Recently, many people see that semiconductors are performing well, and they don’t know where they got the news, but they plan to hold long-term, dollar-cost average into SOXL, SNXX, MUU, and even TQ, hoping to make big money with leveraged ETFs.

This is completely unrealistic. Such people probably opened their US stock accounts only yesterday and are just starting to learn about investing.
Using backtesting to calculate returns, planning to make money with a strategy that never changes—this is destined to fail in Wall Street history. You can look up what a black swan event is and learn how Long-Term Capital Management (LTCM) went bankrupt
Many people see simple backtests done by others and feel like they have discovered some huge bug, like cheating the game.
This is a very naive, beautiful, and innocent idea, and it will be ruthlessly eliminated by the market sooner or later