Veteran analyst drops a shocker on Cisco

Cisco (CSCO) should be on cloud nine right now. The software giant posted another strong quarter. Everything, on a fundamental level, is going well: record revenue, double-digit growth, and a tangible surge in AI-driven demand; if these are not reasons to celebrate, I do not know what is. Yet its share price fell following the earnings release.

What gives? Ugly gross margins.

I’ve covered tech cycles for several years now. What I’ve learned thus far is that the tech market has its own set of inflection points.

When demand shifts, the first thing stock investors don’t always do is “re-rate the stock.” It’s “panic about the cost line,” especially when the new growth engine needs a lot of hardware.

Cisco CEO Chuck Robbins said it best:

And Bank of America’s Tal Liani came out with a succinct yet telling analysis. Cisco’s numbers look better than the market is giving it credit for and the guidance looks conservative.

CSCO shares were last trading hands for $76.85 on Tuesday, a key detail because BofA’s note pegged the stock at $85.54, meaning the post-earnings sell-off discreetly widened the upside from approximately 11% to nearly 24% of its $95 target.

Cisco just delivered an AI demand shock

Photo by Bloomberg on Getty Images

The quarter Wall Street should be talking about

Cisco’s fiscal Q2 (ended Jan. 24) delivered:

Revenue:$15.3 billion (up 10% year over year)
Non-GAAP EPS: $1.04
Networking revenue: $8.29 billion (up 21%)
RPO: $43.4 billion (up 5%)

CFO Mark Patterson, on the occasion, framed it in a wonderful manner:

These are great comments from Cisco; it’s a strong reaction from a mature infrastructure company amid possibly weakening demand.

So why the sell-off? When the spotlight turned to Cisco’s adjusted gross margin, shares tumbled about 7% in extended trading.

Overall, the gross margin was 67.5%, down 1.2% year over year.

The “BofA shocker”: Cisco may be sandbagging the back half

Here is where the BofA note flips the script and we go from “earnings recap” to “setup.” BofA’s core argument does not indicate that margins are fine. What it does do is indicate that the guidance math looks conservative. And conservative guidance in a demand upcycle is where upside tends to hide.

The 4Q revenue cadence looks oddly light

BofA picks up on some key indicators. According to the note, Cisco’s 3Q revenue guidance reflects around 9.5% YoY growth versus Street expectations around 7.3%, which is strong.

More Tech Stocks:

However, the implied 4Q guidance calls for only approximately 1.4% quarter-over-quarter growth, despite normal seasonality.

If, like me, you have experienced this before, you are familiar with the crucial line. When any company sees a demand tailwind, that seasonality doesn’t usually disappear. Instead, it’s deferred into the guidance as a cushion.

2) The implied 2H AI order number looks “too low”

BofA also wants investors to focus on what the guidance implies about $1.6B in 2H AI orders, approximately half the $3.4B seen in 1H, which management frames as “lumpy and non-linear.”

The language is not uncommon in hyperscaler-driven cycles. And it’s exactly where the estimated upside can live if the timing breaks.

The other BofA point investors should not ignore: this isn’t just AI

BofA’s note highlights an aspect that the market often overlooks:

Product orders when you take out AI are exceptionally strong, up 10% YoY.

The drivers BofA marked out are looking like a broad networking turn:

Campus switching strength
Wi-Fi 7 expansion (BofA cites +80% QoQ growth)
Data center conversion improvement
Further growth in servers

These data points are worth noting because they reduce the single-point-of-failure risk of hyperscaler timing. When in cycles, you require multiple engines, not one.

Cisco also holds product momentum on the AI plumbing side. In anticipation of earnings, Cisco unveiled a new AI networking chip, aiming to penetrate deeper into the AI data center stack. All of this, according to Cisco, will lead to an improvement in efficiency by rerouting data around network problems “within microseconds.”

The real risk: memory costs are the toll for winning AI hardware

Now comes the hard part: discussing the bear case, and believe me there is one.

BofA, within its note, carries out two pressing issues hitting gross margin:

Mix shift to equipment and cloud-related products
Memory pricing shock (BofA cites a 400% YoY hike in memory pricing) pushing gross margin down about 200 bps sequentially

It’s worth noting that Reuters, in explaining Cisco’s margin pain, said it is attributable mainly to rising global memory chip prices being pulled higher by AI infrastructure demand broadly.

The Wall Street Journal reported Cisco’s purchase commitments increased 73% in the past 90 days moving quickly to lock down supply. It is yet another sign that the cost pressure we are talking about isn’t theoretical.

I also like BofA’s counterpoint. Cisco can maneuver through the situation with the help of pricing. The note reveals campus switching and compute pricing are up approximately 3%, and further increases are also possible thanks to memory trends.

That is Cisco’s entire strategy. It’s not a situation like, “Will memory costs rise?” In essence, they already have. The question on everyone’s mind is, “How quickly does Cisco’s pricing and efficiency response show up in the numbers?”

BofA thinks the underlying trend in Security is better than the headline

BofA says that security revenue was down 4.4% YoY, largely thanks to Splunk’s transition from term licensing to SaaS. It can have a distorting impact on comparability.

However, what is the long-term play? Management predicts that security orders ex-Splunk will end the year with about 10% growth, saying that there is more demand for newer solutions, which BofA says now make up 33% of security revenue.

To me, that’s a “masking effect” investors often do not take into account. The business, as per statistics, looks weak in reported revenue, while the product cycle beneath the surface is improving.

What to do with the stock? BofA reiterates Buy and the valuation gap just widened

BofA’s conclusion is straightforward:

Buy rating reiterated
$95 price objective maintained
2026E EPS lifted to $4.15 (from $4.11)
2027E EPS lifted to $4.45 (from $4.40)
Revenue growth accelerating to 8.5% in 2026
Operating margin stable around 34%
$6.6B returned to shareholders year-to-date

The “OM stable” call is what makes this different from a basic margin fear. BofA is basically saying, “Yes, gross margin is under pressure, but Cisco can still make the operating model work.”

And here’s the best part: at $76.85, the stock has a lot more room to grow toward a $95 target than it had when BofA wrote the note at $85.54.

BofA sees a classic “margin scare” masking a demand inflection

BofA did not say this was a fantastic quarter. Instead, it called it an “execution quarter,” with a margin concern that the market is obsessed with and a demand scenario that the market is, possibly, underpricing.

In my experience, the ideal times to buy in the IT cycle occur when investors treat a temporary cost shock as a permanent loss.

Cisco still has to show that it can keep prices up even when memory prices go up. If BofA is correct, however, and the back-half estimate truly is cautious, this drop may wind up seeming less like a warning and more like a gift-wrapped reset.

And they don’t last long.