$49.7 million is not a huge ETF flow by mega-cap AI standards, but for a small-cap tech fund that had bled assets for four straight years, it marks a clean break in investor behavior.
The Invesco S&P SmallCap Information Tech ETF (PSCT) has pulled in $49.7 million in net inflows in 2026 as of May 27, ending four consecutive years of outflows, according to CryptoBriefing. The signal is not that investors have abandoned Nvidia or Microsoft. It is that the AI trade is no longer being treated as a one-stock, one-sector, one-market-cap story.
$49.7 million says AI capital is moving beyond the obvious winners
For years, the AI trade had a simple structure: own the giants building or monetizing the first wave of infrastructure. That meant large-cap hardware, hyperscale cloud, and platform software names. The new flow into PSCT suggests investors are now hunting for second- and third-order beneficiaries: smaller companies tied to data centers, networking, optical connectivity, chips, testing equipment, and specialized infrastructure.
That shift matters because ETF flows often show a change in appetite before every underlying stock has a clean, mainstream narrative. Buying a small-cap tech ETF is different from betting everything on one speculative AI supplier. It spreads exposure across a basket while still targeting the same thesis: AI spending may ripple outward from the largest platforms into smaller vendors.
“The AI trade has broadened quite materially,” said Oren Shiran, portfolio manager for Lazard US Systematic Small Cap Equity ETF. “Small-caps have become a real part of the second and third order of AI.”
MLXIO analysis: that phrase — “second and third order” — is the core of the trade. Investors are not just buying AI models or GPUs. They are buying the pipes, components, equipment, and adjacent technologies that may benefit if AI infrastructure spending keeps expanding.
For readers tracking how capital moves when a crowded trade starts to broaden, MLXIO has covered similar allocation questions in crypto funds, including $1B Bitcoin ETF Dump Sends Investors Into HYPE Funds, and in market breadth through 493 Stocks Wake Up as S&P 500 Earnings Hit 2021 Peak.
A 54% small-cap tech surge has outpaced the S&P 500 tech index
The flows are backed by performance. The S&P 600 small-cap tech index has gained almost 54% this year, compared with a 20.1% rise in the S&P 500 technology index, according to Trivariate Research data cited in the source material. That gap is the widest since before 1995.
| AI trade segment | Source-supported signal | Investor read-through |
|---|---|---|
| Mega-cap technology | Large-cap AI leaders still dominate the original AI playbook | High expectations raise the bar for future upside |
| Small-cap technology | PSCT has $49.7 million of 2026 inflows after four years of outflows | Investors are seeking AI exposure beyond the largest names |
| Small-cap semiconductors | Expected profit growth of nearly 40% in the second quarter | Earnings leverage may be strongest where AI demand is directly tied to chips and infrastructure |
| Broader small-cap tech | Expected earnings growth of 7% next quarter, excluding loss estimates for MARA Holdings | The rally is not evenly supported by fundamentals |
The distinction is crucial. This is not a blanket small-cap bid. The source material points to a technology-focused rotation, with investors seeking small-cap exposure in areas more directly tied to AI infrastructure rather than simply buying the entire small-cap universe.
MLXIO analysis: that selectivity strengthens the signal. A broad small-cap rally could simply reflect general risk appetite. A targeted move into small-cap technology suggests investors are actively mapping AI demand into narrower parts of the market.
MaxLinear’s sharp rally shows both the opportunity and the danger
The most striking company-level example is MaxLinear (MXL). CryptoBriefing says MaxLinear is among several small-cap tech names that have posted triple-digit gains in 2026. Reuters-linked source material adds that MaxLinear posted a 43% revenue jump in its most recent earnings report, citing demand from hyperscale customers.
The company makes high-speed optical connectivity solutions, which are used to move large volumes of data between servers in AI data centers. That makes the story more concrete than a generic “AI exposure” label. There is an identifiable product category, an identifiable customer type, and a demand channel tied to data center buildouts.
But the caution is just as important. The source material says MaxLinear and VIAVI have swung between quarterly profits and losses since their inception. Ichor Holdings has jumped about fourfold this year even though it last reported a quarterly net profit in December 2022.
That creates a sharp split inside the small-cap AI trade:
- Real demand: Companies with measurable revenue tied to AI infrastructure have a stronger case.
- Earnings uncertainty: Inconsistent profitability makes rallies more fragile.
- Narrative risk: Some stocks may be rising because investors expect AI benefits before those benefits appear in results.
- Debt sensitivity: Higher government bond yields worldwide could dent appeal because smaller tech companies are more economically sensitive and often rely more on debt-fueled growth.
The price of small-cap tech stocks has more to do with “speculation and less on changes in fundamentals relative to large-cap stocks,” said Hal Reynolds, senior portfolio manager at Los Angeles Capital Management.
That skepticism does not kill the trade. It narrows the test. The market is willing to pay for possible AI leverage. Now companies must prove the leverage exists.
Rates and earnings make this more than a simple AI story
Rates are still part of the setup, but mainly as a risk factor. The source material notes that higher government bond yields worldwide could dent the sector’s appeal because smaller technology companies are more economically sensitive and often rely on debt-fueled growth.
That creates a two-sided rally test: AI demand on one side, financing pressure on the other. If AI-related orders improve revenue while borrowing conditions remain manageable, small-cap tech earnings can move quickly in percentage terms. If yields rise and debt costs become more restrictive, the same stocks can become more vulnerable.
Still, the earnings data does not support a simple “all small-cap tech wins” view. Semiconductor small caps are expected to post profit growth of nearly 40% in the second quarter, while the broader small-cap tech sector is expected to grow earnings by just 7% next quarter, excluding loss estimates for MARA Holdings.
MLXIO analysis: the market is already separating infrastructure-linked names from the broader basket. That separation should deepen as earnings reports force companies to quantify AI demand rather than simply reference it.
A useful parallel inside digital assets is benchmark dispersion: strong individual winners do not always move the whole index. MLXIO explored that dynamic in 13 Crypto Winners Can't Make CoinDesk 20 Budge an Inch. The same logic applies here. A handful of AI-linked small caps can surge while the broader group remains uneven.
Small-cap AI rallies still need proof
The relevant warning is straightforward: small-cap rallies can move faster than the evidence supporting them. When a trade shifts from mega-cap leaders into smaller companies, the upside can be sharper, but the burden of proof also rises.
The current move has a more concrete catalyst than a generic risk rally: investors are looking for actual AI-linked demand in revenue, orders, and customer activity. MaxLinear’s cited revenue growth is one example of the kind of evidence the market wants to see.
That distinction is useful, but not conclusive. A company can show AI-related demand and still fail to convert it into durable margins or consistent profits. A stock can rise on a valid business catalyst and still overshoot. A sector ETF can attract inflows and still contain weak companies alongside real beneficiaries.
For investors, the due diligence list is straightforward:
- AI revenue: Is management quantifying demand, or just using AI language?
- Customer quality: Are hyperscale or enterprise customers actually buying?
- Margins: Is growth improving profitability, or only boosting sales?
- Cash burn: Can the company fund growth without repeated dilution or expensive debt?
- Earnings consistency: Are profits recurring, or still swinging quarter to quarter?
The next evidence will come from earnings, not ETF flows
The small-cap AI rotation looks real enough to watch, but not clean enough to buy blindly. PSCT’s $49.7 million inflow reversal shows investors are broadening the AI trade. The 54% surge in the S&P 600 small-cap tech index shows the market has already rewarded the thesis. The earnings split shows why selection still matters.
The confirming evidence would be more companies reporting AI-linked revenue growth like MaxLinear’s cited 43% jump, especially if that growth comes with better margins and steadier profits. The weakening evidence would be higher yields, fading AI capital spending, or earnings reports where management talks about AI but cannot attach it to orders, revenue, or customer wins.
The practical takeaway: small-cap tech may be where the AI trade becomes more interesting, but also more punishing. The label will not be enough. The next winners have to show the numbers.
Disclaimer: This MLXIO analysis is for informational and educational purposes only. It is not financial, investment, legal, tax, or professional advice. It does not provide buy, sell, hold, price-target, portfolio, or personalized recommendations. Verify information independently and consult qualified professionals before making decisions.
The Bottom Line
- Investors are expanding the AI trade beyond Nvidia, Microsoft, and other mega-cap leaders.
- PSCT’s $49.7 million inflow marks a reversal after four straight years of outflows.
- Small-cap tech stocks may offer broader exposure to AI infrastructure demand without relying on a single winner.
