Broadcom, Cisco, Arista Networks And A Financial Stock On CNBC's ‘Final Trades'

Broadcom reported fiscal Q3 2026 revenue of $29.59 billion versus $29.36 billion consensus and adjusted EPS of $3.32 versus $3.24 expected; its shares rose 3% to $357.61. Deutsche Bank initiated Buy coverage on Arista Networks with a $220 target and Cisco with a $135 target. Itaú Unibanco's Q2 EPS of $0.21 and $9.186 billion in sales both missed consensus, while its shares fell 1.1%.

Analysis

The investable read-through is a widening bifurcation within AI infrastructure rather than a blanket networking trade. AVGO captures accelerator and custom-silicon content, while ANET is leveraged to the scale-out Ethernet fabric required once AI clusters move from pilot deployments to multi-site production; both have a more direct hyperscaler capex beta than CSCO. CSCO’s higher enterprise and campus mix makes its upside dependent on a broader IT-budget recovery, so analyst-target-driven strength without evidence of order acceleration is more likely to compress its relative valuation gap only temporarily.

Over the next 1-3 months, AVGO guidance, disclosed AI-networking/custom-silicon backlog, and hyperscaler capex commentary are the relevant catalysts—not small earnings beats or sell-side initiations. The key risk to AVGO/ANET is a customer-concentration air pocket: any delay in a major cloud customer’s cluster buildout can produce sharp multiple compression because expectations embed sustained AI growth. A sustained rise in supply-chain lead times or evidence that proprietary interconnect displaces Ethernet at the largest clusters would be a specific falsifier for ANET.

ITUB is the differentiated risk case: a modest operational miss matters less than the interaction of Brazilian rate cuts, credit costs and BRL translation. Falling policy rates can support loan growth and reduce funding costs, but an adverse BRL move or rising delinquencies would erase USD-holder returns quickly; this is not a clean single-quarter mean-reversion setup. Consensus may be underestimating that lower-rate Brazil favors duration-sensitive domestic financials only if fiscal credibility prevents renewed inflation and currency pressure.

Key Decisions for Investors

• Maintain/establish long AVGO on a 6-12 month horizon, sized below a core semiconductor position due to hyperscaler concentration. Add only on post-results weakness if AI revenue/backlog guidance remains intact; reassess if next-quarter AI growth guidance decelerates materially or a top customer signals capex cuts.
• Run a 3-6 month pair: long ANET / short CSCO in equal-dollar exposure. The thesis is superior AI fabric revenue sensitivity versus enterprise-networking cyclicality; exit if CSCO reports clear order-growth reacceleration while ANET backlog or cloud revenue commentary weakens.
• Do not chase CSCO solely on the new price target. Set an alert for evidence of improving product orders, security/software ARR acceleration, and enterprise spending recovery; absent those data, treat relative rallies versus ANET as potential short-entry opportunities.
• Keep ITUB at watch/underweight rather than buying the earnings miss. Revisit after Brazilian inflation, central-bank guidance and nonperforming-loan trends confirm that easing is translating into loan growth without BRL stress; a material BRL selloff or rising credit provisions invalidates a constructive thesis.