September Is Historically the Worst Month for Stocks. A Pattern From 2000 Says This Could Happen Next.

The article warns that the S&P 500’s valuation looks historically rich, with the Shiller CAPE at ~41 versus the long-run average ~18 (near the dot-com-era peak of 44), implying vulnerability to a potential sell-off. It ties this setup to multiple macro overhangs including sticky inflation, rising energy prices, hawkish bank signals, and high U.S. Treasury yields, while noting risks such as an added geopolitical shock (Iran). Despite the caution, it argues investors should avoid panic selling and stay invested in quality companies.

Analysis

The actionable signal is not “September” by itself; it’s that investors are starting the month with unusually little valuation cushion. In that setup, even a garden-variety de-rating can overwhelm single-digit earnings growth, and the first casualties are the longest-duration names tied to a continued multiple expansion story — especially NVDA and the rest of the AI complex.

Near term, the risk is mechanical rather than macro-philosophical: crowded positioning, lower summer liquidity, and systematic de-risking can turn a modest catalyst into a self-reinforcing drawdown. If Treasury yields stay elevated, the equity risk premium on megacap growth compresses further, which should pressure QQQ relative to defensives and value before it shows up in the broad index.

Contrarian view: the market may be overpricing the odds of an outright air pocket and underpricing the ability of large-cap earnings to “grow out” of expensive valuations if nominal GDP remains firm. What would falsify the bearish setup is a clear break lower in real yields or an earnings season showing AI capex translating into faster revenue monetization; that would revive the premium on leaders like NVDA and keep the index pinned despite the scary chart.

Key Decisions for Investors

• Buy 4-8 week QQQ put spreads into any strength; use this as a tactical hedge against September de-rating rather than a crash bet. Risk/reward is favorable if implied vol stays contained and yields remain sticky; invalidate if QQQ reclaims recent highs and 10Y yields roll over.
• Rotate part of growth exposure from NVDA to NFLX on a relative basis for the next 1-3 months. NFLX has a less duration-sensitive multiple and should hold up better if the market pays for cash flow visibility instead of long-dated AI promise.
• Pair trade long XLP or XLU / short QQQ into month-end as a cleaner expression of multiple compression than an outright index short. This should work best if breadth narrows and systematic funds de-risk; cover if the market breadth expands materially or yields fall.
• For existing NVDA longs, sell upside calls or trim ahead of the next CPI/Fed window rather than waiting for a drawdown. NVDA is the highest beta to a valuation reset; the trade breaks if the stock holds above its 50-day through the next macro print.