September’s setup is a volatility event around the 15th-18th FOMC, BOJ, ECB and BoE cluster, with QQQ, SPX and crypto still in rotation rather than a clean risk-off break. The core trade calls stay selective: bullish BTC over gold, bullish TSLA, bullish MSTR/STRC, bullish MU on HBM scarcity, but bearish near-term NVDA and wary of chasing extended moves in SK Hynix, UMAC and other spike names. A second theme is mean reversion: overbought pumps are treated as sell zones, while crushed charts like copper, ISRG and a final unnamed stock are framed as potential bottoms only if catalysts and earnings repair the damage.
Volatility is only a little elevated now, but the real spike risk is the September central-bank cluster: FOMC on Sep. 15-16, Bank of Japan on Sep. 17-18, plus ECB and BoE, all flagged as potential VIX catalysts.
QQQ just printed a sell signal last week, slipped a bit, and is expected to turn back up; it is not oversold or overbought, and its next move is tied to NVDA, TSLA and MU.
SPX had a mild IDSS dip and is framed as a weak pullback, not a washout, while an “I13” divergence showed crypto and other non-traditional assets ripping as the S&P was down.
BTC printed a sell signal after a rampage, with resistance at 80k-82k and the top/bottom indicator turned down there; 40k and 50k are dismissed as non-events, 60k is possible but less likely, and 90k is preferred first if ETFs keep buying and Michael Saylor keeps buying.
BTC seasonality is a September drag: the month is described as normally BTC’s worst, typically down 5%.
SOL is boxed by resistance at 102, 103 and 112 while consolidating between 100 and 104; ETH has resistance near 2,481, a buy signal at 2,390, and a 430 target within a year, with price around 2,49 and sitting at a kill zone that could turn.
BTC outperformed gold from February into late April/early May, then gold took over; the stance stays long BTC over gold, with rotation back only if BTC reaches roughly 39-40 ounces of gold per BTC.
TSLA accumulated from 300 to a buy signal around 312, then flipped to “up up only”; the target stays 380, the 200-day moving average is 399, and Q4 seasonality is framed as strong.
SpaceX stayed unfazed after a prior unlock and another 7% unlock; the buy-add zone is under 135, last time it did not dip below 131, and the pair-trade rule is Blue = buy TSLA, Orange = buy SpaceX, with a reversal only if a big Elon Web Services compute deal appears.
NVDA printed a sell signal exactly at resistance, with 228 as the key line; trend is turning down, so no loading up here, though hedgers may still hedge because it is described as cheap at these levels.
MSTR is tied to BTC plus Saylor ATM selling to refill STRC; the trade range is 80s to 130-135, even 145, with a bounce off the 200-day moving average, a clean chart and a buy signal despite BTC itself looking less bullish.
STRC is being pushed back to 100, trading around 98-97-96 with an uptrend; the mechanism is Saylor selling MSTR to support STRC, restore confidence, trigger a 12% return flywheel and then recycle into BTC.
CLSK ran from 10.80 to nearly 14 and got sold quickly as overbought; 35% of the balance sheet is BTC, and the bull case is miners no longer having to sell their BTC hoard because AI/HPC revenue can pay the bills.
MU is described as 2.06 standard deviations overbought on sales-meter mean reversion, but HBM supply is said to be under 10 days worldwide, keeping the chart bullish after a break above level five; targets are 1,500 and 1,600 by end-2026 or sometime in 2027.
SK Hynix is called a blinder because supply is sold out through 2028; the Korean ticker 000660 turned down around 1.5 million won, then a buy signal hit and it shot up, while the ADR moved from 156 a few days ago to 199.87 today after being up another $13 intraday.
SK Hynix is treated as a do-not-chase name after a huge move: mean-reversion tools showed a red dot with a 72% chance of falling, June 26 was cited as the last big mean-reversion spike near an all-time high around 3 million won, and a proposed alert sits at 14.96 Korean won on level three ATR.
ALAB flipped bullish with the orange snake turning blue, got a buy signal, and is expected back to 500 by year-end or early next year; connectivity plays tied to the AI revolution are the holding context.
AMD is boxed between 457 and 580 and already halfway through the range at 521, so adding here is called no-man’s land; a return to 580 is possible but uncertain.
PLTR missed the expected rollover at 169/160, ran to 189, then rolled over after nearly 27 days; the hedge stays on for a fall back toward 158-160, while 200 is still viewed as reachable in 6-12 months.
UMAC is extremely strong but overbought and rolling over; there is no daily buy signal, only a 12-hour signal, and the guidance is to wait unless already in from the teens.
COSTS/Comfort Systems is framed as an industrial services winner for data centers with Houston roots; despite small profits, the breakout since April 2026 is expected to carry it back to 2,000 within 6 months unless earnings disappoint.
POET is intriguing as an adjacent AI name because it has turned profitable, has small debt and is very oversold, but no buy signal is in place yet even though a 100% surprise upside remains possible.
NKE is dismissed: dead, shrinking, no growth, and not worth touching despite the cheap look and possible accounting oddity; it is explicitly not a sports buy.
Copper is turning back up after taking a hit: it is very oversold, has a buy signal and trend turn, and is expected to keep making new all-time highs until 2030; if it ever revisits the 200-day moving average, the call is to buy leaps hand over fist.
ISRG is in a bad downtrend from about 600 to 350 despite strong medical-robotics leadership and very good financials; earnings beats of 17-18% on 12% revenue growth are not enough, and a blowout would at least target the 200-day moving average around 460 before caution returns.
The final unnamed stock has an ugly chart, is making money now after a prior 90% earnings miss, has good revenue but an earnings mess and a big market cap relative to sales; it needs profitability to return before the stock turns, with an alert set for another drop and a deeper fundamental review if that level breaks.