They're Shorting Into All-Time Highs. The Data Says That's a Problem.

ETF flows, COT positioning, and career risk all pointing the same direction. Just not the direction the shorts want.

Six hundred million dollars into SOXS. Let that number sit for a second.

While semiconductors rip to new highs, traders are pouring money into the leveraged short semiconductor ETF. At the same time, they’re selling out of TQQQ and SOXL. Buying the short, selling the long. Into a rally.

Three datasets tell the same story this week:

ETF flows show aggressive inflows into inverse vehicles (SOXS, METU, SSO) and outflows from leveraged long products (TQQQ, SOXL). People are betting against this tape with real money.

COT futures data shows large speculators still net short ES contracts. They got slightly more short last week. On the Nasdaq side, they’re not adding shorts but they’re selling longs, getting flatter as price rises. Bitcoin is the same pattern. As it breaks out of the 60K-80K base, speculators are selling into the move.

Career risk is the part nobody models. End of June is quarter-end. Imagine sitting across from an investor and explaining you not only missed the rally but bought the short instruments. You’re flat or down in a year the market ripped. At some point, the math stops mattering and the career pressure forces action. That action is buying high-beta names to catch up. Fast.

Now, some of this is just hedging. If you own semiconductors and think they’re overbought short-term, buying SOXS lets you hedge without selling the position and triggering capital gains. That’s rational. But the COT data showing net short speculators on ES is harder to explain away as hedging. That looks like top-picking.

Here’s what it means practically. On any downtick, there’s underlying demand waiting. The SOXS holders will take profits on a 5-10% semi pullback. The ES shorts will cover. The Nasdaq sellers will re-enter. That creates a bid underneath corrections that doesn’t exist in normal positioning environments.

The systems I run at Stats Edge don’t try to predict when that bid shows up. They’re positioned for it structurally. The pullback and mean reversion systems are designed for exactly this kind of regime. They’ve also had drawdowns in it. I publish those in The Drawdown Memo when they happen. No hiding.

The free 25-Year Backtest PDF at letters.statsedgetrading.com walks through how I test across regimes like this one, including the positioning environments that broke most strategies. This is the framework on pages 2-3, applied to what’s happening right now.

For real-time alerts with entries, stops, and sizing across all three systems, that’s Stats Edge Pro at $149/month with a 30-day money-back guarantee.

— Michael Nauss, CMT, CAIA, CDMS