AI, identity wars, and digital battlefields make cyber defense mission-critical, but picking the champion isn’t easy
It’s been a long time since I last wrote about cybersecurity companies. Not because the sector became less important — quite the opposite. In an age defined by AI, cloud computing, identity warfare, and actual geopolitical conflict increasingly playing out through cyberattacks, cybersecurity might be one of the most crucial industries of the next decade.
That also explains why valuations across the sector are extremely high. And honestly, for the eventual winner(s) of this race, those valuations might very well be justified. But for the rest, I have my doubts. The problem, and the reason I’m not buying aggressively right now, is simple: I don’t know who the winner will be. The space is arguably more crowded than any other. And that’s not really where I like to put my eggs into an arbitrary basket.
Yes, Palo Alto is the largest. Yes, CrowdStrike is growing the fastest and seems to be the market favorite right now. But so what? What happens if Microsoft or Google decide to push even harder into security and bundle it deeply into their ecosystems? Could they eventually crush legacy players or at least compress margins across the board and make investors rethink current valuations? That uncertainty makes conviction in specific players significantly harder.
I owned Fortinet for a while after its disastrous earnings, betting on a quick turnaround. That turnaround never really materialized, growth lagged, and my long-term confidence faded. I exited with roughly a 15% gain. This investment only reinforced my concerns about stock-picking in a still quite early-stage, explosive growth industry. The more I looked into the field, the more worried I grew about owning Fortinet. At the moment, I’m looking at two very different value propositions that have been beaten up over the last few months (one more than the other): SentinelOne and Zscaler.
SentinelOne focuses on endpoint protection with its Singularity platform, which uses AI-driven threat detection to protect devices like laptops, servers, and workloads directly and in real time. This puts SentinelOne in more direct competition with CrowdStrike and Microsoft Defender. Companies that you really don’t want to be in competition with. The technology is impressive, and growth remains strong, with revenue still growing north of 20% annually. Free cash flow is exploding, even if it’s from a small base, which shows operational leverage starting to emerge.
But SentinelOne has a clear and painful problem, and that’s profitability. The company actually took a step backward this year, and breakeven is not even in sight yet. And the stock reflects that: SentinelOne seems to be a perpetual loser, almost damned to eternal disappointment. Nothing excites investors anymore, which is also why the stock has retreated consistently since its IPO in 2021 and now sits almost 70% below the listing price.
Compressed valuation figures reflect that performance, with an EV/revenue multiple of around 4, which is not only well below its historical average of around 8, but also by far the cheapest name in the cybersecurity sector. The market doesn’t see SentinelOne as a clear future leader (at least not anymore). Not the next CrowdStrike, but rather more of a potential turnaround or acquisition candidate. The second is where I also see substantial upside for investors, since SentinelOne represents an obvious target for a tech giant pivoting more into endpoint security.
Zscaler, on the other hand, operates at a different layer entirely — arguably also a more interesting one. It’s not about protecting devices, but about securing access. Zscaler is the poster child for Zero Trust architecture: users don’t get network access at all, but rather get application access, verified continuously based on identity, device, and context. This makes Zscaler the more future-first, cloud-focused player and more embedded into enterprise infrastructure with an entirely new approach. Sounds cooler? That’s also what the market thinks.
That strategic position shows in the numbers. Zscaler is much closer to profitability, projected for FY27. Revenue growth is still projected above 20% annually through 2028, and free cash flow is already substantial, growing in the mid-20% range with margins around 27%. Unsurprisingly, the market rewards that quality: EV/revenue sits around 10 and the FCF yield around 2.5%. Not completely outrageous given the growth and position, but definitely not cheap.
That leaves me on the sidelines. In short: SentinelOne is cheap but risky. Zscaler is high quality but expensive (even if cheaper than some larger competitors). And most importantly, I still can’t confidently say which model will ultimately dominate enterprise security in a world essentially controlled by a few hyperscalers. If I had to bet on a long-term winner, I’d probably put my money on Microsoft or Google.
Cybersecurity is inevitable, and demand will only grow. But until the hierarchy becomes clearer and valuations more reasonable, this is a space I’m only watching. It’s a fascinating sector without a doubt — just not an easy one to underwrite today.
$S (-0,11 %)
$ZS (-2,39 %)
$CRWD (-3,79 %)
$PANW (-3,8 %)
$FTNT (-2 %)
$NET (-8,03 %)
$MSFT (-1,42 %)
$GOOGL (-1,22 %)
$GOOG (-1,13 %)
