Reddit Deep Dive

What the market has actually been trading over the past two years is three layers of narrative stacked on top of each other. The first layer is the IPO narrative: a community platform founded in 2005 that long struggled to turn a profit finally went public in March 2024 at $34 per share and closed up 48% on its first day, which itself carried the flavor of a scarce asset returning to public markets. The second layer is the earnings inflection: the 2024 statements recognized IPO-triggered stock compensation as a one-time charge, blowing the net loss out to $484 million; but after the stock compensation impact clearly receded in 2025, the company swung to a full-year net profit of $530 million, with adjusted EBITDA of $845 million, operating cash flow of $691 million, and free cash flow of $684 million. The third layer is the AI narrative: Google and OpenAI are both paying customers for Reddit's content and powerful platforms that could reshape the search distribution entry point. In 2024 the market was willing to pay a high price for "AI training data is the new oil," and in 2025 it quickly shrank its tolerance because of Google's algorithm and AI search eroding click distribution.

Reddit's past stock rises and falls have never strayed from this main line. After the IPO, the market first valued it as a "scarce high-growth UGC asset"; then, seeing Q4 2024 revenue of $428 million and revenue climbing through Q2 to Q4 2025 to $500 million, $585 million, and $726 million with profits rising in step, the stock kept re-rating; by the time the Q4 2025 results came out, even though revenue and profit were both strong, DAUq came in slightly below expectations, management explicitly admitted that a Google algorithm change had once hit search visibility, and the stock dropped sharply in a single day. In September 2025 RDDT closed at an all-time high of $270.71; by 2026-06-12 the stock was back to $162.10, a pullback of about 40% from the peak. The real reason is that the market is no longer willing to pay the kind of extreme growth premium of 2025 for opaque search distribution risk.

The most important bull-bear divide for this stock right now is whether AI shows up first as incremental revenue or first as traffic erosion for Reddit. The optimists see that Reddit owns a public corpus of more than 2.5 billion posts and comments, which management calls "the oil of the modern internet"; even though Q1 2026 other revenue was only $39 million, its cash gross margin is extremely high and it remains a second curve with enormous strategic value. The pessimists see that Reddit's web traffic is still heavily influenced by third-party search; Similarweb estimates that about 60.29% of desktop visits to reddit.com come from organic search; and Pew Research found that on search result pages carrying a Google AI summary, users click through to traditional web links only 8% of the time, versus 15% when there is no AI summary. If AI Overviews and AI Mode keep turning the act of "clicking into Reddit to read human experience first" into "getting a compressed answer directly on the search page or in a chat box," what is most fragile for Reddit is the top-of-funnel that advertising relies on to grow, not licensing revenue.

My qualitative portrait of Reddit is: in the middle of a valuation reset. It has already crossed the threshold of "can it make money," and has even revealed the rare characteristics of being capital-light, high-margin, and fast at converting cash; but it has not yet crossed the threshold of "is growth truly controllable." In Q1 2026, the company's net margin was 31%, adjusted EBITDA margin was 40%, and operating cash flow margin was 47% - numbers that are not common among social media peers. At the same time, Q1 2026 global DAUq grew only 17% year over year, the U.S. only 7%, while logged-out users grew 26% year over year, markedly faster than the 7% for logged-in users. This means Reddit is running fast right now, but the speed does not come entirely from deeper user habits; a substantial part also comes from a more fragile search entry point and a shallower visit layer. The company has proven its monetization efficiency but has not yet fully proven its traffic sovereignty.

The Company's Vertical History

Reddit appeared in 2005 because the internet at the time lacked a place that "organized content by interest rather than by relationship." The company began operating in 2005, was acquired by Condé Nast in 2006, and only spun out of the Advance/Condé Nast system in 2011 to become an independent private company again. This path matters: Reddit found its product form very early but for a long time was not pushed into serious commercialization. The S-1 bluntly states that it did not begin "meaningful monetization efforts" until 2018. This explains why Reddit's asset accumulation came long before its revenue explosion, and why the story it told capital markets when it listed was one of "dormant inventory finally starting to be monetized" rather than a new-platform story.

The company has two most important long-term legacies from its early days. The first is the subreddit structure. The S-1 defines Reddit as a "community of communities," where each subreddit is a customizable small community, posts and comments together create value, and in many contexts comments matter as much as the original post. The second is its search attribute. Reddit was not discovered by search only later. In 2017 the company publicly reviewed the 12-year evolution of its internal search, from full-text search using Postgres in 2005, to PyLucene in 2007, Solr in 2008, outsourcing to IndexTank in 2010, migrating to CloudSearch in 2012, and rebuilding into a more modern search stack in 2017. By December 2023, Reddit already averaged more than 35 million search queries per day on its own site. This shows that "Reddit = a forum found via Google" is only half the truth; the other half is that the company itself has been turning massive historical content into a searchable knowledge base all along.

The real turning point came after Steve Huffman returned in 2015. The S-1 shows that Huffman has served again as CEO since July 2015; CTO Christopher Slowe also returned in late 2015 and later took charge of technology again. The effect of this management combination was to slowly remake the company from a "community product" into a "scalable platform company." Starting in 2018, Reddit finally began commercializing in earnest; by 2023, it had written "the public archive of modern internet human conversation" into its prospectus and begun restating its value from a traffic asset to a data asset. Around 2024, large-model companies such as Google and OpenAI partnered with it, turning this abstract value into cash for the first time.

Reddit's listing path also carried its own community flavor. The company priced its IPO at $34 per share on March 20, 2024, issuing 22 million Class A shares for total proceeds of about $748 million, and began trading on the NYSE under the ticker RDDT on March 21. More unusually, Reddit set up a directed share program for users and moderators, allocated across six priority tiers by karma, moderation contributions, and the like, and these IPO-purchased shares were not subject to lockup restrictions. On one hand this made the IPO narrative more "Reddit," and on the other it naturally added to the float and volatility in the early period after listing. The stock closed its first day at $50.44, up 48%, and the market quickly lifted it from its $6.4 billion offering valuation to a market cap of nearly $9 billion.

The four stages after listing are very clear. The first stage was "telling the story." In 2024, capital markets were buying a scarce platform, an open corpus, and AI partnerships. The second stage was "proving it can make money." Q4 2024 revenue of $427.7 million, net profit of $71 million, and adjusted EBITDA of $154.3 million let the outside world see scaled profit for the first time. The third stage was "proving growth is not a one-off." From Q2 to Q4 2025, revenue climbed from $500 million to $726 million and net profit rose from $89 million to $252 million. The fourth stage was "being asked to prove the quality of growth." In early 2025, because a Google algorithm change hit logged-out traffic, Reddit's financials were strong but its user data came in slightly below expectations and the stock slumped; by Q1 2026, the market's focus was no longer "can Reddit make money" but "can it keep control of the user entry point after AI changes distribution."

Pulling out the financial history on its own, Reddit's "earnings inflection" has to be split into two layers. On the surface, a $484 million net loss in 2024 and a $530 million net profit in 2025 is a dramatic reversal; but the real structural significance lies in three lines holding true simultaneously. First, 2025 revenue growth was still as high as 69%, showing the profit improvement was not bought by cutting growth. Second, 2025 operating cash flow was $691 million and free cash flow was $684 million, with capital expenditure of only $6.7 million, showing that once this company scales, cash conversion is very fast. Third, the enormous 2024 loss included $843 million of stock compensation and related taxes, mainly tied to the satisfaction of liquidity conditions on IPO-triggered RSUs; in 2025 this dropped to $387 million and the income statement regained readability. In other words, the most misleading thing about the 2024 report is that it disguised a one-time accounting charge as "the business still doesn't work"; the most valuable thing about the 2025 report is that it exposed Reddit's true profit structure.

Looking at the balance sheet, Reddit also does not look like a typical "cash-burning social stock." As of Q1 2026, the company had $2.77 billion in cash, cash equivalents, and marketable securities; as of year-end 2025, it had about $494.7 million of undrawn revolving credit, with only a small amount of letters of credit and no real interest-bearing borrowing pressure. What really needs watching is two kinds of soft constraints rather than debt: first, at year-end 2025 the company's non-cancelable purchase commitments to third-party cloud infrastructure totaled about $372 million; second, platform governance relies heavily on moderators and community rules, an operating model with extremely low capital expenditure but not low governance expenditure.

The stock-price history is more honest than the financial reports. Closing its first day at $50.44 in March 2024 shows the market was initially betting on the "scarcity narrative"; in 2025, driven by strong results and the AI licensing theme, the stock surged all the way to its all-time closing high of $270.71 on 2025-09-18, with its market label shifting from "new IPO" to "high-quality growth platform"; the valuation then gave back quickly, falling to $162.10 by 2026-06-12. The current price is still far above the IPO first-day close but about 40% below the peak. This trajectory looks a lot like a stock that was just discovered by capital markets and then quickly asked to hand in a second round of answers. The story is never lacking; the hard part is proving the story won't be rewritten by Google.

Business Model, Industry, and Horizontal Competitors

Reddit's business machine runs clearly. Advertising is the main body, and content licensing is a second curve that is very valuable at the margin but still small in scale. In 2025 advertising revenue was $2.062 billion, up 74% year over year; other revenue was $140 million, up 22% year over year. By Q1 2026, advertising revenue rose another 74% year over year to $625 million, while other revenue grew only 15% to $39 million. Management added on the earnings call that performance-oriented revenue already made up more than 60% of advertising revenue in Q1 2026, that conversion-oriented lower-funnel revenue was still growing triple digits year over year, and that active advertisers grew more than 75% year over year. This set of data is key: Reddit's ad business has moved from "having communities so it can sell ads" to "having clear commercial intent so it can sell performance ads." The licensing business, by contrast, looks more like a high-margin strategic chip than the main engine today.

The cost structure is equally clear. Reddit's gross margin has long stayed around 90%, at 91.2% for full-year 2025 and 91.5% in Q1 2026. Capital expenditure is extremely low, only $6.7 million in 2025, almost negligible; the company itself explains in its annual report that free cash flow is meaningful precisely because it relies on third-party infrastructure and does not need to pair revenue growth with heavy capital investment. For investors, this brings two consequences. The good consequence is that operating leverage is very pretty - once revenue is lifted, profit and cash flow can run even faster. The bad consequence is that you cannot only look at "asset-light"; you also have to look at its dependence on external distribution channels and cloud infrastructure. The former determines the quality of growth, the latter determines fixed commitments.

The moats that truly hold for Reddit, in my view, are only three. The first is a "structured library of human experience." This refers to the topic-based knowledge organization formed jointly by subreddits, posts, comments, votes, and rules. The second is "high-intent traffic." Management discloses that about 40% of Reddit conversations carry commercial attributes, and that 84% of shoppers say they feel more confident about a purchase decision after researching on Reddit; this makes it more valuable in performance advertising than many platforms that rely only on broad-entertainment dwell time. The third is "supply efficiency from community self-governance." Moderators are volunteers and community rules are customized by communities, which lets Reddit support more than 100,000 active communities at relatively low capital cost.

But Reddit also has "false moats" that the market over-mythologizes. The most common is "data exclusivity." Reddit's public corpus is indeed scarce, but the annual report already says plainly that almost all current licensing contract value comes from two mid-term partners, that these contracts may not be renewed, and may be renewed at lower prices and with fewer services; meanwhile, some companies still use Reddit content without a contract. Put another way, the data value is real, but data pricing power is far from secure. Another exaggerated moat is "Google search favors Reddit." This looks more like an external dividend of a tailwind period than a controllable company asset. The company itself was hit once by a Google algorithm change as early as Q4 2024.

On governance, Reddit has to take a discount. The company uses a multi-class share structure: Class A has one vote per share, Class B has ten votes per share, and Class C carries no voting rights; only Class A is the publicly traded RDDT. The annual report states directly that this structure lets pre-IPO holders continue to control significant voting power and limits ordinary shareholders' influence over major matters. Add to that the company's very short listing history, which means the verifiable sample of management's capital allocation that outside shareholders have is not long, and the governance discount here is a reality that should be written into valuation assumptions rather than an abstract concept. The good side is that the company authorized a $1 billion buyback in February 2026, showing that management is at least willing to send the signal at the capital level that "we have crossed the cash-consumption period."

Put Reddit back into the industry and it sits in a very special position. It does not depend on a social graph and massive display advertising the way Meta does, nor does it aim from the start at "discover - save - buy" the way Pinterest does, still less does it build communication scenarios around cameras and young users the way Snap does. Reddit's position looks more like an "intent network between search and social": users often come here with a question rather than with a relationship. So what it competes for is a portion of search traffic, a portion of performance ad budget, and a portion of AI training data budget. This niche has advantages and disadvantages. The advantage is that content is more long-tail and longer-lived, and comments themselves keep replenishing the inventory. The disadvantage is that the entry point is not firm, especially on the web. Similarweb estimates that about 60.29% of reddit.com desktop traffic comes from organic search; and the company's Q1 2026 data shows logged-out DAUq of 74.8 million, up 26% year over year, markedly faster than the 52.0 million and 7% for logged-in. This practically writes "strong top-of-funnel, weak deep relationships" right into the quarterly report.

Looking across four companies, Reddit's group position is clear:

Sources: latest-quarter earnings releases and market quotes for Reddit, Meta, Pinterest, and Snap.

The business reasons behind this table matter more than the numbers themselves. Meta's strengths are scale, a closed-loop of data, and the maturity of its ad system; Reddit can hardly catch up on total size, but it has its own uniqueness in the advertising scenario of "high-intent human-experience answers." Pinterest is the public comparable most worth contrasting, because it likewise serves high-intent discovery and is likewise using AI to remake advertising; the difference is that Pinterest's content leans toward visual inspiration and consumption decisions, while Reddit leans toward text discussion and genuine feedback. Snap's significance lies in reminding investors that a social advertising company, even with many users, can be stuck for a long time below the profit-realization threshold; the reason Reddit's current valuation is not low is precisely that the market believes it has already crossed the line Snap has never stably crossed.

If we add "AI data holders" into the picture, Reddit looks different again. Companies like Thomson Reuters sell high-value, strongly copyrighted data embedded in professional workflows; News Corp represents publishers rediscovering bargaining power over their existing content because of AI training. Reddit is neither as closed and professional as TRI nor as traditional and copyright-heavy as NWSA; what it sells is the largest, fastest-updating, most topically mixed, and most emotionally rich human conversation on the open internet. This determines that the ceiling of its licensing business could be very high, but that its contract stability, pricing logic, and legal boundaries are all less mature than those of professional databases.

Current Fundamentals and the Bull-Bear Divide

First, put the most recent five quarters side by side:

Source: Reddit's official quarterly earnings releases.

This set of numbers shows three things. First, growth did not collapse after profit was released; instead it accelerated further in the second half of 2025, and Q1 2026 still held 69% revenue growth. Second, advertising is the absolute main engine, and the Q1 2026 acceleration came mainly from advertising, not from licensing. Third, the reason margins could lift so quickly is that Reddit's incremental advertising drops almost straight down at a gross margin above 90%. In the 2025 report, the company attributed revenue growth to an increase in ad impressions, followed by price increases; on the Q1 2026 earnings call, management emphasized that ad growth was driven jointly by impressions and pricing, and that lower-funnel, performance ads already exceeded 60% of the mix. This is the result of the ad product becoming more like a performance advertising platform, rather than being squeezed out purely by raising ad load.

What the market is trading now is no longer just "Reddit is an AI corpus" but a "triple re-rating": one, the ad business is moving from the edge of brand budgets into a more measurable lower funnel; two, the company has proven it can not only make money but also generate unusually good cash flow; three, AI on one hand raises licensing value and on the other pushes Reddit Answers, search, and shopping products to internalize user intent. On the Q1 2026 earnings call, management spoke of Reddit Max ad automation, search DAUq/WAUq and query volume growing "meaningfully year-over-year," search being one of the drivers of retention and DAUq, and the product search catalog being connected into answer pages. This shows the company is actively trying to turn "Reddit inside external search" into "search and answer products inside Reddit."

The problem lies right here. The most important bull logic the market sees right now has four points. First, the Q1 2026 ad trend is very hard: revenue up 74% year over year, active advertisers up more than 75% year over year, and performance revenue already more than 60% of ad revenue, showing Reddit is moving from "a distinctive traffic platform" to "an ad platform that can deliver performance." Second, U.S. daily actives are only 53.5 million while management's explicit target is 100 million, and U.S. weekly actives are close to 200 million, so lifting usage frequency rather than expanding awareness is logically more feasible than pure new-user acquisition. Third, internationalization is still early - Q1 2026 international DAUq grew 26% year over year, far above the U.S. 7%, and international revenue grew 76%, showing that machine translation and product localization are widening the ceiling. Fourth, content licensing, though small, gives Reddit non-advertising pricing power in the AI era. Even if it is only a few points of revenue today, that is enough to change how the market views this company's "asset nature."

The bear logic is also hard, and harder to dismiss with a single "AI is an opportunity." First, Reddit's most fragile growth comes precisely from its most fragile user group. In Q1 2026 logged-out DAUq grew 26% year over year, logged-in only 7%; U.S. DAUq grew only 7%. If external search distribution is rewritten by AI, the first to come under pressure will be this top-of-funnel. Second, the Google algorithm change already proved in Q4 2024 that it can hit Reddit; although management says such changes usually do not significantly affect traffic over the long term, the market will not easily forget this one. Third, Pew Research data already shows that when an AI summary appears, the probability of users clicking web links is nearly halved; Reuters Institute has also documented publishers' concerns about AI search diverting web traffic. For any website that depends heavily on search distribution, this is not an abstract risk. Fourth, licensing revenue is concentrated. The annual report says plainly that almost all current contract value comes from two partners, revenue that is strategically precious but financially still unstable.

There is also a detail the market easily overlooks: starting in Q3 2026, Reddit will stop disclosing logged-in and logged-out DAUq, keeping only U.S./international DAUq and WAUq. For the company, this is simplifying disclosure; for investors, it is one less layer of transparency on the most sensitive question. And what most needs to be verified right now is precisely whether traffic quality has improved. A reduction in disclosure dimensions does not by itself mean bad news, but it will cause the market to factor more uncertainty into the valuation.

Valuation, Risk, and Tracking Metrics

First, a cash-flow walk-through. In 2025 Reddit's operating cash flow was $691 million, free cash flow was $684 million, and capital expenditure was only $6.7 million; in Q1 2026 operating cash flow was $312 million, free cash flow was $311 million, and capital expenditure was about $1 million. Hedging full-year 2025 against Q1 2026/Q1 2025, TTM revenue is about $2.47 billion; valued against the current price of $162.10 and market cap of $32.83 billion, the current market-cap-basis P/S is about 13.3x, the net-cash-adjusted EV/Sales is about 12x; the market-given TTM P/E is about 46x; and on a rough TTM free cash flow basis, the FCF yield is about 2.5% to 2.7%. This is not a cheap price, nor is it the kind of extreme premium seen at the 2025 peak, but it is still a valuation built on continued high growth.

Historically, RDDT's valuation center has already been rewritten once, clearly. At the IPO, the market gave it an imagination premium of "scarce new stock + AI theme + not yet stable profit"; at the 2025 peak, the market gave it the growth-stock label of "high growth, high margin, low capex"; and by June 2026, with the P/E back to over 40x, it is essentially in the de-bubbling middle ground after the 2025 surge. Using Macrotrends data, the stock has pulled back clearly from the all-time closing high of $270.71 in September 2025; but measured against revenue, profit, and cash flow, the current valuation is again far above the IPO period. In other words, the valuation has been killed, but not killed down to the point of being "viewed only as an ordinary advertising platform."

The peer comparison does not yield a "cheap" conclusion. Meta is clearly cheaper, with a P/E of about 20.6x, but its growth scale and cash-flow scale are simply not in the same league; Pinterest's P/E is about 42x with revenue growth of 18% and 631 million users, and the price the market gives it comes more from e-commerce discovery and AI ad optimization; Snap is still at a GAAP loss with a lower market cap, showing the market will not automatically give a social platform a high premium just because it has "many young users and an improving ad business." Reddit's current valuation premium reflects "growth markedly faster than Pinterest/Snap, a structure lighter than Snap, and a narrative newer than Pinterest." This premium has its reasons, but it is not a margin of safety.

Below is a three-scenario valuation under a research framework. Here I use 2027 free cash flow as the main anchor, supplemented by EV/Sales cross-validation, because Reddit is no longer a loss-making platform you can only judge on revenue, but is also not yet mature enough to be judged on static P/E alone.

The above is only a valuation scenario under a research framework and does not constitute investment advice. Valuation inputs come mainly from the company's 2025 annual report, the Q1 2026 release, and the current market cap.

Pulling out these three scenarios and the margin of safety on their own, the conclusion is clear. At the current $162.10, there is no discount relative to the conservative scenario of $110-130, and the margin of safety is zero. The base scenario can barely cover the current price, but the return range it offers is not thick. The most fragile assumption is that "search distribution changes will not weaken U.S. user growth and ad quality in step," not licensing revenue. If I haircut this assumption in the base scenario by 30%, I calculate fair value falls back to around $135. Run one more plainer stress test: if profit grows zero over the next three years and the valuation does not expand either, the investor's annualized return is roughly close to 0%. This is not a bad company, but the price right now does not carry a wide margin of safety. My conclusion on margin-of-safety sufficiency is: no.

Among the risks, the ones that could truly cause permanent capital loss are, in my view, five. First, search distribution being rewritten by AI. Probability medium-high, impact high. What deserves the most attention is U.S. DAUq, logged-out growth, and search-ecosystem commentary, not site-wide traffic. Second, licensing revenue is concentrated and its legal boundaries are undefined. The annual report already flags that almost all contract value comes from two partners, and the company is suing Anthropic, showing that "who is willing to pay, who will steal, and who can be stopped" is still taking shape. Third, the ad cycle and budget visibility. On the Q1 2026 earnings call, management said clients have not pulled budgets broadly, but they do lean more toward monthly decisions, and visibility has declined. Fourth, the governance discount. What ordinary shareholders buy is one-vote Class A, not control. Fifth, although stock compensation is already far below the 2024 anomaly, 2025 still had $387 million of stock compensation and related taxes, which will keep diluting the "cheapness" of the valuation.

I will keep tracking the following dashboard:

Among these metrics, the most critical are really U.S. user frequency and traffic quality, not margins. As long as U.S. DAU stays stuck, logged-out keeps running far ahead of logged-in, and the company no longer discloses this breakdown, the market will re-price it as an "external traffic platform" rather than an "internal high-frequency platform." Conversely, as long as U.S. frequency starts to accelerate, on-site search and Answers growth keeps getting quantified, and licensing renewals are smooth, Reddit's valuation ceiling will lift again.

There are mainly four research uncertainties. First, Reddit does not publicly give a true source-by-source breakdown of search referral share; Similarweb's 60.29% is only a desktop estimate. Second, the full financial terms of the Google and OpenAI contracts are not public; all the market can see is the revenue result and management's qualitative comments. Third, disclosure of usage data for Reddit Answers, on-site search, and shopping integration is still very limited. Fourth, the company has been listed for a short time, and capital markets have not yet truly stabilized its "normal valuation band."

Main reference sources include: Reddit's 2024/2025 annual reports, the 2024 S-1, the 2024 IPO pricing announcement, the official earnings releases for Q1-Q4 2025 and Q1 2026, and the Q1 2026 shareholder letter and earnings call transcript; Reuters reporting on the IPO, Google/OpenAI licensing, the Anthropic lawsuit, and stock-price volatility; Pew Research's study on click-through rates for Google AI summaries; Similarweb's estimate of reddit.com traffic sources; and the latest-quarter disclosures from Meta, Pinterest, Snap, Thomson Reuters, and News Corp.

The Vertical-Horizontal Synthesis

Vertically, what Reddit has truly proven is the ability to slowly grind a long-undervalued community asset into a business system that is profitable, replicable, and able to charge in the AI era. Its strongest point is content form, not total users: anonymous, genuine, contentious, organized by topic, with comments carrying more information density than posts. Once such content crosses the scale threshold, it can naturally serve advertising, search, recommendation, and model training all at once. Over the past two decades, libraries of this kind of public and continuously updated human conversation have not grown more numerous on the internet; instead they have become ever scarcer amid closed ecosystems and the short-video wave. The reason Reddit can earn such high margins, run such low capex, and convert cash so fast today is all rooted here.

But horizontally, its weaknesses are not short-term either. Meta's entry point and ad system are both firmer, Pinterest's consumption intent is cleaner, and Snap's youth-oriented social relationship chain is more natural. Reddit's advantage is "human-experience density"; its weakness is "incomplete entry-point sovereignty." As long as users much of the time still first search "problem + reddit" on Google rather than opening Reddit directly, searching inside Reddit, and leaving deeper behavior inside Reddit, Google remains its most important and most dangerous upstream. Google and OpenAI are simultaneously its partners and potential traffic diverters, and this tension is the most distinctive thing about Reddit as a company and the hardest for any linear valuation method to truly digest.

So my final judgment on this stock is very clear: the company's quality is higher than many thought at the 2024 IPO; the stock price, as of June 2026, is still not cheap enough to ignore those structural uncertainties. What the current valuation rewards is no longer just the past year's beats on advertising and profit, but also an early bet on two futures: one, that U.S. DAU frequency can keep improving, and two, that AI search will not permanently weaken Reddit's top-of-funnel, and may even conversely steer more high-intent questions toward on-site search and Answers. As long as one of these two futures falls through, the base valuation will drift downward; only if both are delivered will the optimistic valuation be sustainable.

I summarize the bull case in five points:

• The ad business has evolved from a marginal role in brand budgets into a performance advertising platform where performance makes up more than 60% of the mix.

• The company's gross margin is above 90% and capex is extremely low, and both 2025 and Q1 2026 proved very strong operating leverage.

• U.S. weekly actives are about 200 million and daily actives 53.5 million, so the ceiling for frequency improvement is still very high.

• Internationalization has just entered its realization phase, with International DAUq and revenue growth both markedly above the U.S.

• Content licensing, though not large, gives Reddit a second kind of pricing power that traditional ad platforms lack.

I summarize the bear case in five points:

• Traffic quality is still on the shallow side, with logged-out growth markedly above logged-in, making it the most exposed to AI search and algorithm adjustments.

• Pew Research has empirically shown that AI summaries compress web clicks, which is a positive risk exposure for a search-oriented platform like Reddit.

• Licensing revenue is highly concentrated among a few partners, with strategic significance greater than financial stability.

• The multi-class share structure means the governance discount persists long-term, and ordinary shareholders cannot influence key decisions.

• The current valuation still requires sustained high growth, and a margin of safety does not exist.

In my pre-mortem, the two scripts I worry about most are these. Script one: by 2027, Google's AI Overviews/AI Mode become the default entry point for more search intent, Reddit's U.S. logged-out traffic sees a mid-double-digit decline, U.S. DAU stays stuck below 55 million for a long time, ad revenue growth drops below 20%, and the market no longer gives it 12x EV/Sales but values it at 7x to 8x, with the stock returning to the $80 to $100 range. The scariest thing about this script is that revenue still grows, but valuation and growth rate move down together - a case of "the company is not entirely broken, but shareholders lose half first." The evidence supporting this worry is the Google algorithm disturbance that already happened in Q4 2024, not imagination, and the real compression of web clicks by AI summaries.

Script two: by 2027, Google/OpenAI cut prices at renewal, or a large client shifts to self-built/alternative data sources; at the same time, lawsuits against unauthorized users like Anthropic do not quickly form an industry constraint, leaving licensing revenue stagnant at a level that is "symbolic but not material in scale." The market then stops treating Reddit as an "AI data asset" and values it only as an ad platform. Even if the ad business is fine, the stock could be compressed from today's over-40x P/E to around 25x. This is the second halving path: rooted in the AI premium not being delivered, not in advertising collapsing.

My final research conclusion is: Reddit has gone from "a newly listed stock telling an AI and community story" to a genuine internet platform with high margins, low capital consumption, and continuously improving ad efficiency. Its problem is no longer whether the business model can hold, but whether the traffic entry point its model relies on to expand can stay stable in the AI search era. Unlike many platform companies, Reddit simultaneously enjoys the licensing revenue AI brings and bears AI's erosion of search clicks. Precisely because both directions are real, it is not the kind of company you can conclude on with a single "AI is good for it."

Near the current $162.10, I think Reddit is no longer absurdly expensive, but it is by no means cheap. It is suitable for a high-quality watch list, and suitable for existing holders to keep holding and track the key variables closely; it is not suitable for chasing a heavy position on the slogan "the oil of the modern internet" without a margin of safety. The truly better buy point should appear when the market is disappointed again over traffic worries but the company's core cash-creation ability has not been damaged. At such a time, the odds will improve markedly.

【Company Profile Score】

• Fundamental quality: High

• Growth: High

• Moat: Medium

• Financial soundness: Strong

• Management credibility: Medium

• Valuation attractiveness: Low

• Risk level: High

• Suitable investor type: Long-term growth

【Investment Rating】

• Rating: Hold

• One-line investment thesis: ad efficiency and cash flow are already proven, but the Google/AI distribution risk has not been adequately buffered into the price.

• Three-tier price signals: Ideal buy price: $110-130

• Acceptable holding price: $145-180

• Clearly overvalued price: $220-250

• Current price classification: acceptable to hold

• Whether it is worth waiting for a better price: Yes. If the price returns below $130 while U.S. DAUq does not stall below 5% for two consecutive quarters, the odds will improve markedly; the opportunity cost of waiting is that if the company proves early that U.S. frequency is improving and on-site search internalization is succeeding, the stock may no longer return to that range.

• Target holding period: 1-3 years

• Expected annualized return: conservative -12% to -15%; base 0% to 5%; optimistic 18% to 22%

• Maximum loss risk: about 40% to 50%; the trigger is Google/AI distribution pressure stepping U.S. traffic quality down while the market stops awarding an AI data premium.

• Signals that trigger reassessment: If U.S. DAUq is below 5% year over year for two consecutive quarters

• If advertising revenue YoY falls below 25% while management talks of continued search-ecosystem deterioration

• If other revenue YoY falls below 10%, or management hints that Google/OpenAI renewal terms are weakening

• If adjusted EBITDA margin is below 30% for two consecutive quarters

• If the company no longer discloses enough traffic-quality signals and deep logged-in usage cannot be verified by other proxy metrics

【Ideal Buy Price】110-130 USD Basis: corresponds to the implied value range of the conservative scenario, and retains enough discount against the single largest risk of "search distribution being rewritten by AI."

【Valuation Range】

• current: 162.10 (as of the 2026-06-12 close)

• bear (conservative · ideal buy zone): [110, 130]

• base (fair · acceptable holding zone): [145, 180]

• bull (optimistic · above the clearly overvalued line): [220, 250]

Other Tickers Mentioned in This Report

• META.US — the scale and margin benchmark for digital advertising platforms, used to gauge whether Reddit's premium is reasonable.

• PINS.US — one of the closest public comparables to Reddit, likewise benefiting from high-intent traffic and AI ad optimization.

• SNAP.US — a social advertising platform comparison set, used to compare user scale, profit realization, and valuation discount.

• GOOGL.US — Reddit's most important search distribution entry point and AI data licensing partner, and also its largest external variable.

• TRI.US — a professional data/content holder, contrasting the difference between Reddit's "public corpus" and a "professional database."

• NWSA.US — a traditional content asset holder, offering another reference for data monetization through AI licensing and copyright bargaining.

• SHOP.US — Reddit's ecosystem partner in the SMB and e-commerce ad conversion chain.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.