Cisco just had a massive rally. HSBC says the stock still has momentum

HSBC upgraded Cisco to buy and lifted its price target to $137 from $77, citing a more structural AI revenue opportunity. Cisco’s fiscal Q3 EPS came in at $1.06 versus $1.04 expected, and revenue was $15.84 billion versus $15.56 billion consensus, while Q4 EPS guidance of $1.16-$1.18 topped the $1.07 estimate. Shares jumped 13% on Thursday and are up 50% year to date.

Analysis

The key takeaway is not that Cisco beat a quarter; it is that the mix of demand is shifting from cyclical order recovery to a more durable AI-related spend stack. That matters because networking is one of the few parts of the AI capex chain where monetization can show up quickly in P&L, which tends to expand the market’s willingness to pay a premium multiple for a legacy infrastructure name. If this is a structural rather than transitory AI revenue stream, the rerating can persist for several quarters even if hardware growth elsewhere normalizes.

Second-order winners are the ecosystem players tied to enterprise network refreshes, campus upgrades, and adjacent security spend. The likely pressure point is on peers selling generic switching/routing exposure without Cisco’s installed-base leverage or pricing power; they may face longer sales cycles and more aggressive discounting as customers consolidate AI/network budgets around fewer vendors. Supply-chain beneficiaries are more subtle: tighter contract terms and lower memory intensity imply Cisco can protect margins without relying on broad component inflation, which reduces upside for low-quality component suppliers and shifts margin capture toward software-attached infrastructure.

The near-term risk is not demand collapse but multiple compression if the market concludes the story has moved too far, too fast. A 50% YTD move already prices in a lot of the good news, so the stock becomes more sensitive to any evidence that AI-related orders are lumpy, hyperscaler demand is front-loaded, or enterprise upgrades merely pulled forward spending from future quarters. The biggest reversal catalyst would be guidance that confirms the beat was mix-driven but not sustainable, especially if gross margin pressure starts to outrun offsetting pricing and operational discipline.

Consensus appears to be underestimating how quickly Cisco can convert AI adjacency into free cash flow rather than just revenue. The contrarian view is that this is still a quality compounder, not a true AI secular winner, so the upside from here is likely more in earnings durability than explosive top-line acceleration. That suggests the trade is better expressed as owning it for relative defensiveness and gradual multiple expansion, not chasing it as a momentum beta proxy for AI infrastructure.

Key Decisions for Investors

• Stay long CSCO on pullbacks over the next 1-2 weeks; the post-earnings re-rating can extend as analysts reset estimates, but trim strength if the stock approaches a full revaluation of the upgraded target range.
• Pair trade: long CSCO / short a less differentiated networking peer such as ANET or JNPR over 1-3 months if you believe enterprise AI/network refresh demand will concentrate in the installed-base incumbent rather than pure-play exposure.
• Buy CSCO call spreads 1-2 quarters out to express upside with limited premium burn; risk/reward is attractive if the market continues to price AI revenue as structural, but capped upside is appropriate after the sharp YTD move.
• For relative value, use CSCO strength to underweight lower-quality hardware suppliers tied to generic campus/network refreshes; they face margin compression if Cisco keeps using pricing and contract discipline to defend share.
• If CSCO fails to hold the post-earnings gap within 2-4 weeks, fade the move with a small tactical short or put spread: the main risk is the market deciding the AI story is already fully discounted.