The Russell 2000 closed at a record 3,052.85 on Thursday, clearing a previous peak of 3,036.98 that had stood since its last run at the highs. On Friday it edged up again, roughly a quarter of a percent, and finished the week at another record.
That leaves the small-cap index up about 23% for 2026 — roughly nine percentage points ahead of the S&P 500 and comfortably clear of the Nasdaq 100’s 19%. For the first sustained stretch since the brief rotation of July 2024, the 2,000 smallest names in the American market are beating the largest ones, and not by a rounding error.
Here is the part that should stop you. The same week produced the weakest retail sales print in more than a year: total sales fell 0.6% in July to $763.6 billion against forecasts for a 0.1% gain, with online spending down 2% and auto dealers down 1.8%. Strip out autos and gasoline and the picture is softer but still negative, at −0.2%. Small caps sell to the American consumer far more directly than mega-cap tech does. They set a record on the week the American consumer pulled back.
Our take: This is not a growth trade dressed up as a value trade. It is a rate trade. Small caps carry more floating-rate debt, refinance more often, and earn a larger share of revenue at home — which makes them the most leveraged expression in the market of one specific view: that the Fed is finished hiking. July’s cooler CPI and PPI made that view cheap to hold. The retail sales miss made it cheaper still, because weak demand argues against further tightening. The rally and the bad data are not in tension. They are the same bet, read from opposite ends.
What is actually driving it
The valuation gap. Small caps entered 2026 at roughly 18x forward earnings against about 26x for the S&P 500 — a discount near a 30-year extreme. Even after a 23% run, the main small-cap ETF proxy still trades at something like a 34% earnings-multiple discount to large caps. The gap narrowed by rallying into it, not by closing it.
Domestic revenue. With tariffs, export controls and currency swings all live variables, a company that earns its money inside the United States has fewer ways to be surprised. That has been a quiet bid under the whole cohort since spring.
Rate expectations. The Fed on hold through 2026 removes the single biggest drag on balance sheets that borrow short. Odds of a September hike fell after last week’s inflation prints.
The uncomfortable part
Long rates have not cooperated. The 10-year Treasury sat near 4.65% on Friday after testing 19-month highs around 4.75% earlier in the week. Small caps are supposed to hate that. They rallied anyway, which means the market is pricing a specific sequence: front-end relief arrives before back-end fiscal pressure bites.
That sequence can hold. It can also break in one print. And the index itself is a blunt instrument — a meaningful slice of Russell 2000 constituents earn nothing at all, which is why headline multiples for the index vary wildly depending on whether the calculation includes negative earners. Beating the Nasdaq 100 for eight months does not make a cohort of unprofitable borrowers into a quality trade.
What to watch
- The August retail sales print. One soft month is noise; two is a trend, and the second one would reprice the domestic-revenue argument that half this rally rests on.
- The 10-year. A move back through 4.75% with no small-cap damage would be a genuine regime signal. A move through it with damage tells you the rotation was always a duration bet.
- Breadth inside the index. If the gains keep coming from the profitable third, the rotation has legs. If they broaden into the loss-makers, it is late-cycle behaviour.
- Index migration. Every small-cap success story eventually graduates — the mechanics of a company moving up to the S&P 500 quietly drain the Russell of its winners.
For the week the Dow fell about 0.6%, the S&P 500 rose roughly 0.4% after clearing 7,800 for the first time, and the Nasdaq was close to flat. The index that led is the one nobody spent three years talking about.