I am going to cut and paste the abstract summary below and then turn it into English:
Abstract:
I study outcomes for a variety of "do-nothing" portfolios constructed from constituents of the S&P 500 index, from 1971 to 2025. These portfolios maintain their positions even for those stocks that exit the index. The findings include (i) initially equal-weighted portfolios outperform initially value-weighted portfolios over the full sample period, (ii) value-weighted "do-nothing" portfolios essentially match the index on average, (iii) portfolios constructed from the largest constituent stocks have recently outperformed portfolios constructed from all index constituents, but this is atypical, as over the full sample the largest-stock portfolios performed quite poorly, and (iv) narrow portfolios of randomly selected component stocks generate average returns similar to the index, but underperform the majority of the time, and more so over longer periods and for narrow portfolios.
\-------------------------
(i) equal weighting every stock in the index from the starting period did the best
(ii) matching the market-weighted index (the way the index is currently constructed) matched the index. Of course it did since that is the way the index is constructed.
(iii) owning the largest mkt cap companies in the index was a bad idea even if it has been a very good idea most recently.
(iv) it is very hard to beat the equal weighted index with a more concentrated portfolio, i.e. stock picking consistently is very hard.