A fund that calls itself thematic does not always look very different from the market it is supposedly cutting across. Across 392 funds with usable holdings data, the typical thematic ETF holds 32% of its weight in names also found in the S&P 500. That median masks a wide spread. Semiconductors funds sit at 64% in the S&P 500, Defense and Infrastructure funds at 41%, while Gold miners funds sit at just 5% and Cannabis funds at 0%. At the far end, 73 of the 392 funds carry more than sixty percent of their weight in index names already.
Active share, which measures how different a fund's weights are from the S&P 500 book, runs the opposite way in several themes. Gold miners, Nuclear and uranium, and several metals and mining themes show a median active share of 100%, meaning essentially none of their weight lines up with the index the same way, even though some of those same funds hold low in-index weight by the first measure. Artificial intelligence funds show a median active share of 80%, Semiconductors funds 85%, and Metaverse and immersive tech 80%, the lowest active share readings in the table alongside Humanoid robots at 94% and Quantum computing at 84%.
Fees sit at a median of 0.60% across 419 funds in 54 themes. Measured a different way, against only the slice of each portfolio that is not already in the S&P 500, the median fee for that difference is also 0.60%. Theme by theme the two numbers sometimes match exactly, as in Defense, Infrastructure, Gold miners and several mining and metals themes, and sometimes diverge, as in Artificial intelligence (0.70% versus 0.80%), Semiconductors (0.30% versus 0.40%), and Cannabis, where the fee for the difference (0.60%) is lower than the headline fee (0.80%).
One-year total returns varied widely by theme, from +82.2% in Semiconductors and +61.5% in Genomics and gene editing to −32.1% in Blockchain and crypto equities and −28.5% in Nuclear and uranium. Measured against the S&P 500 over the same year, some themes saw every fund beat it, including Semiconductors (9 of 9), Biotech (8 of 8), Cybersecurity (7 of 7), Copper (6 of 6) and Genomics and gene editing (3 of 3), while others saw none do so, including Defense (0 of 14), Gold miners (0 of 9), Internet and e-commerce (0 of 13), Clean energy (0 of 16) and Blockchain and crypto equities (0 of 13).
Separately from how each fund compares with the S&P 500, some funds overlap heavily with each other. Across the full fund set, 129 pairs hold portfolios that are more than half identical by weight. The closest pairs include GII and IGF in Infrastructure at 97% weight overlap, IBB and IBBQ in Biotech at 86%, and AUAU and GDX in Gold miners at 84%. In cases like these, holding both funds in a pair is close to holding one of them twice, and the two funds in a pair can carry different fees for that similar exposure, as with IBB at 0.44% against IBBQ at 0.19%.
Readers can look up any individual fund's in-index weight, active share, fee, and one-year return, along with its closest overlapping pairs, on ETFIQ's fund pages.