Page One · Markets · IN04

A Record 7,818.93 Crossed the Line We Named as Our Falsifier. We Were Wrong on Timing — and We’re Still Not Chasing

Nvidia is up 4.5% in a week, Meta 24% since Aug. 13, and the Magnificent Seven closed at a record roughly $25 trillion. The index crossed our line and the average stock didn’t come along — so we say wrong on timing, and keep the reserve on its dates.

By Sean Anees Saifi · Capital Wealth · Published Wednesday, October 7, 2026 · Source: The Wall Street Journal, Wednesday, October 7, 2026, whose market figures are the Tuesday, October 6 close (Page One)

Key Points

The S&P 500 rose 0.58% Tuesday to a record 7,818.93; the Nasdaq composite added 0.45% to 27,599.79, also a record; the Dow gained 253.38 points, or 0.49%, to 51,521.28. The VIX closed at 15.01.
The leaders are the usual few, the Journal’s Hannah Erin Lang reports: Nvidia (NVDA) is up 4.5% in the past week to an all-time high, Meta Platforms (META) 24% since Aug. 13, and the Magnificent Seven’s combined value closed at a record of roughly $25 trillion.
Breadth is thin: by Tuesday’s close, fewer than half of S&P 500 members were trading above their own 200-day moving average, and that share has shrunk steadily since August. On the NYSE, 134 stocks hit new 52-week lows against 50 new highs, the Russell 2000 slipped 0.59%, and the equal-weight S&P 500 trails the cap-weighted index over the past month.
Rates eased but stayed high: the 10-year Treasury yield fell 0.04 point to 5.27%, a day after settling at 5.310% and touching 5.349%, a 24-year intraday high. Odds of an October Fed hike slid from around 70% to around 25% after Williams and Jefferson spoke — and lower still after Friday’s jobs report, Nick Timiraos reports.
House ledger: the September letter named a record close above 7,798.99 before Oct. 2 as the falsifier of its fall caution. The Oct. 2 close was 7,722.72; Tuesday’s record came four days after the deadline. By the rule written in advance, we grade it wrong on timing — and nothing is bought on the strength of a view.

7,818.93

S&P 500 record close Tuesday, up 0.58%

$25 trillion

Magnificent Seven combined market value, a record Tuesday

134

NYSE new 52-week lows Tuesday vs. 50 new highs

5.27%

10-year Treasury yield Tuesday, down from 5.310% Monday

A lone hiker stands on top of a tall, narrow rock pinnacle that rises out of a sea of fog, with the sun low on the horizon casting warm light on the rock and cool blue tones across the clouds below. — The index set a record with fewer than half its members above their own 200-day average — a record carried by a few, not shared by most.

In one line: The S&P 500’s record 7,818.93 crossed the line the desk said would prove its fall caution wrong — so we say it, wrong on timing — but a record carried by a handful of AI names while most stocks sag is a reason to check concentration, not to chase.

Let’s get the uncomfortable part over with. In September we wrote that if the S&P 500 printed a record close above 7,798.99 before Oct. 2, our fall caution was wrong on timing and we’d say so in print. It didn’t — the Oct. 2 close was 7,722.72. Then on Tuesday the index closed at 7,818.93, a record, four days past our deadline. A rule you only honor when it flatters you isn’t a rule. So: we were wrong on timing. Here’s the record, honestly reported, and here’s why we still aren’t buying it.

The Journal’s Hannah Erin Lang puts it plainly: one group of stocks is doing the lifting. The S&P 500 rose 0.58% and the Nasdaq composite 0.45%, both to records, and the rally hasn’t flinched at a Fed hike — the first in three years — or at an oil price that a months-long war drove to $100 a barrel. Nvidia (NVDA), the world’s most valuable company, is up 4.5% in a week to a new high. Meta Platforms (META) has gained 24% since Aug. 13, the date of the S&P’s previous record. The Magnificent Seven — the four hyperscalers plus Tesla (TSLA), Nvidia and Apple (AAPL) — closed Tuesday worth roughly $25 trillion, a record, up from $24.8 trillion the day before.

A record most stocks missed

Then there’s everybody else. Healthcare, banks, staples, small caps and the Dow’s blue chips are all losing ground, Lang writes, and by Tuesday’s close fewer than half of the S&P 500’s members sat above their own 200-day moving average — a share that has been shrinking since August. Tuesday’s own tables say the same thing: 134 NYSE stocks made new 52-week lows against 50 new highs, the Russell 2000 fell 0.59% while the big index rose, and the equal-weight S&P 500 is trailing the cap-weighted one over the past month. Truist’s Keith Lerner put the mood in one line: “Tech is almost seen as defensive in some ways.” Investors aren’t selling it; they’re hiding in it.

Bonds gave the rally a hand. The 10-year yield eased 0.04 point to 5.27% after Monday’s 5.310% settle and a 5.349% intraday print, a 24-year high. And the Fed’s two lieutenants, John Williams and Philip Jefferson, talked October off the table — Nick Timiraos has the odds of a hike this month falling from around 70% to around 25%, and lower after Friday’s jobs report, with December now the consensus. Lower hike odds, a 10-year still above 5%, an index at a record. That’s the week.

Our read

Investments/Risk (IN04), and a line for the ledger first. Our September caution had a falsifier, the falsifier printed four days late, and the rule we wrote on Sept. 26 says what happens next: we were wrong on timing, we say so, and the week-of-the-vote purchase stands. What the rule doesn’t say is chase. The North Star books add on dates and drawdowns — the September letter’s levels and the week of Nov. 3 — never on a Tuesday because the index felt good. The letter’s conditions for new equity money still fail: core inflation ran hot in August and the Fed raised in September. Nothing new is bought.

Now the planning point, which matters more than the mea culpa. A record set by seven companies while most stocks sag is a concentration reading, not a buy signal. If you own an S&P 500 fund, you own more of those seven than you did on Aug. 13, whether you chose to or not; if you also hold Nvidia, Meta or a tech fund on the side, the overlap has quietly grown. We hold the names — NVDA in 31 model books, META in 20, AAPL in 24 — at weight, nothing added, and keep the equal-weight RSP in 2 books as the honest scoreboard. Pull up the statement and ask what share of your equity sits in those seven. If the answer surprises you, that’s the finding — and a rebalance you decide on a sunny day beats one the market decides for you in the rain.

What It Means For Your Portfolio

Watch — wrong on timing; check concentration, don’t chase

No portfolio action: the record crossed our falsifier line and we say so, but with fewer than half of S&P 500 stocks above their 200-day average, the move is a prompt to measure your Magnificent Seven exposure, not to add to it.

General planning principles, not advice for anyone in particular. Add up what a cap-weighted index fund, any single tech names and any growth fund give you in the same seven stocks; many investors find the total is well above what they’d choose on purpose. Rebalance to the weight you’d pick fresh, in tax-advantaged accounts first.

Keep the buying rules you wrote when you were calm: dates on the calendar, drawdown levels, contributions on schedule. A record that arrives with weak breadth and a 10-year near 5.27% is no reason to tear them up — and no reason to sell what you own, either.

Book a 15-Minute Review → Back to Edition No. 180 →