Okta runs the login layer for large organisations, the software that decides who can sign in to which applications and what they can touch once inside. That job is called identity and access management, or IAM.

In terms of the business model, Okta sells identity software by subscription, priced mostly per user and increasingly per identity, including non-human ones.

As the identity layer of the security stack, Okta is a direct way to own the agentic AI theme from the AI plumbing side, based on the argument that every AI agent needs a managed identity.

The revenue splits into two clouds.

Upstream, it runs on Amazon Web Services and federates more than 18,000 applications, including platforms such as Salesforce, Workday and ServiceNow.

Okta’s most important competitor is Microsoft (MSFT), who bundles the rival product Entra ID into Microsoft 365.

Across the security stack, Okta’s peers are CrowdStrike (CRWD) and Palo Alto (PANW), which are consolidating identity into broader platforms. Zscaler (ZS) sells adjacent zero-trust access.

If Okta’s business tumbles, Microsoft captures the workforce economics through Entra, while CrowdStrike and Palo Alto absorb the security-platform budget.

Downstream sit enterprise security teams in every vertical and, increasingly, AI-agent deployments.

Okta's edge is switching costs plus neutrality.

Once identity is wired into every application, directory and workflow, ripping it out is disruptive and risky, which keeps customers in place. On neutrality, Okta positions itself as the independent identity layer that works across every cloud and does not compete with its customers' own applications, a claim Microsoft cannot credibly make.

As shared in our morning systematic portfolio emails, Okta currently weighs 4.5% of the Systematic Portfolio, ranked 7th out of the 8 current holdings.

Here is a simulated history of Okta’s multi-model signals. Okta’s long/flat position is currently determined by the votes among the ML, Trend, and Regime models (this has proven to be the most robust combination for Okta based on backtests, but not necessarily for other tickers). In this case, the Neural model (currently Risk On) is a spectator in the position voting, but is still very relevant for qualifying Okta as a portfolio candidate.

An immediate observation is that the ML model has mostly sat out of the past two years’ price action. It is also flat today. However, in the past quarter, the Trend and Regime models have been mostly Risk On, keeping Okta mostly long and capturing a lot of the rally since May.

Given Okta rides comfortably above its major moving averages, the model that will sway Okta from Long to Flat next is likely the Regime model. The Regime model has not flashed any warnings since the beginning of July.

The debate that decides Okta’s stock is simple. Okta is the leading independent choice for identity and access management, yet Microsoft gives away a capable rival inside its Office bundle, so the question is whether Okta can grow into new identity work faster than Microsoft can absorb the old work.

Base case: revenue grows from the guided 9% to 10% and settles as a low-teens growth story, with the agent-identity narrative slow to show up in the numbers. As growth normalises, the price premium fades.

Holding about 6x forward sales on roughly $3.5 billion of revenue, plus the $2.2 billion of net cash, values the equity near $132, about 5% below today. Investors benefit from the earnings growth, but risks include multiples flattening or even contracting.

Bull case: non-human and agent identity, governance and privileged access reaccelerate bookings, and net retention pushes above 110%. Revenue reaccelerates toward $3.7 billion and the market pays about 8x forward sales, roughly $180, or +30%. The trigger is two straight quarters of current-RPO growth, the contracted revenue due within twelve months, running above the mid-teens.

Bear case: Microsoft Entra bundling accelerates workforce displacement, growth slips to mid-single digits, and the premium unwinds toward the cheaper end of the cyber group near 4.3x sales. That implies roughly $95, or -31%. The trigger is current-RPO growth guided below 8% or a soft first look at fiscal 2028.

The move in the past month has been almost entirely about the multiple, not the business. A wave of analyst upgrades in July tied to the AI-agent identity story lifted the stock to the 52-week high of $154.62.

The latest quarterly growth beat the previously expected 9-10% YoY growth, which helped a very positive earnings reaction.

The remaining performance obligation (RPO) is the total contracted revenue not yet recognised, and it grew 16% last quarter to $4.72 billion.

Current RPO, the portion due within twelve months, grew 12% to $2.5 billion, ahead of the 9% reported revenue growth. This suggests that demand is running slightly faster than the latest growth trend. Net revenue retention, which measures how much existing customers grow their spend over a year, sits at 107%.

The highlight of the quarter isn’t the numbers, but the management devoting their entire focus to an Agentic AI future in which machine identities outnumber human ones.

This is an essential framing because Okta’s growth has stagnated over the past three years (see blow). If the Agentic AI argument turns out to be even half true, the company can reaccelerate its growth back to 20%+ (not seen since 2024).

Historical + Forward Revenue Growth Estimate

Okta for AI Agents became generally available in April 2026, so it is not yet reflected in the numbers. Auth0 for AI Agents targets developers embedding agents into products. According to the management, over 90% of surveyed customers have agents in production, but only 22% are confident those agents are governed. This is not a surprise given the recent headlines of OpenAI, Anthropic, and Meta’s rogue AI agents hacking other companies.

Guidance:

However, the guidance above does not bake in meaningful revenue from AI Agent products. My read is that the management is deliberately conservative.

Okta’s gross and net margins have improved by 100bp each vs a year ago. This is good to see.

SBC declined in the quarter but is largely behind revenue growth, which is good for margin management.

The company repurchased and retired just over 3 million shares for $241 million in Q1, leaving $680 million of the $1 billion authorisation launched in January. The fact that management has deployed share buybacks supports the share price appreciation.

Here are the key financial metrics from the past four quarters. I designed the table this way so we can focus solely on the key top-line, profit, and cash flow metrics.

Here are Wall Street's forward estimates. These growth estimates are in line with the management guidance and have not baked in meaningful Agentic AI acceleration, suggesting that the Agentic AI theme is more of a 2027 story than this year’s.

Okta’s valuation multiples have expanded significantly since the last earnings, reaching the highest point in the past three years. This can move sideways or continue if revenue growth can follow in the near term.

However, the valuation expansion has been across the board for the cybersecurity sector, and OKTA’s valuation actually looks in line with its peers on a revenue-growth-adjusted basis.

In terms of the implied target, today’s 8x appears expensive relative to history (which we flagged before). If we use the mean valuation of 5.7x from the past 3 years, the implied stock price is $123.

Assuming low double-digit revenue growth in the next 5 years, EBIT margin (GAAP) rising towards 25% in the long term (more than double, after accounting for SBC), and WACC cheapening with a 20% debt structure.

The bullish bet from here is that Agentic AI creates an enormous rise in TAM that OKTA grows by 20%+ and re-rates. Microsoft’s bundling strategy may be less effective in that world too, as AI Agents don’t all require an Office subscription.

Therefore, I would like to think of Okta’s price as a “fair value” in today’s business plus a long-term call option premium on Agentic AI acceleration. The market values that option premium at around $20 today, and it may rise and fall with the deployment / delay in Agentic AI revenue stream.

OKTA has been trending up in a very wide and volatile range since the 2022 low. It is currently pushing the top of the price channel following the latest exuberance in the cybersecurity sector.

From a TA point of view, this isn’t a typical entry point - I would prefer to buy the pullback instead.

The aim of this article isn’t to apply discretionary timing to our Systematic Portfolio - of course, that is anyone’s choice, as I don’t give investment advice.

It is to highlight how the market perceives Okta, the street narrative, the actual cold numbers, and therefore the key risks in owning the name. I am personally comfortable understanding what could go wrong, so I am not surprised when the Signals turn sour or underperform.