Han Dieperink: Party like it's 1999

December 9, 2025

_{This column was originally written in Dutch. This is an English translation.}

By Han Dieperink, written in a personal capacity

Anyone who invested during the 1990s will remember the dot-com bubble well. Anything with “.com” in its name skyrocketed, even if the company had never made a penny in revenue. Investors shook their heads pityingly at banks, utility companies and manufacturers that had been paying dividends for a hundred years. Boring, outdated, old economy. Money flowed in only one direction: towards technology, media and telecommunications, the so-called TMT shares.

The result of this trend was that these “boring” shares became increasingly cheaper. When the bubble burst in March 2000, it was precisely these undervalued shares that remained stable, while the internet stars plummeted by 80% or 90%. In hindsight, the best way to time the peak was child's play: slowly sell what everyone is buying and buy what everyone is dumping. Contrarian investing in its purest form.

Today, seven companies – Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta and Tesla – determine how high the S&P 500 stands. Together, they account for 35% of the index. The remaining 493 shares hardly participate. If you look at the S&P 500 Equal Weight, in which each share has equal weight, that index lags far behind its larger brother in terms of market capitalisation.

The difference between the two indices is now almost as extreme as it was at the end of 1999. Back then, it was due to the “internet”; now it is called “AI” and “cloud”. The dynamics are identical: all the fresh money, all the institutional money, all the FOMO money is flowing to the same handful of names. In order to buy them, shares elsewhere have to be sold. As a result, energy, basic materials, pharmaceuticals and consumer goods are falling to valuations we haven't seen in twenty years.

Fortunately, there is also talk of “This time it's different”. And to be fair, there are differences. The Magnificent Seven are not loss-making start-ups. They are printing billions in profits, have unassailable market positions, and are still growing rapidly. Nvidia sells every chip it can make, Microsoft and Amazon dominate the cloud, Apple is in the pockets of two billion people. Compared to Cisco at 140× earnings or Pets.com with no revenue, current valuations are almost modest. For a real bubble, most of the Magnificent Seven stocks would have to double.

But that is precisely where the pitfall lies. Precisely because these companies are really good, no one feels the urgency to sell. The euphoria is not hysterical, it is calm, calculated and confident. And that makes it more dangerous. Because the longer a reasonable story is stretched out, the higher the eventual correction can be if growth disappoints for even one quarter or if interest rates do not remain low forever. History teaches us that extreme concentration and extreme valuation differences never last forever. Sooner or later, the tide will turn. Sometimes with a bang (2000, 2022), sometimes more insidiously through years of underperformance by the former darlings. But it always turns.

Does this mean you should sell everything now and load up on energy stocks and supermarket chains? No, because the rotation may not happen for years – and those who jump in too early will suffer greatly from “missed profits”. Furthermore, in the late 1990s, the stock market ensured that value investors in particular were smoked out. After the dot-com bubble, there were hardly any value investors left. This time around, too, there are value investors who have been saying since 2017 that “it's really going to happen now” and who are now lagging behind by 15-20% every year.

The wisest approach is to take a middle ground. Gradually take profits on positions that have quintupled in five years and reinvest part of them in sectors that are now trading at a 30% or 40% discount to their own history. Not because you think you can time the top – no one can – but because your portfolio is slowly shifting to a situation where a correction of the Magnificent Seven can no longer affect you.

Because one thing is certain: when everyone says that “this time it's really different” and that “you'd be crazy not to be in AI”, the difference between expensive and cheap is usually already historically large. And that's exactly when the most boring shares in the world are often the most exciting.

History rarely repeats itself exactly, but it always rhymes. And the rhyme we are hearing now sounds suspiciously like that of 1999. We are still on the good side of the bubble for investors, the upward phase. The scenario also resembles the Goldilocks scenario of the second half of the 1990s in several respects, and in many ways it is even better. Hence the label Uber-Goldilocks. In that scenario, we will end 1998 this year and then see the bull market stabilise, the last and most euphoric phase of the bubble. Party like it's 1999.

More columns by Han Dieperink