Small Caps Enter September: The Rally Meets the Warsh Test (September 1, 2026)

The scoreboard still reads like a small-cap decade compressed into eight months. The Russell 2000 closed its best first half since 1991, printed a record 3,052.85 on August 14, and at the peak sat up roughly 23% for the year — about nine points ahead of the S&P 500's 13.9% and comfortably past the Nasdaq 100's 19%. Semiconductors did disproportionate lifting: by one count, 16 of the index's top 50 performers this year are chip-related, with names like Aehr Test Systems (+518%) riding the AI supply chain down-market.

Then the tape went quiet. In the last full week of August the Russell 2000 slipped 1.49% while the S&P 500 added 0.50%, and small caps closed the month at 2,956 — about 3% below the record — even as the large-cap benchmarks logged a positive August (S&P 500 +3.0%, Nasdaq +4.0%). A leadership index that outruns everything for eight months doesn't stall by accident. It stalls in front of a calendar.

That calendar belongs to Kevin Warsh. Roughly 100 days into the job, the Fed chair used Jackson Hole to warn that the central bank may have “work to do” on inflation, to call 2% PCE a “firm, fixed target,” and to pointedly refuse a pre-commitment on September — “committed to a discipline, not to a decision.” This is 2026's inversion: the live question isn't when the Fed cuts, it's whether it hikes. Futures put the odds of a September hike near 55% in mid-August before July's producer-price print cooled to 4.7% year-over-year (from 5.5%) and trimmed them toward 35%. The inflation trade hasn't gone anywhere, though: WTI crude pushed to $85.76 into month-end on U.S.–Iran tensions (up 2.8% for August), gold rose 9% in a month to $4,481, the 10-year sits at 4.75%, and the 30-year at 5.25%.

Small caps care more than anyone. Russell 2000 balance sheets lean on floating-rate and short-maturity debt, so a hike — or even a hawkish hold with a 5%-plus long end — flows into interest expense in quarters, not years. The same sensitivity that supercharged this rally when hike odds fell is exactly what unwinds it if Warsh moves.

The cushion is real, which is why dips keep getting bought. Small caps still trade around 19x earnings versus roughly 29x for large caps — a 34.5% discount, and 58% on price-to-book — while consensus now models 38% Russell 2000 earnings growth for 2026, revised up from 23% at the start of the year. “It's both a valuation catch-up story and a fundamental story,” as Alger's Amy Zhang framed it this summer. Cheap, with accelerating earnings, into a coin-flip Fed — that's the whole September argument in one sentence.

The Tape:

• Russell 2000: 2,956 at the August close — record 3,052.85 (Aug 14); best first half since 1991.

• Final August week: Russell 2000 −1.49% vs. S&P 500 +0.50%; August overall: S&P +3.0%, Nasdaq +4.0%.

• The inflation trade: 10-yr 4.75%, 30-yr 5.25%; WTI $85.76; gold $4,481 (+9% in August).

The Setup:

• Sept 4: August payrolls — the release Warsh flagged as an input to September's decision.

• PCE inflation lands ahead of the Sept 15–16 FOMC; hike odds near 35% and headline-sensitive.

• Levels: the 2,956 close vs. the 3,052.85 record — a 3% band that resolves with the Fed.

Bull: A 34.5% valuation discount plus 38% consensus earnings growth means dips get bought while the label still says expansion.

Bear: A Warsh hike — or a 5.25% long end that sticks — hits floating-rate borrowers first, and small caps are the floating-rate index.

The Screen's read: Treat the next ten sessions as a data gauntlet, not a trend change. Above the August lows, pullbacks toward the low 2,900s look like rotation inside a leadership run; a hawkish surprise on the 16th is the scenario that turns rotation into distribution. Position sizing beats prediction here.

SCN editorial — independent commentary for information only, not investment advice.