The networking giant is growing again. The awkward question is whether investors have already priced in the comeback.
Cisco built the plumbing. Wall Street suddenly wants the skyscraper
Cisco has done something Wall Street has spent years asking it to do: grow.
Revenue reached $63.33 billion in FY2026, up 11.8%, while net income surged 30.3% to $13.27 billion. Diluted EPS rose 30.6% to $3.33. Networking revenue jumped from $28.30 billion to $34.67 billion.
And yet, looking at the share price, you might assume $Cisco(CSCO)$ had turned up to the AI party wearing last decade’s outfit.
The shares closed at $109.59 on 20 August, well below the 52-week high of $130.37. More intriguingly, the analyst consensus remains Buy, with a $136.05 price target implying 24.15% upside.
That disconnect is the real Cisco story.
The market isn't asking whether Cisco can benefit from AI anymore. It is asking whether that growth is sufficiently profitable, durable and differentiated to justify the valuation investors have now attached to the company.
The old Cisco is becoming inconvenient
The first thing I notice in the numbers is how much the business has changed.
Networking generated $34.67 billion in FY2026, accounting for more than half of total revenue. Security contributed $8.23 billion, collaboration $4.30 billion and observability $1.10 billion.
But the quiet heavyweight remains services, at $15.03 billion.
That matters because Cisco isn't simply becoming a hardware company chasing AI infrastructure. It is attempting something considerably more interesting: using an enormous installed base of enterprise customers to participate in faster-growing infrastructure markets while retaining the higher-margin economics of its established services operation.
That is the bull case.
The bear case is that AI infrastructure eventually makes Cisco more hardware-heavy just as investors are paying it a higher multiple for becoming a growth company.
In other words, Cisco could discover that getting invited to the AI party and getting paid properly for attending are two different things.
The margin maths nobody should ignore
This is where the financials become much more revealing than the headline revenue number.
Cisco's gross margin was 64.6% in FY2026, down slightly from 64.9% in FY2025. Product gross margin was 63.2%, while services produced a considerably stronger 68.8%.
That difference is not cosmetic.
If Cisco's growth increasingly comes from infrastructure products, the company could produce impressive top-line numbers without enjoying the same incremental profitability that investors have historically associated with its more service-heavy model.
There is evidence on both sides.
Operating margin recovered from 22.1% in FY2025 to 25.4% in FY2026. That is a substantial improvement, but I would not call it a record. Cisco achieved operating margins of 27.35% in FY2023 and 27.2% in FY2022.
Free cash flow provides another useful reality check. It fell 3.9% to $12.77 billion, taking free-cash-flow margin from 23.46% to 20.16%.
So earnings are accelerating faster than cash generation.
Cisco's profits are catching up. Cash flow, rather less enthusiastically
That doesn't make the growth bad. It simply means I would like to see the cash follow the accounting profits before becoming too enthusiastic.
ROIC tells a more complicated story
There is another number I would not gloss over: return on invested capital.
Cisco's latest ROIC is 20.65%, down slightly from 21.55% for FY2026 and dramatically below 43.84% in FY2023 and 37.81% in FY2022.
But there is an important technical wrinkle here. The 20.65% figure is the spreadsheet's current measurement dated 20 August 2026, whereas 21.55% is the FY2026 figure for the fiscal year ended 25 July. They should not be presented as consecutive fiscal years.
The longer-term comparison is nevertheless worth attention. Cisco's acquisition of Splunk materially enlarged the capital base, making a straight comparison with the pre-acquisition years less clean. Still, today's ROIC is well below those earlier levels, raising an important question: has Cisco's growth become more capital-intensive?
A company can grow revenue and earnings impressively while becoming less efficient at deploying capital.
That is precisely the sort of detail a quarterly beat can obscure.
Competitive advantage: boring is beautiful
Cisco's competitive strength is not that it has suddenly invented networking.
It is that customers already use it.
The installed base, enterprise relationships and breadth across networking, security, collaboration, observability and services create an ecosystem that newer competitors cannot replicate overnight.
And Cisco has something else going for it: recurring economics.
Total annualised recurring revenue stands at $31.10 billion, while services revenue remains remarkably stable at $15.03 billion, barely changed from $15.05 billion the previous year.
That stability gives Cisco a valuable financial cushion while it pushes into higher-growth infrastructure.
The competitive question, therefore, isn't simply whether Cisco can win AI networking contracts.
It is whether it can use those contracts to pull more profitable software, security and services revenue through the same customer relationships.
If it can, the incumbent advantage becomes considerably more interesting.
The valuation has already noticed
This is where I become less comfortable.
Cisco trades at 33.20 times trailing earnings and 21.45 times forward earnings. The market capitalisation is $431.94 billion, up 62.9% in the latest comparison.
That is a very different Cisco from the one investors once bought primarily for its dividend and dependable cash generation.
The forward multiple looks reasonable beside many AI beneficiaries. But 'reasonable for AI' can become a dangerous phrase. It still assumes earnings growth arrives.
The analyst target of $136.05 is fascinating precisely because it shows the Street hasn't abandoned the story. Consensus remains Buy and implies 24.15% upside from the current share price.
So we have an unusual setup: the market price is sceptical while the analyst consensus remains optimistic.
The fundamentals improved. The market's mood remains considerably less stable
Someone is going to be wrong.
Cisco as the second act of AI
This may ultimately be the most important part of the story.
The first phase of the AI trade was about computational horsepower. The next phase is about the infrastructure surrounding it.
Networking is essential. Data centres cannot scale without it, and enterprises cannot deploy increasingly demanding workloads without upgrading the plumbing underneath.
That puts $Cisco(CSCO)$ in a potentially valuable position.
But second-order AI beneficiaries will not necessarily enjoy the extraordinary economics of the first-order winners. Hardware competition remains fierce, and margins matter.
This is why Cisco's transition needs to be judged on more than revenue growth.
If networking growth pulls through higher-margin security, software and services, Cisco could become something Wall Street once thought unlikely: a mature technology company with a credible AI growth engine.
If instead AI simply shifts the revenue mix towards lower-margin products, investors may eventually decide that Cisco has become a bigger company rather than a better one.
The interesting bit is what happens next
I don't think the latest numbers settle the Cisco debate. They sharpen it.
The bull case is easy enough to understand. Revenue is growing, earnings are growing faster, networking has become a major engine, recurring revenue provides ballast and the balance sheet gives management room to execute.
The bear case is equally coherent. The valuation has rerated dramatically, free cash flow has not kept pace with earnings, operating margins remain below their FY2022–23 peaks and ROIC is substantially below those earlier levels.
And sitting between the two is that $136.05 analyst target.
I find that more revealing than another Buy-or-Sell proclamation.
Cisco no longer needs to prove that it can grow. The numbers have already done that.
What it needs to prove is that AI growth can improve the quality of the business rather than merely increase its size.
The infrastructure is tangible. The expectations are considerably heavier
If Cisco manages that trick, today's scepticism could look rather short-sighted.
If it doesn't, Wall Street may discover that the most expensive thing about an AI revolution isn't always the hardware.
Sometimes it's the expectations.
@TigerStars @Daily_Discussion @Tiger_comments @Tiger_SG @Tiger_Earnings @TigerClub @TigerWire
Comments