A10 Networks: The Q2 Beat Confirms The Business, But The Valuation Still Demands Too Much

A10 Networks: The Q2 Beat Confirms The Business, But The Valuation Still Demands Too Much

A10 Networks (NYSE: ATEN) beat Q2 earnings, raised guidance, and the stock fell anyway. The question isn't whether the business is running-it is. The question is whether a stock that's up 74% year-to-date, now trading at nearly 50x earnings on guided mid-teens growth, is actually cheap after a 20-day pullback of only 17%. It's not. Hold.

What happened

A10 Networks reported Q2 2026 revenue of $80.1 million, up 15.5% year-over-year, and non-GAAP EPS of $0.25, beating the consensus estimate of $0.24 by a cent on both revenue and earnings lines. Product revenue, the harder part to forecast, jumped 25% to $49 million, driven by security-led demand. Non-GAAP gross margin sat at 80.3%, operating margin at 25.5%, and adjusted EBITDA - earnings before interest, taxes, depreciation, and amortization, a rough cash-earnings proxy - hit 30.5% of revenue, or $25.4 million.

The company raised full-year revenue guidance from 10-12% to 12-14% and EPS guidance from 12-14% to 14-16%. Management also expects full-year free cash flow to exceed the roughly $65 million it generated in 2025.

Shares fell 2.25% in after-hours trading to $30.15 and have pulled back roughly 17% over the past 20 sessions. The stock sits at $30.84 today, still well above the 52-week low of $16.52.

Why the beat didn't move the needle

This is the fourth double-digit growth quarter in five, and the operating profile is strong. But the market isn't paying for past quarters. It's pricing in the next two years of execution at a premium that demands more than what the guidance provides.

The stock trades at roughly 50x trailing earnings, 55x forward, and about 7.4x sales. EV/EBITDA sits near 31x. Against guided revenue growth of 12-14%, the PEG ratio (price-to-earnings divided by the growth rate, a rough measure of whether a premium multiple is justified by the speed of growth) runs between 3.6 and 4.6. That is not a valuation reset. That is a stock that has already run hard and taken a modest pullback while still demanding near-perfect execution for years.

The 17% drawdown over the past 20 sessions may look steep in isolation. But the stock is still roughly 87% above its 52-week low. A selloff is only a bargain when valuation resets faster than the business deteriorates. Here, the business hasn't deteriorated at all, but the valuation hasn't reset meaningfully either.

What's under the hood

The product mix shift is the most compelling operating signal. Product revenue at $49 million, or 61% of total, is growing at 25% while service revenue ($31.1 million) grew more slowly. That's the right direction - products carry higher margins and A10 has been pushing its security-led application delivery controller and AI security products into both enterprise and service-provider channels.

Management flagged expanding Microsoft relationships, improving enterprise pipeline quality, and a gradual recovery in service-provider demand in the Americas. None of those are hard data points yet, but they fit the pattern of a company trying to grow into a larger total addressable market beyond its traditional mid-market niche.

Deferred revenue of $154.8 million provides visibility. At roughly two quarters of current run-rate revenue, the backlog is adequate but not a lock on future growth. The balance sheet shows $357.3 million in cash and marketable securities against $413.6 million in total debt, leaving the company with a manageable net debt position of roughly $56 million and flexibility to invest or return capital. Free cash flow of $26.9 million in Q2 is serviceable.

The TrojAI acquisition, closed in June, is the forward-looking bet. Management says native AI security solutions could reach the market in one to two years. That timeline matters because it means TrojAI won't be a revenue driver in 2026 or even most of 2027. It's an option on AI security, not a current cash-flow contributor.

Valuation versus the business

Here's the arithmetic. If A10 hits the midpoint of its raised guidance - roughly 13% revenue growth and 15% EPS growth - full-year EPS works out to approximately $0.95 on about $320 million in revenue. At $30.84, the stock trades at roughly 32x that forward EPS. That sounds more reasonable on its face, but it assumes 2026 execution hits the midpoint, 2027 growth doesn't decelerate, and the multiple doesn't contract as the market reassesses what "AI-adjacent network security" is worth after the initial excitement fades.

The closest operational peers in the application delivery and network security space are F5 Networks and Radware. Neither is a perfect analog - F5 carries a larger legacy enterprise footprint with different growth dynamics - but what's clear is that the market isn't assigning a deep discount to A10 for its size, competitive position, or execution risk. That gap between a solid company and a compelling stock is what keeps the rating at Hold.

The dividend yield of 0.78%, at a 39% payout ratio, is safe but too thin to compensate for valuation risk. It's a floor, not a driver.

Risks

• Growth deceleration. Service-provider demand, historically a laggard, is only "gradually recovering." A miss there drags the full-year below even the raised 12-14% range.
• Product concentration and competition. Product revenue is growing, but A10 competes against F5, Radware, and increasingly cloud-native alternatives from AWS, Azure, and Google. Share loss in any major segment would be material at this scale.
• AI timeline risk. The TrojAI story is real, but 1-2 years to market means 2026 and 2027 are effectively pre-TrojAI from a revenue standpoint. The current valuation can't be built on a product that isn't shipping yet.
• Customer concentration. A10's revenue base is weighted toward a modest number of large service providers and enterprises. Losing a single top-10 customer would be a meaningful hit.

The call: Hold

The Q2 beat and raised guidance confirm the operating engine is running. Product growth at 25%, 30.5% EBITDA margins, and a clean balance sheet are all good facts. But the stock at 50x trailing earnings on guided mid-teens growth is pricing in sustained execution with no margin for error.

The rating would flip to Buy on a deeper pullback - into the mid-$20s on the stock, where the forward multiple compresses closer to 25-28x and the risk/reward actually resets. Alternatively, a quarter showing product growth accelerating past 30%, tangible TrojAI integration progress with bookings, or a clear inflection in service-provider demand would justify the current price.

Until then, the next proof point is the November 3 earnings report. If Q3 continues the double-digit growth trajectory and management maintains or raises the full-year outlook, the valuation becomes more defensible. If growth slows or guidance softens, the downside from here is meaningful.

The business is solid. The price, at this point, is not.