Herman Jin
Formerly in Asia FICC at Goldman Sachs, Macro & Semiconductor Investing
The 3x leveraged semiconductor ETF SOXL will certainly go to zero within five years, because its daily rebalancing keeps adding exposure as prices rise and selling at lows as they fall. Leveraged ETFs go to zero even in the biggest bull markets. That answers @nongxiaodefa, who joked about going all in on SOXL and opening the mystery box in five years. x.com x.comÂ
SOXL holders hoping the bull run resumes more slowly misunderstand leveraged ETFs, which move fast in both directions. US leveraged semiconductor ETF assets went from $14B to $163B in three years and now sit at $100B, a 39% drawdown. Even after that, assets remain seven times where the trade started, so the bubble has not popped, only exhaled, and those trapped in 3x products face more of the same. x.comÂ
The 3x leveraged semiconductor ETF is exposed to a chip-stock reversal because leveraged ETF decay worsens as volatility rises: the bigger the bubble gets on the way up, the faster capital is destroyed on the way down. Market makers have not marked up future chip volatility enough, making way-out-of-the-money 2028 puts highly likely to triple, though 5x-plus returns would require nailing the timing. Unleveraged semiconductor funds carry no such decay. x.com x.comÂ
Semiconductors are exposure worth owning even through a 3x leveraged ETF, so SOXL is worth buying on a dip. High implied volatility comes with leverage rather than signaling the uncertainty it would in lower-quality names. The risk is a sharp drop, such as 40% in a week. x.com x.comÂ
War escalation and the midterms will keep investors from increasing gross or net semiconductor exposure. SOXX will take out its July lows before any sustained rally; against a suggested 5% dip and 50% rebound, the path is '5% down then 100% down.' x.com x.comÂ
SOXX's extreme selling sets up a semiconductor relief bounce next week before another leg lower. Per Heis, SOXX had its third-highest red-volume week and tied its longest nine-day losing streak; the four other top red-volume weeks averaged an 8.4% gain the following week. This is not a heavy bet: preserve capital, keep short-term setups small and accumulate core dips for the long term. x.comÂ
Leveraged semiconductor and Nasdaq 100 ETFs should beat a 3x position in the average individual name because market correlations are falling and dispersion is rising. Low correlation among basket members reduces the volatility drag that hurts levered ETFs, as seen when CPU stocks pumped while memory dumped. x.comÂ
Semiconductor earnings have outpaced prices, making the selloff look more like forced positioning than a broken growth case. Since January 2025, the SOX rose about 114% while reconstructed forward earnings rose about 198%; the estimate is approximate. Hedge funds sold global tech at a greater-than-three-standard-deviation pace in October 5-16, a pattern seen near earlier lows, though another sharp flush could come before a reversal. x.comÂ
Leveraged ETFs such as the 3x semiconductor bull fund should never be owned: it lost a third of its value in a single-day crash while individual chip stocks fell only 5%-10%. x.comÂ
Semiconductors could gain from desks buying back exposure sold ahead of leveraged-product rebalancing if markets recover even mildly. The expected selling flow was about $16B, and a strong signal also favored leveraged semiconductor exposure through SOXL and Nebius. There is no clear read on whether the market has bottomed over a longer horizon. x.comÂ
3x leveraged ETFs like the ProShares UltraPro QQQ and Direxion Semiconductor Bull 3x are not foolproof long-term holdings and would suffer badly in a dot-com or GFC-style decline. They target 3x daily returns, and the daily leverage reset compounds losses in volatile periods. The Nasdaq one launched in 2010 at the start of a ZIRP bull market; backtested to the late 1990s it would still be underwater. A 2x software ETF holding Adobe, Microsoft and Salesforce was discontinued this April. x.comÂ
