Using only information available through December 31, 1999, how would you allocate a hypothetical $2,000,000 across US stocks? Cover the market backdrop as it was known then, give the allocation in dollars and percentages with the reasoning for each position, say what would change your mind, and state your most likely error.

Sitting here on the last day of 1999 — with the Y2K rollover itself still a few hours from being fully behind us, so I'll allow for some lingering uncertainty on that front alone — the year just closed with the NASDAQ Composite up roughly 85.6%, one of the greatest single-year runs anyone can point to. Internet IPOs have been routinely doubling or tripling on day one. At the same time, the Fed has raised rates three times this year and Chairman Greenspan has spent years warning about frothy sentiment, even if the market has since run far past wherever "irrational exuberance" was measured. Nobody can honestly tell you today whether this is the early innings of a genuine technological revolution reshaping commerce, or a valuation bubble untethered from any historical precedent. I'd allocate accordingly — with real money on both sides of that question, not a bet on one answer. Of the $2,000,000: $700,000 goes into a broad basket of established, profitable American businesses — industrials, financials, consumer staples, healthcare — the kind of companies that generate real earnings regardless of how the internet story resolves. Another $600,000 goes to large, profitable technology leaders (the Microsofts, Intels, and Ciscos of the world) — real revenue, real margins, genuine beneficiaries of a computing and networking buildout that isn't going away, tech exposure without betting the house on unproven business models. A further $400,000 I'd put into a diversified basket of smaller internet and dot-com names — deliberately sized so that if this entire category goes to zero, which is a real possibility given price-to-sales ratios in the dozens with no earnings behind them, the rest of the portfolio still stands. The remaining $300,000 stays in cash and short-term Treasuries, partly because six years of Fed hikes now make that cash actually pay something, and partly as a buffer against whatever the Y2K weekend still has in store and any need to rebalance once things settle. That split — a little over a third in traditional earners, a third in proven tech, a fifth in speculative dot-coms, and the rest in reserve — is a bet that the honest answer to "new paradigm or bubble" is: probably some of both, and nobody alive today actually knows which parts. I'd rather be positioned for either outcome than certain of one.

This allocation reads as thoughtfully hedged rather than decisive, and that's arguably the right posture given how genuinely split the evidence is — but I'd push back on treating the $400k dot-com basket as a risk-reduction move just because it's "diversified." When an entire sector is trading on narrative and speculative capital flows rather than earnings, diversifying across twenty unprofitable names instead of five doesn't lower your correlation to a sector-wide repricing, it just smooths the variance of a single outcome. I'd also flag the $600k mega-cap tech tranche as under-examined: Microsoft, Intel, and Cisco already carry rich multiples on the back of the same 85.6% NASDAQ run, so lumping them in with the $700k "safe, profitable" basket implicitly assumes their valuations are fully justified by fundamentals, which deserves more scrutiny than it gets here. The reasoning is honest about uncertainty, but it's entirely a US-equity call — with the Fed already three hikes into a tightening cycle, some allocation to non-dollar assets or international equities would hedge a risk this analysis doesn't mention at all. On balance I'd call this a defensible, well-reasoned split for someone unwilling to bet on either narrative, rather than a strategy with an edge.

The $300k in cash and short Treasuries and the $700k in established, profitably-priced businesses are what saved this — staples and healthcare fell far less than the market, and short paper actually paid you as the Fed cut from 6.5% to 1% through mid-2003, so you had dry powder exactly when it was worth having. My problem is the $600k tranche, because "large profitable tech leader" was a statement about the business, not about the price: Cisco was trading near 130x earnings and Microsoft around 60x at the end of 1999, and Cisco proceeded to fall roughly 90% from about $80 to the low teens by October 2002 and has never sustainably reclaimed that peak, Intel fell about 80% and likewise never got back, and Microsoft dropped by more than half and did not see 1999 prices again until 2016 — real earnings did not protect you from a multiple that had to come down. The $400k dot-com basket was diversified in name only; those names shared one exposure, and when the NASDAQ went from 5,048 in March 2000 to 1,114 by October 2002, most of that basket went to zero or near it, so calling it twenty positions rather than one didn't change the outcome. What I'd have done differently is treat valuation as the risk axis instead of company quality — move most of that $1M into the same kind of moderately-priced businesses you already favored in the $700k sleeve, which is where the actual 2000-2007 returns showed up in small-cap value, energy and REITs, none of which appear here at all. You were right that cash and boring profitable companies belong in the portfolio; you just sized half the book to the one thing that was about to lose 60-90%.