- Buys
- 1 of 4
- Three are needed, so Reddit stays out of the basket.
- Revenue growth, third quarter
- 53%
- $895M, on top of nearly 70% growth a year earlier.
- Possible fall in daily users, fourth quarter
- Low double digits
- The company's own warning, blamed on falling search referrals.
- Free cash flow after stock pay, trailing
- $0.81B
- $1.20B reported, less $385.8M of stock compensation.
Quality, decade-long holdingExcellent economics, but it cannot yet defend a twenty-year view of the audience.
Deep value, margin of safetyIts model failed with an error before it could vote.
Macro and reflexivityA regime break with a dated test at the fourth-quarter results. Not paid to guess first.
GrowthThe visitors leaving earn little. The logged-in core that earns most is still growing.
Session of October 31, 2026. Investor B's model failed with an error, so three seats voted. The script sheet missed short-term investments, buybacks, guidance and user figures, so the chair added them by hand from the company's third-quarter shareholder letter. The committee has not reviewed Reddit before.
Bottom line
Reddit reported its best quarter ever and, in the same letter, said its search traffic is draining away. The committee split on which fact sets the value.
On reported free cash flow, the price asks for about 2% a year of growth. Count stock compensation as a cost and the ask rises to about 7%. Fourth-quarter guidance implies revenue growth of roughly 38% to 42%, so neither bar is hard today. The question is how long revenue per user can rise while the top of the funnel shrinks.
Investor D says the visitors Reddit is losing were never worth much. Investors A and C say those visitors become tomorrow's logged-in users. The first test is the fourth-quarter results: did global logged-in daily users hold the third quarter's 53.5 million?
The business in 60 seconds
Reddit is one business: topic communities, moderated mostly by volunteers, with advertising sold against the conversation. Advertising was $847M of third-quarter revenue. The other $48M is mostly data licensing, access to an archive of more than 27 billion posts and comments for companies building AI models (Exhibit 99.2, 8-K filed October 29, 2026).
Logged-in users are about 41% of the daily audience. Together with app users they produce, in the company's words, "over 80% of the monetization". Gross margin was 91.5%.
Class B shares carry ten votes. Under voting agreements signed at the 2024 listing, the chief executive may be deemed to beneficially own about 37.3% of all common stock (Schedule 13D/A, filed August 22, 2024). That figure is two years old; we could not read a current one from the 2026 proxy.
| Fiscal year | Revenue | Operating margin | Free cash flow | Diluted shares |
|---|---|---|---|---|
| 2022 | $666.7M | -25.8% | -$100.3M | 57.3M |
| 2023 | $804.0M | -17.4% | -$84.8M | 59.1M |
| 2024 | $1.30B | -43.1% | $215.8M | 145.5M |
| 2025 | $2.20B | 20.1% | $684.2M | 202.1M |
| 12 months to Sept. 2026 | $3.09B | 29.4% | $1.20B | 201.9M |
10-K filings on SEC EDGAR; only four fiscal years are filed. The 2024 loss includes $801.6M of stock pay at the listing. Last row trailing, from the 10-Qs.
What changed
- A traffic warning. The letter says search referral traffic "was down meaningfully in Q3, with the decline accelerating later in the quarter". October lost more than 1 million referrals a week on average. The letter does not name the search engine.
- Slower guidance. Fourth-quarter revenue of $1,000M to $1,030M and Adjusted EBITDA of $480M to $510M. Against about $725.6M a year earlier, that is growth of 38% to 42%. No 2027 figure.
- Logged-in only. The company will "begin fully testing" a logged-in-only experience "in select markets outside the U.S." this quarter.
- Buybacks began. $339.96M of Class A stock in nine months, against none a year earlier (10-Q, filed October 30, 2026).
- The ad market is strong. Alphabet grew third-quarter revenue 24.7%, Meta 26.1% and Pinterest 11.9% (each company's 10-Q, filed October 22 to 29, 2026). Alphabet runs Google Search and is, by one unconfirmed report, in licensing talks with Reddit.
The bull case on Reddit
- Revenue has already decoupled from users in the U.S. U.S. daily users fell 2% over the year while U.S. revenue rose 50%.
- The users being lost earn little. U.S. logged-out users fell 4%. Revenue per user is $13.82 in the U.S. against $2.43 abroad, so the mix shift lifts the average.
- The international runway is long. International daily users grew 24% to 79.6 million and revenue 67%.
- Survival is not in question. Cash and short-term investments were $3.03B at September 30, 2026, with no debt.
- Growth needs almost no capital. Third-quarter operating income rose 88.8% on revenue up 53%.
The market is treating all traffic units equally.
Investor D, Buy
The market may be confusing user loss with value loss.
Investor A, Watch, arguing against its own vote
The bear case on Reddit
- The funnel that built the logged-in base is closing. Search referrals fed the logged-out audience that becomes members, and login-only tests narrow it further.
- The paying core is barely growing where it matters. U.S. logged-in daily users grew 1%. The U.S. produces 80% of revenue.
- Reported cash flow overstates what an owner keeps. Stock compensation is a third of reported free cash flow, and buybacks have only offset it: fully diluted shares fell 0.3% in the quarter.
- Earnings carry almost no tax. Tax was $11.5M on $751.6M of pre-tax income in nine months, with no date given for a full rate.
- The data option and the traffic problem share a cause. The answer engines keeping readers away are the archive's buyers.
That is the right defence, but it is a defence, and it shrinks the funnel that has fed the logged-in base.
Investor A, Watch
the market is not being handed a value stock, it is being handed an illusion of one, and the bulls citing cheapness are anchoring on the wrong figures.
Investor C, Watch, on the headline multiple
Where the committee splits
One of three seats voted Buy. All three name the same critical environment, AI answers replacing the search click, and read the same user figures in opposite directions.
Do lost visitors matter if the paying users stay?
| Position | Seats | The figure it rests on |
|---|---|---|
| No, yield per user carries it | D | U.S. revenue per user up 53% |
| Yes, visitors become members | A, C | U.S. logged-in daily users up 1% |
What settles it. Global logged-in daily users in the fourth-quarter results against 53.5 million. A and C break below that; D below 53.0 million.
From the chair, to Investor D. Your thesis rests on logged-out visitors being low value, yet U.S. logged-in users grew only 1%. If search visitors are where future members come from, what in the fourth-quarter letter would tell you the funnel is still filling?
Is Reddit cheap, or does it only look cheap?
| Position | Seats | The figure it rests on |
|---|---|---|
| Cheap | D | 14.8x trailing reported free cash flow |
| Fair once costs count | A, C | $385.8M of stock pay; a 1.5% tax rate |
What settles it. The 2026 10-K, due by March 1, 2027: the tax rate, and stock pay against Investor A's line of 15% of revenue.
From the chair, to Investor C. In your base case, year-three free cash flow is near $2.0B, a third above Investor D's, from a 38.8% margin you call flattered by stock compensation. Why does the margin you distrust in the multiple carry your scenario?
| Investor | Verdict | What breaks its thesis |
|---|---|---|
| A | Watch 6 | Logged-in users down year over year; revenue below the $1,000M floor; 2026 stock pay above 15% of revenue; licensing revenue down in any 2027 quarter. Any one moves it to Avoid |
| B | No response | Did not answer this session |
| C | Watch 6 | Logged-in users below 53.5M; revenue below the floor; licensing revenue down in the first-quarter 2027 10-Q. Daily users falling well past the guided path move it to Sell |
| D | Buy 7 | Logged-in users below 53.0M; advertising growth below 20% in the first-quarter 2027 10-Q; 2026 operating margin below 18%. A sustained fall in logged-in users moves it to Sell |
Valuation
What the price requires. These figures are ours, and the seats never saw them. Our discount rate is the 10-year yield of 4.98% plus 4.5 points, 9.48%, with 3% growth after year ten. On a $17.71B market value, trailing reported free cash flow must grow 2.0% a year, putting year three at $1.27B. At 8.48% that is -0.1% and $1.19B; at 10.48%, 4.0% and $1.35B. Reported free cash flow excludes stock pay, which holders bear. On $0.81B after stock pay, the need is 7.2% and $1.00B, in a band of $0.94B to $1.06B. A full tax charge would raise it again. We chart the reported basis, because two of three seats built their scenarios on it.
Reported free cash flow is 6.8% of market value and 8.2% of enterprise value, net cash counted, against the 10-year's 4.98%. After stock pay, 4.6% and 5.5%.
Hatched: what our hurdle needs at three discount rates. Solid: each seat's weighted scenarios. Investor A counts stock pay and a 21% tax, so it belongs against the adjusted band. B cast no vote.
| Investor | Bear | Base | Bull | Weighted, year 3 | Against the price |
|---|---|---|---|---|---|
| A | 30%, about $0.5B | 50%, about $1.1B | 20%, about $1.8B | $1.06B | inside the band, adjusted |
| C | 30%, about $0.9B | 45%, about $2.0B | 25%, $2.7B to $3.0B | $1.88B | above the band |
| D | 25%, $0.65B | 55%, $1.50B | 20%, $2.40B | $1.47B | above the band |
Fiscal 2029 is year three. A range appears as the seat gave it, and the weighted column takes its midpoint.
Two seats sit above the reported band, and A sits at the top of the adjusted one. Investor C expects the most cash and still voted Watch:
At a fair multiple with fat tails on both sides, the distribution is close to symmetric, and the expected value of guessing is not obviously positive, which is the whole reason for Watch.
Investor C, Watch
Investor D's Buy sits against its own entry line: enterprise value below 12.0x trailing free cash flow. On our figures it is 12.2x.
| Investor | Entry discipline |
|---|---|
| A | Owner earnings yield on enterprise value of 7% or more, measured after the search decline is absorbed |
| C | Up to about 15x trailing reported free cash flow after results inside guidance; about 10x without them |
| D | Enterprise value below 12.0x trailing free cash flow, or a reported yield above 8.5% |
Comparables
| Company | Market cap | Growth, last FY | Operating margin | P/E | FCF yield |
|---|---|---|---|---|---|
| Reddit (RDDT) | $17.7B | 69.4% | 20.1% | 17.9x | 6.8% |
| Pinterest (PINS) | $11.1B | 15.8% | 7.6% | 62.3x | 11.4% |
| Snap (SNAP) | $9.2B | 10.6% | -9.0% | n/a | 7.7% |
| Meta (META) | $1.57T | 22.2% | 41.4% | 20.9x | 1.8% |
Each company's filings and Yahoo Finance prices, computed alike. Snap is loss-making, so has no P/E.
Reddit grew more than four times as fast as Pinterest or Snap and trades below Meta's earnings multiple. Part of that is tax, since Reddit's earnings are almost untaxed. The rest is the referral risk. For another attention business priced on its audience, see our Netflix review.
Key risks
Ranked by the desk, most severe first. Raised by names each source that flagged it.
| Risk | Why it matters | Raised by | First signal |
|---|---|---|---|
| High. Referral losses reach logged-in users | They carry over 80% of monetization | Company letter, Analyst, Investors A, C, D | Logged-in users below 53.5M |
| High. People stop posting | The archive ages and the licence loses value | Analyst, Investor A; forum (unverified) | Licensing revenue down year over year |
| High. Google is traffic source, licensee and rival | One party can reprice traffic and data together | Analyst, Investor C; talks reported by Briefs Media, not confirmed in a filing | A licence renewal or lapse on Form 8-K |
| Medium. Stock pay and a near-zero tax charge | Owner cash flow is a third below reported | Desk arithmetic, Analyst, Investors A, C | Stock pay above 15% of revenue in the 10-K |
| Medium. The ad cycle | Advertising is 95% of revenue | Analyst, Investor C | Ad growth below 20% |
| Medium. Control | Outside holders cannot force change | 10-K Item 1A, Analyst, Investor A | Form 144 sale notices |
The filings settle one question: with no debt, financing is not a risk.
What to watch
| Date | Event | What it decides |
|---|---|---|
| Not announced | Fourth-quarter results, Form 8-K; last year's was filed February 5, 2026 (EDGAR) | Logged-in users against 53.5M; revenue against $1,000M |
| By March 1, 2027 | 2026 Form 10-K (SEC deadline) | Tax rate; stock pay against 15%; operating margin against 18% |
| By May 10, 2027 | First-quarter 10-Q (SEC deadline) | Ad growth against 20%; licensing revenue year over year |
| Ongoing | Form 8-K filings | A data licence signed, renewed or lost |
Owning it onchain
Two tokenized versions exist, xStocks' RDDTx and Backpack's RDDT.US. We have not judged whether either is investable: we checked on Saturday, October 31, 2026, with the US market closed for the weekend.
Data
| Item | Figure | Source |
|---|---|---|
| Price | $91.73, October 30, 2026 close | Yahoo Finance |
| Shares outstanding | 193,093,731, three classes | 10-Q cover, filed October 30, 2026 |
| Market value | $17.71B | Price times cover shares |
| Enterprise value | $14.68B | Market value less $3.03B of cash and short-term investments |
| 10-year Treasury yield | 4.98%, October 29, 2026 | FRED |
Method. Four investors with different philosophies, each on a different AI model, read one identical fact sheet and voted independently. A name needs three Buys to enter a basket. How the committee works.
Correction. This article first said neither token was investable, on a missing quote for RDDTx and a 3.10% round-trip cost for Backpack's RDDT.US. Both were checked on a Saturday, with the US market closed. We now judge that cost only during regular US market hours, and we have removed the figure and the verdicts.
As of October 31, 2026, BagsCapital manages no outside capital and holds no position in Reddit. Nothing here is investment advice.