Baron SMID Cap ETF Q1 2026 Commentary

Baron SMID Cap ETF fell 10.56% in Q1 versus a 3.52% decline for the Russell 2500 Growth, with software and AI-related positioning driving much of the underperformance. The letter highlights strong underlying fundamentals in holdings such as Netskope, Waystar, Samsara, Dynatrace, and Guidewire, while several positions were sold due to acquisitions or valuation discipline, including Kratos, Arcellx, Penumbra, Clearwater Analytics, and Exact Sciences. Management remains constructive on long-term returns, arguing the market’s AI winner/loser rotation and short-term dislocations are masking portfolio company cash-flow growth.

Analysis

The quarter’s real signal is not just factor rotation; it is a temporary repricing of software duration under AI-disruption narratives. That creates a two-tier market where businesses with proprietary, non-public, or physically anchored data can be mispriced alongside genuinely vulnerable incumbents. The most durable beneficiaries are not generic “AI software” names, but companies whose workflows are embedded, regulated, or paired with hardware — that group should regain multiple expansion first once investors differentiate model risk from actual product risk.

The best second-order setup is in the “picks-and-shovels to AI” cohort, but with a selectivity problem: supply-chain winners with scarce manufacturing or optical/IP bottlenecks can compound faster than the obvious semiconductor proxies if AI capex keeps broadening. At the same time, the latest rally in defense and networking names shows that portfolio flows are chasing visible budget and capex commitments, which means selloffs in those names are likely to be shallower and shorter unless spending assumptions roll over. The risk is that these leaders become consensus-owned momentum trades; if hyperscaler spending decelerates even modestly, their multiples can compress faster than fundamentals.

The consumer drag looks more like an energy-price pass-through and promotional efficiency issue than a structural demand break. That matters because the rebound path is asymmetric: if energy stabilizes and promotional ROI normalizes, the high-growth consumer basket can snap back quickly over 1-2 quarters. Conversely, the market is still underestimating how long the anti-software narrative can persist; if AI fears remain unchallenged through next earnings season, multiple compression may continue even as fundamentals hold up, forcing active managers to own a longer time-horizon through position sizing rather than thesis abandonment.

Key Decisions for Investors

• Go long GWRE / DT / IOT as a basket for 3-6 months: these are the cleanest examples of proprietary-data software that should re-rate once AI-disruption fears fade; target 15-25% upside with 10-12% downside if the selloff persists.
• Maintain or add to COHR on pullbacks over the next 1-2 months: it has the strongest operating leverage to AI optical demand, but size carefully because the trade is now partially consensus and vulnerable to any hyperscaler capex air pocket.
• Use FLUT vs DKNG as a relative-value pair: short FLUT, long DKNG for 3-9 months on cleaner U.S. exposure and better FCF conversion; favorable if prediction markets remain small and promotional intensity normalizes.
• Buy KTOS on weakness for a 6-12 month horizon: defense spending is a secular budget story, but trim into strength because the stock is likely to trade ahead of delivery if contract flow stays strong.
• Consider a tactical long WAY with a 6-month view: pricing remains disconnected from cash flow growth, and revenue-cycle automation is one of the least LLM-disruptable software niches; risk/reward is attractive if execution stays on track.