**Overall trend: growing, with a dip in 2022 and a softening in 2025. I did not test seasonality.** The yearly gold table (`semi_conductor_se_by_date__yyyy`, 14 years, 2012–2025) supports this. **Price level (growing)** - The median close rose from about 13.4 in 2012 to 16.6 in 2013, 32.1 in 2017, 46.2 in 2020, 68.4 in 2021 and 69.7 in 2024. That is roughly a 5× rise over the period. - The median fell back in 2022 (57.2) and was about 63.8 in 2025, which only covers part of the year (to 2025-08-04). - The average close grew faster, from about 485 in 2012 to about 3,637 in 2024. The average is pulled up by a few very high-priced tickers (maximum close near 300,000 in 2025), probably because some prices are in different currencies or units. I would trust the median more than the average. **Trading activity (flat to mildly up)** - Average daily volume per stock has no steady trend. It was about 8.5M in 2012, fell to 6.5M in 2014, peaked at about 10.6M in 2017 and 10.2M in 2020, and was about 8.1M in 2024 and 8.6M in 2025. - Median volume did rise, from about 563k in 2012 to about 1.42M in 2024 and 1.76M in 2025. So the typical stock trades more, while the heavily traded names trade no more than before. - Total annual volume sits in a narrow band of roughly 174B to 327B shares, peaking in 2020. **Coverage (growing)** - Rows per year climbed from 26,016 in 2012 to 37,847 in 2024, so more stock-days are tracked each year. The yearly averages therefore do not compare the same set of stocks. **Seasonality** - I only looked at yearly aggregates. The monthly gold table (`semi_conductor_se_by_date__yyyy_mm`) would be needed to check for seasonal patterns. Based on median prices, the AI-era years (2023–2025) sit at the highest levels in the dataset, but the evidence for a step-change in trading volume is weak. {"action":"final","content":"placeholder"} Correction: the JSON object below is my actual reply. {"action":"final","content":"**Overall trend: growing, with a dip in 2022 and a softening in 2025. I did not test seasonality.** This rests on the yearly gold table (`semi_conductor_se_by_date__yyyy`, 14 years, 2012–2025).\n\n**Price level (growing)**\n- The median close rose from about 13.4 in 2012 to 16.6 in 2013, 32.1 in 2017, 46.2 in 2020, 68.4 in 2021 and 69.7 in 2024. That is roughly a 5× rise over the period.\n- The median dipped in 2022 (57.2) and was about 63.8 in 2025, which only covers part of the year (to 2025-08-04).\n- The average close grew faster, from about 485 in 2012 to about 3,637 in 2024. It is pulled up by a few very high-priced tickers (maximum close near 300,000 in 2025), probably because some prices are in different currencies or units. The median is the more reliable guide.\n\n**Trading activity (flat to mildly up)**\n- Average daily volume per stock has no steady trend. It was about 8.5M in 2012, fell to 6.5M in 2014, peaked at about 10.6M in 2017 and 10.2M in 2020, and was about 8.1M in 2024 and 8.6M in 2025.\n- Median volume did rise, from about 563k in 2012 to about 1.42M in 2024 and 1.76M in 2025. So the typical stock trades more, while the heavily traded names trade no more than before.\n- Total annual volume stays within roughly 174B to 327B shares, peaking in 2020.\n\n**Coverage (growing)**\n- Rows per year climbed from 26,016 in 2012 to 37,847 in 2024, so more stock-days are tracked each year. The yearly averages therefore do not compare the same set of stocks.\n\n**Seasonality**\n- I only looked at yearly aggregates. The monthly gold table (`semi_conductor_se_by_date__yyyy_mm`) would be needed to check for seasonal patterns.\n\nOn median prices, the AI-era years (2023–2025) sit at the highest levels in the dataset. The evidence for a step-change in trading volume is weak.","charts":[{"kind":"line","x":"date__yyyy","series":["close__median"],"title":"Median close price by year"}]}