[Mou Weekly] Run for your life? Or buy the dip? (Week 44, 2026 | Issue #294)

Last Friday, the Nasdaq fell 2.86%, marking its worst day since June; for the full month of October, it dropped 5.2%, the weakest performance since March 2025. Since its peak on Oct 6, the Nasdaq 100 has seen an intraday drawdown of over 10%, the third 10% correction this year; the S&P 500 saw a more modest pullback of 3.1%. Intel plunged 28.6% in a single day, its steepest decline since 1974.

Chart 1: Three corrections for the Nasdaq 100 and S&P 500 this year, plotted on the same scale (closing price, % distance from previous high)

Let's first lay out the facts clearly. Based on closing prices, the Nasdaq 100 has corrected three times this year: from Jan 28 to Mar 30, it fell 11.8% over 42 trading days; from Jun 2 to Jul 29, it dropped 11.3% in 39 trading days; and this time, from Oct 6 to last Friday, it declined 9.1% in just 18 trading days. Ranked third in depth but first in speed. The S&P 500 saw drawdowns of 9.1%, 4.5%, and 3.1% respectively across these three episodes.The divergence in drawdown magnitude between the Nasdaq 100 and S&P 500 was widest this time. In March, all stocks fell together; this time, AI names are falling while others are rising.

Two Types of People Exiting

Those who exited last Friday fall into two categories.

The first group believes a crash is imminent and has liquidated positions to wait on the sidelines. Before Friday's open, the market-implied probability of an "S&P gap up" had fallen to just 14%. The VIX approached 24 intraday, closing at 19.4. Some investors spent roughly $2.8 million buying VIX call options, betting that the VIX would remain above 23 through mid-November. An October survey of fund managers showed tech overweighting hitting its highest level since August 2020. The most crowded trades tend to run fastest when they fall.

The second group is forced out due to excessive leverage. Hedge funds reduced their total exposure by 7.8% from peak levels—the largest cut on record for Morgan Stanley dating back to 2016; net leverage dropped to 48.6%, a yearly low. The pace of shorting was five times that of long positioning. Three-times leveraged semiconductor ETFs plummeted 19.8% in a single day on Friday. These investors aren't necessarily bearish; they simply have no choice.

The reasons for selling differ between the two groups, but the exit prices are identical. The first group sells their conviction; the second sells their margin. Conviction can be revised; margin cannot.

A fair word for those currently standing on the platform: their position is advantageous. They hold cash, have no unrealized drawdowns on their books, and any earnings report next week won't hit them. However, a small position doesn't guarantee capturing the rebound rally. Consider the speed of the two V-shaped reversals earlier this year: on Mar 30, the S&P 500 reclaimed its previous high in 11 trading days, gaining 10.7%, with 6.9% of that occurring in the first seven sessions; on Jul 29, the S&P 500 set a new high in just four days, up 5.7%, while the Nasdaq 100 surged 9.3% over those same four days. On both bottom-fishing days, the Nasdaq 100 jumped 3.4%.While the platform offers a safe vantage point, it doesn't control when the train departs. Getting off is easy; re-boarding before the wheels turn is hard.

Chart 2: Two V-shaped reversals this year, showing the trend for the first 10 trading days post-bottom (bottom day = 100)

Illustration: Two people on the platform

Money is Shuffling Seats, Not Exiting

To determine if this pullback marks the start of a crash, look at where the capital is flowing.

On Friday, the S&P 500 fell 1%, while the equal-weight S&P remained flat. Healthcare rose 1.5%, energy gained 1.4%, and financials edged up 0.8%. On Thursday, when Google dropped 12%, over 350 S&P constituents were actually rising. For the full month of October, the equal-weight S&P gained 4.9%, finishing the year up 13.5%, outperforming the market-cap weighted index's 11.4%. Apple's market cap surpassed Nvidia's, driven by its restraint in spending on data centers.

It is also clear who is falling. Amazon reported negative free cash flow of $42.4 billion over the past 12 months, with Q3 capex at $77 billion compared to $35 billion year ago; Google posted negative free cash flow for two consecutive quarters, a first since its IPO. While S&P 500 Q3 earnings growth still stood at 30%, the issue lies with cash flow, not profitability. The stocks falling aren't money-makers; they are the biggest spenders.

Safe-haven assets are functioning normally. The 10-year Treasury yield retreated to 4.93%, a seven-week low, as capital rotated out of AI stocks and into Treasuries, healthcare, and energy. Options market readings show a total put/call ratio of 0.81 and a VVIX of 98, indicating hedging activity but no panic.

Chart 3: Sector performance in October, plus the relative strength of the equal-weight S&P vs. market-cap weighted S&P (Oct 1 = 100)

A crash looks like all assets falling together, even Treasuries; a correction looks like money moving from one pocket to another. Last Friday was the latter.

What Makes This Different From The Previous Two

At this point, we shouldn't speak too definitively. This correction features three elements absent in the March and July episodes.

First, the Fed is in a hiking cycle. At the Oct 28 meeting, 10 votes held rates steady while 2 called for an immediate hike; the probability of a December rate increase has now reached 85%. September non-farm payrolls added only 29,000 jobs, and Brent crude sits above $110. The hammer driving prices down and the hand catching them are the same. AI's bill is also being priced by the bond market: cloud vendors are expected to issue $182 billion in debt this year, compared to just $17 billion two years ago.

Second, earnings growth is peaking. Goldman Sachs forecasts S&P 500 earnings growth slowing from 29% this year to 18% next year; Goldman's own risk appetite indicator turned negative for the first time since March. The 30% earnings growth rate has been the foundation of this bull market, and that foundation will shrink next year.

Third, this downturn was triggered by the earnings reports of the AI leaders themselves. March was driven by war and oil prices; July by valuation concerns; this time, it's the reports from Google, Amazon, and Intel. The holes dug previously were made by outsiders; this hole was dug by passengers on the train. The market is voting against AI capex via earnings reports for the first time, and this sentiment won't end with a single bounce.

My View And Conditions For Being Wrong

My subjective view: this is another buy-the-dip opportunity in US equities this year. It may not be correct, and it is heavily subjective, but the basis for this judgment is outlined above: money is rotating, earnings fundamentals are intact, and safe havens are working.

Judgments are guesses; conditions for being wrong are fixed. If two of the following five signals appear, my above judgment is incorrect:

• The S&P 500 weekly close breaks below 7,500, accompanied by a breakdown in the equal-weight S&P, signaling rotation failure;
• The VIX stands above 30, with the term structure fully inverted;
• Credit default swaps for cloud vendors and high-yield bond spreads widen simultaneously;
• Trend-following fund signals flip to short, with Goldman estimating they still hold ~$93 billion in unutilized equity longs;
• Stocks and bonds fall together, the 10-year yield rises while equities drop, reverting to 2022 correlations.

Regarding position sizing, adhering to the old rules of event weeks: I am compressing leverage on my personal account to below 1.1x. Core stock and Call positions remain untouched, while most are hedged with Sell Calls. Buying isn't about guessing the bottom; it's about waiting for none of the five signals above to trigger. Specific operations are detailed on the website's member page.

[Next Week Outlook]

This week has no empty days. After-hours on Monday: Palantir and Apple; Tuesday: Midterm elections and AMD; Wednesday: Qualcomm and ARM; Friday: October Non-Farm Payrolls, expecting 55,000 vs. prior 29,000. Nvidia's earnings on Nov 18 will serve as the final verdict on this correction.

A one-sentence recap of last week: the three cloud vendors' earnings validated AI revenue, but the market's dissatisfaction wasn't with revenue—it was with the bill. On this week's table, Palantir and AMD represent the other end of the AI chain: one collecting money, the other selling shovels; Non-Farm Payrolls is the final card for a December rate hike.

Voting opens Tuesday, earnings come Monday and Wednesday, and Non-Farm Payrolls drops Friday. Since 1934, the Dow Jones has averaged a 2.8% gain in the eight trading days surrounding midterm elections, but this statistic assumes no earnings reports during those eight days force a change of heart.

Watch two things for entry points: the equal-weight S&P holds support, and the VIX stays below 30. If both conditions hold, the pulled-back AI names present a third boarding opportunity this year.

Those fleeing look at how much they've lost; those buying the dip look at what is falling.

(The above represents personal views and does not constitute investment advice.)