital under management to $84.2 billion. They invested exactly $16.7 billion in 2859 growth companies, 60 per cent in information technology, with $3.5 billion going into Internet-related ventures alone. Exactly.
Chatting up NVCA attendees, I learned that nobody knows anything exactly about the future of Internet stocks.
VCs know trees don't grow to the sky, all stocks eventually sell for 10 times earnings, there is too much money chasing too many deals, and Internet stocks have run off the end of a cliff.
But VCs aren't looking down. They are scrambling to take their ripe (and not so ripe) portfolio companies public before the bubble bursts. The trick is to get in on the ground floor and get out early.
When it was my turn to talk, I took it upon myself to predict the exact day the Internet stock bubble will burst.
With Y2K looming, I used the following date-day algorithm, thanks to reader Peter Sprague, for calculating on which day of the week a future date falls.
Take the future date's year and set A to 2 if 1999 or 3 if 2000. Set B to the result of using the date's month as an index into the vector [2,5,5,1,3,6,1,4,0,2,5,0].
If the year is 2000 and the month follows February, add 1 to B, because 2000 is a leap year. Finally, add A, B, and the day of the date's month, divide by 7, and use the remainder ("modulo 7") to select the day: 0 for Sunday, 1 for Monday . . . and 6 for Saturday.
So, for example, January 1, 2000, is 3 (for 2000), plus 2 (for January), plus 1, modulo 7, which yields Saturday. What a relief to know we'll have a whole weekend to work out any Y2K problems that might arise.
Among those Y2K problems will not be the bursting of the Internet stock bubble. Internet stock values are grossly inflated by "dowager corporations looking for Web gigolos to take them to the Internet Riviera", so investors are sure to bail out way before any Y2K problems arise.
Some say Internet stocks are not worth their multiples, but the eventual winners will be, so place your bets. Others say investing in all Internet stocks will prove profitable even though nine out of 10 will fail. This kind of thinking will inflate the bubble, at least through the summer, if I know anything about stocks, which, again, I don't.
The bubble will burst some time after October reports of disappointing third-quarter earnings and warnings about fourth-quarter shortfalls are announced. The pressure to get out ahead of year-end tax-loss taking will build.
Therefore, the Internet stock bubble will burst on -- don't quote me -- November 8, 1999. Sorry they jumped the gun last week.
Just to be sure, let's run a quick check with our date-day algorithm. Taking November 8, 1999, we get 2 (for 1999) plus 5 (for November) plus 8, modulo 7, Monday, which, after a weekend of hand-wringing, is a likely day for a stock market bubble to burst.
Hearing this, an NVCA member asked me how deep the stock crash will be and how long it will last. I said that I don't know -- thousands of points, and it will last a year.
Won't be long before Yahoo is again a race of exclamation point-less men from Gulliver's Travels.
Internet pundit Bob Metcalfe invented Ethernet in 1973 and founded 3Com in 1979. Send e-mail to email@example.com or visit http://www.idg.net/metcalfe