7 Russell 2000 Stocks That Led the First Half of 2026

The Russell 2000 spent the first half of 2026 doing something it had not done in years: leading. The small-cap index pushed to record highs to end June, capping one of its strongest first halves in more than three decades.

The index-level number, though, hides a split that matters for anyone trying to read the back half of the year. The names that actually drove the move were not spread evenly across the index, much like we saw in recent years as a handful of stocks in the S&P 500 drove the majority of indexed gains.

And, like the large- and mega-cap indices, those names clustered in the handful of themes the market spent the spring falling for: AI power and the chips feeding it, defense and space, critical minerals, and the data plumbing underneath the AI buildout.

Most ran to all-time highs somewhere between January and late May/early June, and then a chunk of them gave back 30% to 50% in the late-June selloff that hit the market’s hottest corners hardest before rebounding to end H1. Still, “led the first half” and “still leading” turned out to be two different lists.

This screen collects seven of those standouts and asks the one question worth asking now that the second half has started: does the move keep going, or was the spring the top?

A couple of these sit at the larger end of what still counts as small-cap, since June’s index reconstitution lifted the Russell 2000’s ceiling to roughly $5.7B. The megacaps like Nvidia carried the AI headlines all spring, but these were the small-caps actually re-rating alongside them.

What separates them is what happens next.

Navitas rode the single hottest trade in small-cap semiconductors to an all-time high of $34.17 on June 3, then handed back more than half (trading at just around $14 today).

The company designs gallium nitride and silicon carbide power chips, the components that convert and manage electricity, and it has spent the past year pivoting hard away from phone chargers toward the high-power markets feeding AI data centers and the grid.

The kickoff catalyst was a collaboration with Nvidia to co-develop 800-volt power architectures for AI server racks, a design shift that strips out a conversion stage and packs more compute into the same footprint.

The pivot toward industrial-level power is showing up in the numbers: first-quarter revenue of $8.6 million grew 18% sequentially, with high-power markets now the majority of sales and up about 35% year over year.

The problem is the gap between that revenue and the valuation. Navitas carries a market cap near $3.5B against single-digit-million quarterly sales, it is not yet profitable, and on July 7 Wolfspeed filed a patent-infringement suit targeting its core GaN and SiC lines. Navitas called the claims baseless and said it will fight.

The Nvidia win is real and the pivot is working, but the stock is priced for that pivot to compound without a stumble, and the June pullback suggests the market is already recalibrating. The second half turns on whether the AI-rack ramp shows up in the July 27 quarter and whether the litigation becomes a distraction.

Aehr’s chart tells two stories at once. It hit an all-time high above $120 in early June, then fell about 47%, and both moves trace back to the same tension in its business.

The company makes test and burn-in equipment that stresses semiconductors before they ship, increasingly for the AI processors and silicon photonics parts going into data centers.

Reported revenue is falling while the order book sets records. Third-quarter results showed revenue of $10.3 million, down 44% from a year earlier, and a GAAP loss. In the same quarter, bookings hit $37.2 million for a book-to-bill above 3.5, and effective backlog reached a record $50.9 million. In April the company landed a record $41 million follow-on order from its lead hyperscale customer, its largest ever, with deliveries starting in its fiscal 2027.

The bull case is that bookings lead revenue, and a second-half booking haul above $92 million plus the hyperscaler ramp converts into real growth next fiscal year.

The bear case is that the June high priced the conversion in before it happened, leaving the stock to wait for revenue to catch up to the backlog. There is little ambiguity about demand and a lot about timing.

The second half of 2026 is a show-me stretch: either the backlog starts turning into reported revenue, or the gap between the two keeps the stock stuck.

The California governor primary is live right now. So is the Texas Senate race, the 2028 presidential field, and the question of whether Taylor Swift releases a song for Toy Story 5.

Political and cultural markets on Kalshi move in real time — as news breaks, odds shift. Which means someone who reads closely, follows polls, and actually pays attention to what's happening has a genuine edge over someone who doesn't.

On Kalshi, information is the advantage. The platform is federally regulated under the CFTC, with no house edge and no restricted winners. Every dollar traded reflects real collective intelligence about what's going to happen.

If you're already following politics and culture closely, Kalshi turns that attention into an asset.

Sign up and start trading the events you're already watching.

Start trading now

_{Trade responsibly.}

Perpetua is the odd name here, the one whose thesis has nothing to do with quarterly numbers because it has no product to sell yet. It is developing the Stibnite Gold Project in Idaho, positioned to become the only domestically mined source of antimony, a mineral used in munitions primers and missiles that the U.S. currently imports entirely.

It touched a high around $37 in early March as the whole defense-minerals complex re-rated, and has since slipped about 40%.

What sets it apart from a typical development-stage miner is who is standing behind it. The project is backed by $59.4 million in Defense Production Act funding, it broke ground in October 2025, and after a court denied opponents’ motion for a preliminary injunction in late May, Perpetua pushed ahead with critical-path construction.

In that same announcement, the Department of War identified Stibnite as the only U.S. mine capable of meeting defense antimony demand by 2029. A roughly $2 billion federal debt package is in progress.

All of that is real, and none of it changes the fact that this is a pre-revenue company with a market cap near $2.5B and first antimony production not expected until around 2029. The value here is option value on execution, permitting, financing, and a gold price that sits near $4,100 and is currently a tailwind but need not stay one.

For a name like this, the second half is measured in construction milestones and the federal financing decision, not earnings. The fundamentals are unusually well-supported for a developer but the timeline is long, and the stock will trade on milestones and gold until it does not.

Karman set its all-time high near $118 on January 28, barely a year after its February 2025 IPO, and has since given back more than half.

The company builds systems and structures for missile defense and space, spanning programs its chief executive describes as seeing generational demand, including SM-3, PAC-3, THAAD, hypersonics, and, after a January acquisition, submarine and naval work.

The strength is not in doubt. First-quarter results showed record revenue of $151.2 million, up 51%, a record backlog of $1.0 billion, up 61%, and raised full-year guidance. The 2027 defense-procurement setup gives that backlog a clear runway.

The trouble is in the fine print. First-quarter revenue and earnings landed just short of estimates, the company has disclosed material weaknesses in its internal financial controls and replaced its auditor with PwC, and an acquisition-led strategy has pushed notes payable to roughly $770 million with quarterly interest near $13 million.

The roughly 50% drawdown reads less like a demand reset than a de-rating on execution and balance-sheet risk. The second half depends on whether Karman can string together clean beats and show progress fixing the control issues. The order book buys it time; the market is asking for reliability before it pays up for the growth again.

Of the AI names that hit all-time highs this spring, Innodata is the one with real, growing profits behind the move. It is a data-engineering company that builds the training, evaluation, and safety datasets that large AI-model builders rely on, and it hit an all-time high of $125.14 on June 4 before giving back about 42%.

The move had earnings behind it. First-quarter results showed revenue of $90.1 million, up 54%, adjusted EBITDA nearly doubling to $25.0 million, and diluted earnings of $0.42 per share, and management raised full-year growth guidance to about 40% or more. It also flagged a new engagement with a leading big-tech customer expected to add about $51 million this year, a customer that contributed nothing twelve months ago.

The business is real but the stock still got ahead of it. At the June peak it traded near 60 times earnings, and revenue remains concentrated, with a single large customer driving a meaningful share and the whole model dependent on frontier-AI labs continuing to outsource data work rather than build it in-house.

This is the cleanest set of fundamentals in the group, which cuts both ways: the June high demanded flawless execution, so the second half turns on whether the next two quarters keep validating the 40%-plus trajectory and the new big-tech ramp. If the growth holds, the pullback looks like a reset rather than a top. Earnings, unusually for this list, will do the deciding.

Wall Street is shifting billions into a select group of stocks, and MarketBeat’s updated 10 Best Stocks to Own in 2026 report reveals exactly which ones. Get the 10 names attracting fresh capital before the crowd catches on. Send My Free Report

Dave is the counterexample on this list, the name that has not cracked. The neobank, built around a cash-advance product called ExtraCash and AI-driven underwriting for customers underserved by traditional banks, trades a hair below its 52-week high around $404, and unlike most of this group it is durably profitable.

The numbers are the reason it has held. Full-year 2025 results showed revenue up 60% to $554.2 million, net income up 238% to $195.9 million, and adjusted EBITDA up 162% to $226.7 million, and the board tripled its buyback authorization to $300 million.

The company joined the S&P SmallCap 600 in June. To be fair, Dave’s all-time high record on a strict basis was a brief 2022 post-SPAC spike near $457 that it has not reclaimed, so take its inclusion here as a name pressing against multi-year highs than one setting fresh records. At roughly $5B, it also sits near the top of the small-cap range.

A lending-driven neobank ultimately lives on credit quality. That has held up, with delinquencies improving even as originations grew, but a deterioration in consumer credit would hit the thesis directly, and the stock’s run already assumes continued execution.

This is the one name here still in an uptrend, so its second-half question is different from the rest: whether it clears resistance into clean new highs or stalls under that old 2022 print. With real profits and a buyback underneath it, the setup is more constructive than most on this list, and the burden is on continued credit discipline.

Centrus is the wildcard here, the one name that already had its all-time high and lost it. It ran to $464 in October 2025 on the nuclear-fuel trade and now trades roughly 63% below that.

What earns it a spot on a momentum list is the concrete setup it has heading into the second half.

The company is the only U.S. producer of enriched uranium and, for now, the only commercial producer of HALEU, the high-assay fuel that advanced reactors and the AI-power buildout will need. On July 1 it signed a $900 million Department of Energy contract to move its HALEU cascade to commercial operation, a deal worth up to $1.07 billion with options, sitting on top of a $2.4 billion low-enriched-uranium backlog. A federal ban on Russian enriched-uranium imports tightening toward 2028 hands it a policy tailwind and a domestic-supply mandate.

The selloff was not random, and the risks are specific. Even after the fall, the stock trades around 54 times trailing earnings, first new commercial capacity is not expected until 2029, and the same July announcement funded a competitor, General Matter, with its own $900 million HALEU award. The monopoly is real today and is being deliberately competed away.

This is the one name betting on a recovery rather than defending a peak: the DOE contract and the commercial transition are real second-half catalysts, and the domestic-fuel thesis is as strong as any here, offset by a still-demanding valuation and government-funded competition arriving behind it. Whether the second half is the turn depends on execution on that cascade and where enrichment pricing goes.

Hit reply with thoughts, corrections, or names worth a closer look. If you’re seeing this outside of your inbox, shoot me an email here.

Keep digging and see you next time!