Magnite

Magnite delivered a strong second quarter, with contribution excluding traffic acquisition costs up 17% year over year, led by 36% streaming (CTV) growth. Adjusted EBITDA rose 30%, and margins reached 37%, prompting management to raise its full-year growth and profitability outlook.

Why it matters: Magnite's CTV growth is increasingly looking like more than an industry tailwind. Growth well above the broader streaming advertising market, alongside deeper SpringServe adoption, supports the long-term outlook but does not yet signal the durable switching costs needed for sustained advantages. CTV contribution ex-TAC grew 36% to $97 million, while Magnite's largest CTV customers grew in the mid-to-high 40% range. Continued wins with large streaming publishers and Samsung's adoption of SpringServe support Magnite's case for gaining CTV share rather than merely growing with the market. SpringServe increasingly combines ad serving, monetization, data activation, and demand access within one, easy-to-use operating system. Deeper integration could make Magnite harder for publishers to replace, but sustained share gains and stable take rates are needed to support that view.

The bottom line: We maintain our $21 fair value estimate and no-moat rating for Magnite. Shares trade near our fair value estimate, leaving limited upside despite improving execution. The investment case increasingly rests on Magnite sustaining CTV share gains and SpringServe differentiation while preserving take rates as competition intensifies. We continue to expect CTV to drive most growth. While stronger share gains create upside to our CTV forecasts, DV+'s weaker trajectory and the unresolved risk of CTV take-rate compression keep our long-term companywide growth and margin expectations broadly intact.

Our $21 fair value estimate implies an enterprise value of 4 times our expected fiscal 2026 sales. The main drivers of Magnite’s revenue include growth in CTV advertising inventory and pricing, the expansion of programmatic digital display advertising budgets, the appeal of Magnite’s SSP and ad server solutions, and greater long-term viability in CTV than what Magnite has experienced on the web.

To develop our revenue forecasts, we estimate Magnite’s penetration of the programmatic digital advertising market and overall CTV ad spending. We believe total CTV ad spending is approximately $33 billion and expect it to grow at an average annual rate of 12.5% over the next five years. Within our expectations for increasing CTV inventory, we balance our bullish thesis that consumers will increasingly utilize CTV and publishers may shift some monetization away from integrated "walled garden" tools, with our bearish thesis that walled gardens will continue to dominate and that supply path optimization products like OpenPath will pose a viable disintermediation threat to Magnite’s business model. We also incorporate our perspective that artificial intelligence overviews will reduce monetization opportunities on websites. We then create a baseline projection that includes these views, resulting in Magnite maintaining our estimate of 5% CTV penetration through 2030. This leads to 4% revenue growth over the next five years, a decline from roughly 7.5% average annual growth from 2023 to 2025.

The cost of revenue is the largest expense, primarily comprising cloud hosting fees, traffic acquisition costs, and salaries. We anticipate moderate improvement in the cost of revenue as a percentage of net revenue, decreasing from 38% to 37% by 2035. This slight improvement should result from the platform benefiting from the asymmetry between marginal costs and the marginal revenue opportunities created by adding a publisher. Some of this reduction also stems from our belief that server utilization rates will improve, thanks to co-located workloads within the same server container—leveraging ownership of both ad decisioning in ad servers and auction logic with the SSP. Sales and marketing are the next largest expenses, primarily driven by salaries and advertising costs. Due to its position within a potentially commodifiable layer of the digital advertising supply chain, we anticipate that publisher relationship-building expenses will remain relatively stable as a percentage of net revenue, decreasing only slightly to 24% from 25% by 2030.

Technology and development expenses consist primarily of salaries associated with the ongoing development of the SSP software stack. These expenses have averaged approximately 15% of revenue since 2022, and we expect gradual compression to 11% by 2030 as the platform scales, repurposes existing code, and uses artificial intelligence code assistants for new product development. Overall, this cost structure results in non-GAAP EBITDA margins ranging from 33% to 37% from 2026 through 2035.

In our view, Magnite does not have a moat. Magnite occupies a commoditized layer of the ad supply chain, lacks aggregation benefits, and faces scaled alternatives, which should compress returns on capital for the foreseeable future.

Historically, Magnite’s ability to outearn its cost of capital has been mixed. From 2017 through 2025, we estimate adjusted returns on invested capital exceeded the weighted average cost of capital only twice, and ROICs are on a downtrend, decreasing from 13% in 2017 to 6% in the most recent fiscal year. Overall, Magnite’s historical ability to outearn its cost of capital has been trending in the wrong direction.

Had it not been for diversification into high-growth connected television through the 2021 acquisition of SpotX, we believe that Magnite would have been in existential danger. That’s due to its original tilt toward the open internet, which includes websites outside of so-called walled gardens like Meta or Google Search. Website advertising was commoditized as publishers adopted “header bidding.” Header bidding is a programmatic technique in which publishers can simultaneously offer their inventory to multiple ad exchanges and supply-side platforms like Magnite, forcing SSPs to compete almost exclusively on publisher yield (the ability to attract the highest price per impression for the lowest take rate.)

We believe supply aggregators like Magnite are fundamentally disadvantaged in the ad market. The adoption of header bidding in the late 2010s led to SSP offerings becoming commoditized. Before header bidding, the advertising supply chain was siloed, requiring deep technical integrations between the SSP and publishers (switching costs), and most transactions took place in private auctions that the SSP owned and managed, allowing SSPs to accumulate proprietary data on auction dynamics. These advantages led to stickier publisher relationships and a data moat, culminating in higher take rates. SSPs with the strongest publisher relationships were also prioritized within a waterfall-style auction, where publishers would rank their preferred SSPs and give the top SSP the first look at monetizing the inventory, the next SSP the next look, and so on.

SSPs can add value through dynamic floor pricing strategies or latency enhancements, but we believe that yield maximization algorithms are especially replicable in a post-header bidding world, and that latency wars are expensive. Google already holds the top bid-clearing latency advantage due to the inextricable link between its ad servers (which account for 90% of the market share) and its ad exchange (which holds a 50% share).

If a moat were to appear in the future, it would likely be tied to Magnite’s strategic foothold in CTV. CTV has higher CPMs, which flow into top-line growth and operating profit for Magnite. We expect CTV to grow at a healthy clip over the coming decade, as it remains undermonetized relative to the time consumers actually spend streaming. CTV has fewer structural disadvantages for SSPs, as header bidding-style competition is less pervasive due to streaming applications existing in closed operating systems, which lack the shared technological framework that facilitated the commodification of SSPs on the web.

However, there is still competition among various SSPs within CTV inventory auctions via Prebid for Video, an open-source framework that allows publishers to run auctions among multiple SSPs (much like header bidding on the web). We believe that technical integrations with SSPs in the CTV ecosystem are more complex than those in the web ecosystem, which makes immediate SSP commodification unlikely. However, it still leaves room for take-rate compression to the point where competition can reduce returns on invested capital.

By emphasizing connected television monetization, Magnite could benefit disproportionately from streaming platforms' growing use of programmatic advertising to generate returns on content investment.

Magnite’s integration of its SSP with its ad servers combines downstream yield optimization with upstream ad inventory management, creating a streamlined environment for a complex ad ecosystem–exactly what publishers need.

Potential antitrust remedies could diversify publisher inventory across additional SSPs, enabling Magnite to capture more opportunity.

Artificial intelligence is disrupting how humans consume content on the web - pushing down the value of ad inventory and creating a headwind to the revenue Magnite derives from the website advertising.

Walled gardens, such as Google and Meta, account for most digital media consumption and ad spending. Magnite doesn’t operate within these closed ecosystems, which limits market opportunities.

The Trade Desk created OpenPath to enable a direct connection between advertising budgets and publishers. Supply path optimization tools like this can disintermediate SSPs such as Magnite.