Zscaler Isn't Down 55% Because the Business Broke — It Fell Because Two Sales People Left

The pitch writes itself: Zscaler, a "magnificent" AI-security growth stock down 55%, and Wall Street telling you to buy. The two reasons usually attached are that it is cheaper than its own recent history and that the Street's average target sits roughly a quarter above the current price. Both are true. Neither, alone, is a reason to own it.

So let's start with what actually happened, because the headline buries it. Zscaler did not bleed its way to a 55% discount on failing demand. On May 27 the stock dropped more than 30% in a single session — its worst day ever — after the company disclosed it had lost two senior sales leaders during its fiscal third quarter, and CFO Kevin Rubin said Zscaler was taking a "prudent approach" to guidance. For fiscal 2027 it guided annual recurring revenue growth to just 16% to 17%, below what analysts had been modeling. Every current, live reason the stock is in the dump — still down about 27% this year and roughly half below its 52-week high near $337 — traces back to that one session, not to 2021.

The entire argument is one question

The discipline question is whether a two-person sales shakeup is a business-model problem, in which case the discount is deserved, or a fixable ramp, in which case the market over-reacted. That single distinction is the whole thesis. And on September 3 Zscaler delivered the counter-evidence.

Its fiscal fourth quarter grew revenue 25% year over year to $898 million and pushed annual recurring revenue up 25% to $3.77 billion. Strip out the Red Canary acquisition and net new ARR accelerated to 17% growth — management itself used the word "accelerated" — on the best sales productivity the company has seen since 2022. Non-GAAP operating margin hit a record 24%, and the GAAP loss from operations narrowed to just 2% of revenue.

That is the signature of growing pains, not a broken model: the two leaders are gone and replaced, the growth engine re-accelerated the quarter after they left, and management itself says the rebuild is weighted toward the back half of fiscal 2027. The market's May verdict priced a structural decline in the growth engine; the company reported a fixable wobble and the market, down 27% on the year, has not yet re-priced it.

"Cheap" is doing two different jobs

Here is the honest part. At about 8 times trailing sales, Zscaler is the cheapest of the big security vendors — Palo Alto Networks trades near 24 times sales and CrowdStrike near 39 times. If you are paying relative price for security-software growth, this is the inexpensive seat.

But notice what the pitch does not say. The claim that the stock needs to rise roughly 67% just to return to its own three-year average price-to-sales ratio of 13.2 is mean reversion — a hope that the multiple snaps back because it used to sit there. That is not a business case, and it is exactly the reasoning that leaves investors holding a bag. On the cash-generation basis the company actually guides, Zscaler is no steal: against roughly $900 million of fiscal 2027 free cash flow, an enterprise value near $25 billion works out to about 27 times forward free cash flow, with guidance targeting a 23% free-cash-flow margin.

The divergence, and what has to show up

So this is not the "magnificent stock, down 55%, just buy it" trade. It is narrower. The market's bear story — that AI upends subscription security and the growth engine is gone — is being contradicted by a business that just re-accelerated net new ARR and printed record margins, yet still carries the price of the structural read. That disconnect is real, and it is a legitimate edge if, and only if, the forward math holds.

What converts a warranted re-rating into an actual discount is one observable thing: whether net-new ARR keeps re-accelerating through the back half of fiscal 2027 — management has said the leadership transition keeps first-half growth muted — and whether the company finally closes its GAAP loss. Until that number shows up, the honest call is a watchlist held on forward math, not a "buy right now" handed over by a mean-reversion headline.