Reddit, Inc. [RDDT] — Valuation

Two outputs, two horizons. Reporting only one is the defect this file exists to prevent. As of 2026-07-29 · spot $178.345 (close 2026-07-28) · no position verdict is issued.

0. Verified inputs — every figure traced to a filing

1. Implied-path test — the Valuation Criteria (BINDING)

assets/reverse_dcf.py, solved for the revenue CAGR today's price requires. Solved for; not asserted.

1.1 Parameters — what was solved for, what was held fixed

1.2 The result

This corrects the screen. The screen reported required 37.0% and a margin of +11.9pp. The difference is entirely the two input corrections, both of which the screen got wrong in the same direction:

1.3 Anchoring the exit multiple — growth-matched, and its compression stated

Basis: GROWTH_MATCHED. Comparator set = every name in the 4,018-name scan_all_v2 universe with a positive ev_ebit in (0, 200) and an EV of at least $100m whose demonstrated growth falls within ±50% of Reddit's 48.9% — i.e. 24.5% to 73.4%. n = 135. Reddit's growth is inside the comparator band by construction, so the growth/multiple relationship is estimated on support rather than extrapolated. This is the defect that drove every high-growth rejection in the record (a 1.0–7.5% growth anchor set used to value 15–39% growers) and it does not apply here.

Median EV/EBIT of the matched set: 24.2x.

Implied compression, stated as a number:

Reddit trades at 53.7x TTM EV/EBIT today. The test exits at 24.2x. Compression of 29.5 turns, or −55.0%.

That is a severe compression assumption and it makes the test strict, not lenient. The required 29.0% CAGR is what the price needs even after the multiple more than halves. If Reddit still traded at 53.7x in 2031, the required CAGR would be 10.0% and the margin +38.9pp. The framework's rule — that a base exit multiple may not sit below every stated anchor without a separately argued reason — is satisfied in the other direction: 24.2x is the anchor, and no distant-year haircut has been stacked on top of it.

1.4 Mandated sensitivity — over the exit multiple, never over scenario probabilities

Terminal margin held at 25.1%; every other parameter as §1.1.

The flip point is at an exit multiple of ~12.0x EV/EBIT — half the growth-matched anchor, and less than a quarter of today's trading multiple. At any exit multiple above ~12x, the price requires less than Reddit has demonstrated. This is the parameter that determines the answer, and it is where the range was run.

Secondary sensitivity, over the terminal margin (exit multiple fixed at 24.2x):

1.5 Terminal value as a share of EV

Under the base parameters the entire modelled EV is terminal by construction (reverse_dcf.py discounts a single terminal EV and states so — interim cash flows are deliberately omitted because they cannot change the answer for a name with this growth profile). Terminal value is therefore >60% of EV and the reverse DCF is mandatory as the primary long-horizon output. A forward DCF would be supporting evidence at best and is not produced.

1.6 The honest reading — what 29%/yr actually asks of the business

The framework's test is passed. The judgement that matters sits underneath it, so it is stated explicitly.

29.0%/yr for five years takes revenue from $2,473.6m to $8,836m by 2031. Reddit's revenue identity is DAUq × ARPU × 4. What that path requires:

Current DAUq growth is +17% global and +7% in the United States. Current ARPU growth is +44% global. So the required path is roughly: DAUq growth decays from 17% to ~10%, and ARPU growth decays from 44% to ~16%, and both hold there for five years. That is a substantial deceleration from the present run-rate in both terms — which is precisely why the test passes.

The constraint the test does not see is that US logged-in DAUq growth is +1% (research §3). If US DAUq goes flat, the DAUq leg is carried entirely by rest-of-world, where ARPU is $2.02 versus $9.63 — so mix alone drags blended ARPU down while the model needs it up. That is the real risk in this name, and it is captured in the Downside Criteria, not in the implied path.

Valuation Criteria (BINDING): PASS. Margin +19.9pp. This is the number the strategy ranks on.

2. Twelve-month target — and why the multiple anchor is UNIDENTIFIED

2.1 The name's own multiple history

Built per method: daily EV/Sales from IPO, using as-known TTM revenue stepped in at each filing date (so the series is not forward-looking), with the verified share count and net cash held fixed — which makes the series a measure of multiple movement, which is what a percentile is for. Basis stated, not silent.

2.2 The anchor is declared UNIDENTIFIED — with reasons, and no peer median substituted

Per the brief and per valuation.md, a too-short window or one spanning a regime change must be declared unidentified rather than back-filled. Three independent reasons apply:

1. The window is 2.2 years and one market cycle. 556 sessions, all post-IPO-2024. There is no pre-IPO comparison and no full multi-year cycle inside it.
2. The range is 5.0x wide (7.68x → 38.20x) and the interquartile range alone spans 11.89x–22.84x — a ±40% band around the median. A "median multiple" drawn from that dispersion is not an estimate.
3. There is a documented structural break mid-window. US logged-in DAUq growth ran +32% in 2024Q2 and +1% in 2026Q1; Reddit crossed from loss-making (FY2024 operating margin −43.1%) to +25.1% TTM inside the same window; and management retires the logged-in/logged-out disclosure in Q3 2026. EV/Sales at the start of this series and EV/Sales at the end of it are pricing different businesses in different traffic regimes. The pre-2025-07-28 median is 13.99x; the trailing-12-month median is 18.18x. The series is not stationary.

No peer median is substituted. META at 6.97x EV/Sales, GOOGL at 8.83x and SNAP at 1.47x are named in the Peer Spread Criteria below as context, and none of them is used to set a target. That substitution is the exact defect the rule prohibits.

2.3 The band, reported anyway — because the band is itself the finding

NTM revenue base: $3,444m.

Construction, and its limitation, stated plainly. No Street consensus could be obtained (Alpha Vantage EARNINGS_ESTIMATES shares the same exhausted 25/day cap as the transcript endpoint), so valuation.md step 1 — "start from near-term consensus" — could not be satisfied. The base is instead built forward from the company's own guided Q2 2026 revenue of $715–725m (mid $720m, +44.1% YoY) with YoY growth decaying ~5pp per quarter thereafter: Q3'26E $801m (+37%), Q4'26E $943m (+30%), Q1'27E $829m (+25%), Q2'27E $871m (+21%). Sum $3,444m, i.e. +39% on TTM. This is a house build, not consensus, and it is the weakest input in this memo.

2.4 The 12-month output

12-month target: INDETERMINATE as a point. Band $216 – $402 (+21% to +125%), from Reddit's own interquartile multiple range applied to a house NTM revenue of $3,444m.

The informative fact is not the midpoint. It is that the entire interquartile band sits above spot. Even at the 25th percentile of its own two-year multiple history — a multiple Reddit has traded below only a quarter of the time since IPO — the implied price is +21%. Reddit is at the 41st percentile of its own multiple range while its revenue is 39% higher than a year ago.

Two things this does not say. It does not say the stock goes up: the same band's lower bound (the minimum multiple ever printed, 7.68x) is −19%. And it does not substitute a peer median for the missing point estimate.

Expected direction check (item B16). The framework records 16 of 16 prior house targets sitting below spot, a median 46.1% below Street — a house view about the market, not a valuation. This target band sits above spot at every percentile from the 25th up. It is not another instance of B16.

3. The MEASURED Criteria — scored, blocking nothing

3.1 Momentum Criteria (MEASURED — entry timing, never selection)

Reddit sits at the median of its cross-section on 12-1 momentum and near the bottom of its own 52-week range with an RSI of 38. This governs when to enter a position the thesis already justifies — never whether to own it. Nothing in this section blocks anything, and no language here vetoes the name.

The specific timing read: a 48th-percentile momentum score with an RSI of 38 and a −34% drawdown from the high is a name whose price is not confirming, into an earnings print that is imminent. That is a timing input, and it points to waiting for Q2 rather than to any conclusion about ownership.

3.2 Catalyst Criteria (MEASURED)

Fully dated in RDDT_Catalyst_Calendar.md. Headline: Q2 2026 results are due within days (prior-year analogue 2025-07-31), and the logged-in/logged-out DAUq disclosure is withdrawn with the quarter ending 2026-09-30. Time works for a long, so a dated event is not required to own a compounder.

3.3 Consensus Criteria (MEASURED)

INDETERMINATE. No Street estimate could be obtained — the Alpha Vantage 25/day cap was exhausted by parallel runs before this memo began. A quota gap leaves this blank and blocks nothing (criteria.md). No consensus figure is quoted anywhere in this memo, and none is invented.

3.4 Peer Spread Criteria (MEASURED)

Named peers in the same end-market (digital advertising), from scan_all_v2:

Spread: RDDT trades at 1.9x META on sales and 3.2x META on EBIT, for 2.5x META's growth and 0.61x its operating margin. The margin gap is the whole argument — Reddit at Meta's 41.4% operating margin on today's revenue would be a 32.7x EV/EBIT, not 53.7x. The spread is a bet on Reddit closing a 16pp operating-margin gap, which is exactly what the implied-path test's 25.1% terminal margin does not assume. That asymmetry is worth holding: the reverse DCF is more conservative than the peer spread implies.

3.5 Short Mechanism Criteria (MEASURED — scored on every name, acted on by nothing on this fork)

Requires decelerating growth and exhausted margin runway. - Decelerating growth: YES. +69% (Q1'26 actual) → +44% (Q2'26 guided). DAUq +51% → +17% over eight quarters; US logged-in +32% → +1%. - Exhausted margin runway: NO. Operating margin is 25.1% against a 91.5% gross margin and a Meta comparator at 41.4%. Incremental leverage is visibly not spent — Q1'26 alone ran 27.6%.

One of two conditions met. Short Mechanism Criteria: FAIL (i.e. no short mechanism). Recorded for the relative-value fork; acted on by nothing here.

3.6 Sub-sector Criteria (MEASURED)

Not a healthcare name; the reference taxonomy (LC MedTech, SMID Growth, Genomics/Dx, Pharma, Labs, LC Tools, HCIT, Services, SMID Other) has no applicable tag. Tagged for concentration purposes as Internet / Digital Advertising — Large Cap, SIC 7374. Correlation cluster: META, GOOGL, PINS, SNAP, APP — and, through the AI-referral channel, the entire consumer-internet complex.

3.7 Liquidity Criteria (BINDING)

Full chain evidence in RDDT_Trade_Construction.md. Summary: PASS. The January 2027 chain carries 40,938 contracts of call open interest and 28,823 at the $220 strike alone, with ATM quoted spreads at ~3.3% of mid. The HCA failure — a maximum open interest of 18 contracts across an entire chain, making the default defined-risk spread uninvestable at any size — has no analogue here.

4. Criteria summary

No position verdict is issued. The book decides.