First off, do we agree that larger companies tend to remain in business more than smaller companies do?

...

BUT, there is never a long term in which the top 100 do worse than the bottom 400. And even if the top 100 do equal to the bottom 400, the equal weight still has that "drag" of owning more of the failed companies.

The top companies can wipe out, too. In Canada, we had Nortel at #1, in fact worth more than the next few combined, just a year or two before it went bankrupt. Nokia did something similar (without quite going bankrupt) a few years later, in Finland.

What about the USA? Surely not so vulnerable?

You might have a look at this site: finhacker.cz - Top sp companies by market cap . It lists the top 20 companies in the S&P 500, by market cap, from 1989 to today.

It starts in 1989 with Exxon, IBM, GE, Bristol-Myers-Squibb, and Merck in the top 5 spots. Honourable mention to Verizon, AT&T, 3M, AIG, Boeing and Pfizer, who were in the top 20. With the exception of AIG, none of them are bankrupt, but they have not provided good returns, and with the exception of Walmart, not a single top-10 company is still in the top 10, 35 years later.

Skip ahead to 1995, and drug companies were all the rage, with Merck #4, Johnson and Johnson #6, BMS and Pfizer #12 and 13, and Eli Lilly #18. That last one would have worked out very well, thank you obesity drugs, but all the others have done poorly. As have Coca-Cola (#2!), Procter and Gamble (#5), IBM (#9) and Intel (#10), not to mention AIG and Fannie Mae (#14 and #16), whose shareholders were wiped out 13 years later.

End of 1999 of course tech was the thing, and Microsoft was far ahead of all the others. That investment has worked out well, but of course Intel (#6), IBM (#8), AIG (#14), Qualcomm (#16), AT&T (#19) and Verizon (#20) have been terrible investments since then.

Of today's stars, most will not go bankrupt either, but you are making an awfully big bet on those top 20, which are worth about the same amount as the other 480 companies in the S&P 500 combined. They may seem untouchable today, flying high on recent successes, but whoever is looking at that table in 10 years may wonder what all the fuss was about with Nvidia (#1), Broadcom (#7), Tesla (#8) or Palantir (#19). Of course, my hunch is that Tesla will be #1, but my hunch could well be wrong, and it might also be #50, who knows? Buying these stars of 2025 probably means we are buying huge stakes in the things that are most popular today, but which may well fall back when the wisdom of paying 40, 60, 100 or 200 times earnings for these growth stars starts to look dumb, or when something new comes along that makes the internet and AI look stale.

Regards, DTB