A Warning to All Investors: The AI Trade Is Finally Breaking

1.Why you should care: the AI narrative is getting questioned

“I’m old enough to remember the dotcom bubble. The next 10% correction will be spurred by questions about whether the AI hype is overblown.”

That’s not fear-mongering — it’s history repeating. We’re already seeing headlines and alternative narratives pop up: reports about Alibaba building chips to fill the Nvidia void in China, nitpicking around Nvidia’s data center sales, and weakness in other AI/semiconductor names like AMD, Micron and Broadcom. When mega-cap names run up quickly, the market starts expecting perfection. Miss expectations even slightly and a pullback becomes possible.

2.Macro backdrop: Fed division and seasonal risk

Two big non-technical factors to keep in mind:

• Fed divide: The Fed is unusually split — some officials are more concerned about inflation, others more concerned about the labor market. Those are opposing incentives: worrying about inflation argues against cuts; worrying about jobs argues for cuts. No matter what the Fed does, it can be framed as “too much” or “not enough,” which fuels volatility.
• Seasonality: September and October have historically been the weakest months for stocks. In post-election years especially, we tend to see weakness through September into October. We’re already tracking similar to historical post-election patterns.

The Fed meets on September 17 and the consensus today is for a 25 bps cut. If the Fed waits long between cuts (5–12 months), history shows that near-term returns can be poor — so timing matters.

3.Technicals I’m watching (and why they matter)

VIX: a ready-to-pop setup

VIX has been downtrending all summer with only one spike around 22 on August 1. Historically, the VIX rarely travels into October without a significant spike. I expect VIX to come off its lows in the next couple months.

Notable flow: $14.5 million in VIX 31 calls expiring October 22 were bought recently. That’s a sizable bet on higher volatility into October.

SPY vs QQQ / XLK divergence

There’s clear divergence between SPY and QQQ: SPY made a marginal higher high, while QQQ failed to take out prior highs and shows a lower high. XLK (tech sector) looks like it may be forming a head-and-shoulders pattern, not confirmed yet, but possible.

• QQQ neckline to watch: ~560 (a confirmed close below and retest would validate a head-and-shoulders).
• SPY structure break level to watch for confirmation: close below ~632 (daily structural break).

On the SPY monthly chart there’s bearish RSI divergence: price made a higher high but RSI did not. If the monthly RSI doesn’t get above the prior level (~74.82), that divergence remains and increases the odds of a pullback.

Gold: consolidation ready to breakout

Gold futures have formed an ascending triangle since April and have tested resistance at ~3553 multiple times. If gold breaks and retests that level, the potential for a strong rally is real, especially if rate cuts push the dollar lower (low-90s or high-80s on the DXY would help gold).

4.How I’m preparing, my trading and hedging playbook

I don’t believe in being fully out of the market. My goal is preparedness: set probabilities, hedge cleverly, and be ready to buy into a pullback. Here’s what I’m doing right now.

Hedge: VIX call spreads (cheap, effective)

Instead of paying for outright VIX calls (which suffer heavy theta decay), I’m buying call spreads to hedge. Example I outlined:

• Buy VIX 20 call, sell VIX 25 call, expiring October 22
• Cost example: ~$118 (for the vertical spread)
• If VIX \> 25 at expiry the spread pays the $5 width or $500. Minus cost ($118) = ~$382 net profit (roughly 2–3x).

Why this works: it caps upside but is inexpensive and limits theta decay compared with buying single-leg calls.

Covered calls and selling premium

I don’t plan to sell my core, long-term holdings like Google or the Nvidia stock I bought when it was cheap. But when the market looks stagnant at highs, selling calls against shares is “free money” - generate income while retaining ownership. I’ve sold calls against Nvidia, Tesla, QQQ, AMD and others. That’s a low-risk way to monetize positions you’d otherwise hold.

Shorts: only on confirmed structure breaks

I’m not shorting the market outright at all-time highs unless we get a structural break and a clean retest. Shorting market tops has historically been a low-probability play unless the market confirms the breakdown. If XLK confirms a head-and-shoulders (close below neckline ~258 and a retest), I’ll consider puts or a debit put spread to target the prior consolidation around 242–243.

Example for XLK if bearish setup confirms: buy the 255 put, sell 250 put (a $5 wide spread) - pay ~$2 for $5 of intrinsic width = $3 potential net at expiration (risk-managed directional trade).

5.My buy list if we get a meaningful dip

I’m compiling a list of names I want to re-enter on weakness. If a correction happens I’ll be ready to buy high-quality, long-term winners and some leveraged plays where appropriate.

• SOXL - I entered at ~7, sold near 30. It was one of my biggest winners; I’ll look to re-enter on a pullback.
• Nvidia - huge long-term position. Key buy levels: prior highs near ~153; a 50% retracement from recent high is another historically good buy zone (example levels mentioned: ~150 then ~135 as next buy points if we sell off).
• Google (Alphabet) - I remain bullish. Google recently made a new all-time high and I still see substantial upside vs. business valuations.
• Robinhood - I’m considering adding if it dips; inclusion into the S&P 500 could spark a strong move (historical pattern with other S&P additions).
• BitMine / Immersion / BMR - I’m monitoring these crypto-related names (BMR being an Ethereum fund Tom Lee chairs) as potential midterm positions if crypto starts to break higher.

6.Practical rules I follow

• Investors: find long-term, quality companies and buy with a multi-year horizon. Avoid selling long-term winners based on short-term headlines.
• Traders: manage risk tightly, keep losses small and let winners run. You don’t have to be reckless to trade - use spreads, covered calls and defined-risk positions.
• Have a pre-made watchlist of stocks to buy on a correction. Being prepared beats trying to predict timing.

7.Final thoughts

We’re in a phase where headlines (AI, chips, Fed policy) can trigger outsized moves. That doesn’t mean selling everything, it means having a plan: hedges in place, covered calls to generate income, defined-risk bearish trades if structure breaks, and a buy list to deploy cash on meaningful pullbacks.

Volatility creates opportunity. The job of a trader/investor isn’t to predict the exact day of a drop. it’s to be ready for the probabilities and to act when the market confirms a setup. Stay disciplined, size your trades, and keep a watchlist ready.