SHORT REPORT PUBLISHED: 2026-03-12 08:00 AM ET | Muddy Waters Research | "AppLovin: A House of Mirrors"

ABI tear-down shipped 11:47 AM ET | time-to-publish 3h 47m

NASDAQ: APP Communication Services | Interactive Media & Services | Large-Cap Short Report Active

Short Report Rapid Response – APP / Muddy Waters

AppLovin Corporation | Palo Alto, CA | 1,750 employees | founded 2012 | applovin.com

$284.16

−14.82% | −$49.42

spot | May 26, 2026 11:47 AM ET

Muddy Waters published a 87-page short report on AppLovin alleging the AXON 2.0 ad-attribution model overstates incremental conversions by 38-52%; that AppDiscovery click farms in Vietnam and Indonesia inflate installs; and that ~$1.1B of FY25 revenue is double-counted via a circular flow with related parties owned by Tim Stone affiliates. ABI shipped this 14-allegation triage at 11:47 AM ET – 3h 47m after publication. 14 allegations scored: 2 Substantiated, 5 Partially, 5 Refuted, 2 Inconclusive.

Stock move @ pub

−18.2%

−$60.78 to $273.40 low

Short interest pre

4.7%

of float | pre-pub

Short interest post

9.1%

+440bp | post-pub

ABI verdict score

38/100

moderate cred.

Allegations

across 6 themes

Sub / Part / Ref / Inc

2 / 5 / 5 / 2

ABI triage

APP | Intraday (1-min) | short report dropped 08:00

$333.58 → $273.40 (low) → $284.16 | spike vol >42M shares

LIVE

Borrow rate: 4.8% → 38.5% bps | IV30 jumped from 42% to 78% | put/call ratio 0.38 → 1.91

ABI Bottom Line AI synthesized | analyst reviewed

Verdict from the Rapid Response desk | published 11:47 AM ET | subject to update on management response

Tear-Down v1.2 Next update: 4 PM ET

VERDICT

HOLD | trim 25-30% of position

conviction: medium-high | evidence quality: 5.2/10

Muddy Waters' report contains two genuinely substantiated allegations – (1) attribution-model overstatement is real and material (we estimate 18-25% inflation vs MW's 38-52%, supporting our $0.42 EPS haircut); and (2) the Adjoe-related-party arrangement disclosed in the FY25 10-K Note 18 is structurally what MW alleges, though the dollar amount is closer to $310M (1.9% of FY25 rev) than MW's $1.1B claim. The five refuted allegations – click farms, axon model accuracy, app-store policy bomb, gaming network overstatement, Tim Stone control – are either factually wrong, double-counted, or fundamentally misread public disclosures. We model a base-case 20-30% drawdown (already 18% in), but believe full Muddy Waters thesis pricing (−45 to −60%) is unwarranted. Buy the dip below $245 if Q1 print confirms FY26 guide; cut to zero if Adjoe write-down exceeds $400M or DOJ subpoena surfaces in next 30 days.

If allegations 30% true → fair value $228 (−20%) If 50% true → $185 (−35%) If 80% true → $124 (−56%) If refuted → $358 (+26%)

Allegations Triage Source-verified

Every claim from the Muddy Waters report scored on verdict, evidence strength, materiality if true, and source-checked against primary docs.

#	Allegation (Muddy Waters)	ABI Verdict	Evidence (1-5)	Materiality if true	Source check
1	AXON 2.0 attribution overcounting. MW alleges the model overstates incremental conversions by 38-52% by attributing organic installs to paid sources.	Substantiated	★★★★★	−$0.42 FY26 EPS (~$220M rev haircut)	Singular, Branch, Appsflyer cross-checks confirm 18-25% gap (vs MW's 38-52%). Material but overstated.
2	Adjoe/Tim Stone related-party revenue inflation. ~$1.1B of FY25 rev routed through Adjoe-affiliated entities controlled by Stone family trust.	Substantiated	★★★★★	$310M rev at risk \| ~$0.55 EPS	FY25 10-K Note 18 (Related Parties) confirms Adjoe relationship; ABI estimates $310M (not $1.1B). Material disclosure gap.
3	Click fraud on AppDiscovery network. Vietnam/Indonesia click farms inflate install counts 12-18%.	Refuted	★★★★★	n/a – claim fails	Pixalate, DoubleVerify, IAS three-way audit shows IVT rate 2.1% (below industry 3.4%). MW screenshots are from 2022 era.
4	Apple App Store policy risk. MW alleges undisclosed SKAdNetwork policy violation; 90-day suspension imminent.	Inconclusive	★★★★★	−15-30% rev if suspended	Apple developer policy v17.4 ambiguous on alleged practice. No DPLA enforcement actions in 24m. Tail risk but unverifiable.
5	Gaming Network revenue overstatement. ~$420M overstated via promotional crediting from sister-co Lion Studios.	Partially	★★★★★	$80-120M rev haircut	Lion Studios is owned, not sister. Inter-co eliminations disclosed Note 4. ~$95M of "promotional credit" arrangement undisclosed.
6	AXON model accuracy claim. MW alleges 71% predicted ROAS vs actual 38% delivered to top-10 advertisers.	Refuted	★★★★★	n/a – claim fails	3 of "top-10" cited by MW are not actually in top-50 spenders (Sensor Tower). MW used 2023 internal benchmark, not customer data.
7	Data broker exposure. $2.3B class-action liability from CCPA/COPPA violations across data partnerships.	Partially	★★★★★	$200-500M potential settlement	Two pending class actions (N.D. Cal. 23-cv-04412, 24-cv-00871) confirm exposure. $2.3B MW figure unsupported; realistic exposure $200-500M.
8	Tim Stone affiliated entities own 6.2% of float undisclosed. Hidden control / 13D filing failure.	Refuted	★★★★★	n/a – claim fails	13F filings cross-check shows 1.8% via family trust (disclosed in 2024 proxy). MW conflated separate "Stone" entities (different individuals).
9	SDK telemetry over-collection. SDK collects 47 device fingerprints exceeding stated policy; FTC violation risk.	Partially	★★★★★	$50-150M potential fine	AppLovin SDK v8.2 telemetry list (38 fields per public docs) vs MW packet capture (47 fields) – gap real. FTC unlikely to act sub-$200M.
10	EBITDA add-back abuse. $185M of "non-recurring" stock-comp / restructuring is structurally recurring.	Partially	★★★★★	Adj EBITDA −12% if reclassified	SBC has been 8-11% of rev each year FY22-25 – not non-recurring. ~$140M (not $185M) genuinely structural. Adj EBITDA overstated.
11	Customer concentration hidden. Top-5 advertisers = 38% of rev, not the 22% disclosed.	Refuted	★★★★★	n/a – claim fails	10-K Item 1 discloses top-10 at 31% concentration. MW collapsed unrelated entities into one "customer" to inflate figure.
12	Wurl/CTV acquisition impaired. $430M paid; FV today <$80M; impairment overdue.	Inconclusive	★★★★★	−$0.18 EPS one-time charge	Wurl segment revenue grew 14% YoY (disclosed). FV depends on private market comps – currently impossible to verify without management goodwill test data.
13	Insider selling pattern signals concern. CFO Herald sold $42M shares in 60 days before report.	Partially	★★★★★	Sentiment only	Form 4s confirm $38.4M (not $42M), all under 10b5-1 plan filed Sept 2025 (predates report by 6 months). Pattern real, but pre-planned.
14	Auditor change red flag. KPMG → Deloitte change in Q4-25 signals undisclosed material weakness.	Refuted	★★★★★	n/a – claim fails	Auditor change disclosed 8-K filed 2025-10-04 with standard "no disagreements" letter. Routine rotation per audit committee charter.

Sources cross-referenced: APP 10-K FY25 (filed 2026-02-19), Q1-Q4 10-Q FY25, DEF14A proxy 2025, 8-K filings 2024-2026, Form 4 insider filings 6m window, Singular/Branch/Appsflyer attribution audit, Pixalate/DoubleVerify/IAS IVT audit, FTC CID database, court PACER (N.D. Cal.), Sensor Tower top-spender data

What's True | What's Not

Distillation of the 14 allegations into the two pieces that matter for your book sizing decision in the next 24 hours

● What's actually true (the meat)

Attribution overcounting is real, and material. Independent triangulation across Singular, Branch, and Appsflyer suggests AXON 2.0 inflates incremental conversions by 18-25% – meaningful but ~half of Muddy Waters' 38-52% claim. EPS impact ~$0.42 if accounting catches up.
Adjoe related-party flow is a real disclosure gap. FY25 10-K Note 18 references Adjoe but understates economic dependency. Real number is ~$310M (1.9% of revenue), not $1.1B. Audit committee should re-examine, and a 10-K/A is plausible.
SBC add-back abuse for adj EBITDA. $140M+ of "non-recurring" stock-comp is structurally recurring at this scale – should be in opex. Adj EBITDA overstated 9-12%.
Two real lawsuit tail risks (data broker class actions, $200-500M realistic settlement range) and SDK telemetry policy mismatch (likely sub-$150M fine if FTC pursues).

● What MW got wrong (or recycled)

Click fraud claim is recycled from 2022. MW's "smoking gun" screenshots predate AppLovin's 2024 SDK rewrite. Current IVT is 2.1% – below industry average. Refuted.
"AXON 71% vs 38% delivered" benchmark is fabricated. MW used a leaked 2023 internal slide as if it were customer-realized ROAS. Three of MW's "top-10 advertisers" aren't actually top-50 spenders per Sensor Tower.
Tim Stone control narrative collapses. MW conflated multiple unrelated individuals named "Stone" + entities into one alleged hidden controller. Real position via Stone family trust is 1.8%, properly disclosed in 2024 proxy.
Customer concentration claim falsifies trivially. 10-K Item 1 already discloses top-10 at 31%; MW collapsed unrelated entities to manufacture a 38% figure.
Auditor change red flag is a non-issue. 8-K with no-disagreements letter filed Oct 2025. Routine rotation. MW chose not to disclose this context.

Stock Impact Model | Heatmap

Fair value as a function of what % of MW allegations the market eventually believes & the EPS multiple it accepts. Spot = $284.16.

Fair value sensitivity – % of allegations true × fwd EPS multiple

Base FY26E EPS $13.40 (consensus pre-event)

% TRUE \ FWD P/E

18x

22x

26x

30x

35x

0% true (refuted)

$241

$295

$348

$402

$469

15% true (light haircut)

$205

$251

$296

$342

$399

30% true (ABI base case)

$169

$206

$243

$281

$328

50% true (mw partial)

$132

$162

$191

$220

$257

80% true (mw thesis)

$78

$96

$113

$131

$152

How to use: Find where the market is pricing in (currently ~32% true at ~21x = $191 implied vs $284 spot, suggesting market still discounting heavily). Our base case (30% true @ 30x) lands at $281 – essentially where spot is now – meaning current price already reflects ABI's central scenario. Asymmetric upside if Q1 print clears the air (refutes click fraud, axon claims): fair value re-rates toward $342-399. Asymmetric downside only materializes if Adjoe write-down or DOJ subpoena triggers full 50%-true repricing → $191 (−33% from spot).

Sell-Side Reaction Tracker Live | refreshed every 15min

Every covered analyst with a take on the Muddy Waters report. Click ticker to open original note.

Firm	Analyst	Pre-event rating	Pre-event PT	Reaction (time)	Post rating	Post PT	Stance on MW
Morgan Stanley	Brian Nowak	Overweight	$420	09:14 ET	Overweight	$385	"Sees real attribution issue, refutes Adjoe scale"
JPMorgan	Cory Carpenter	Overweight	$405	09:32 ET	Overweight	$370	"Overdone; constructive on Q1 print"
Goldman Sachs	Eric Sheridan	Favorable	$430	10:18 ET	Favorable	$395	"Most attribution claims unsubstantiated; trim 100bps on Adjoe risk"
Bank of America	Omar Dessouky	Neutral	$310	11:02 ET	Neutral	$275	"Was already concerned on related-party flows; vindicated"
Wells Fargo	Steven Cahall	Overweight	$395	11:31 ET	Equal Weight	$285	DOWNGRADE – "credibility gap too wide pending mgmt response"
Citi	Jason Bazinet	Favorable	$385	12:14 ET	Favorable	$345	"MW report overstates; uses 2022 data on attribution"
Jefferies	Andrew Uerkwitz	Favorable	$410	13:08 ET	Favorable	$365	"Buy the dip; ad-tech fundamentals intact"
Bernstein	Mark Shmulik	Market-perform	$320	14:42 ET	Market-perform	$290	"5/14 substantiated is worse than typical short report"
Stifel	Scott Devitt	Favorable	$400	pending	–	–	"Note pending – preview call at 4 PM ET"
Macquarie	Tim Nollen	Outperform	$390	pending	–	–	"On road; expected end-of-day"

Net Δ across 8 reacting analysts: Mean PT cut $397 → $339 (−14.6%) | 1 downgrade (Wells) | 0 upgrades | 2 still pending | ABI sees 7/8 reactions as directionally aligned with our 30% true / $281 base case.

Borrow & Options Flow | Pre vs Post-Publication

Pre-pub close (March 11 4 PM ET) vs intraday peak post-pub (March 12). Source: S3 Partners, OPRA, Cboe.

Short interest & borrow

S3 Partners + ABI broker survey

Short interest (% float)4.7% → 9.1%

Short interest ($M)$2.84B → $4.91B

Days to cover2.1 → 5.8

Borrow rate (bps)48 → 3,850

Hard-to-borrow flagN → Y (HTB)

Net new shorts (shares)– → +6.4M

CTB squeeze pressure score2/10 → 7/10

Options & vol surface

OPRA + Cboe live

IV3042% → 78%

IV6040% → 68%

Put/Call (volume)0.38 → 1.91

25-delta put skew+2.8 vol → +11.4 vol

$250 put open interest2,418 → 38,491

$200 put open interest812 → 22,604

Implied 30d move±$24 → ±$58

Implied vol surface | 30d

IV by strike, pre (gray) vs post (red)

Historical Precedent | Recent Short-Report Targets

How comparable short reports played out on Day 30 / 90 / 365 from publication. Source: Activist Insight, ABI historical database (2020-2026).

Target	Short seller	Date	Day-0 reaction	30d %	90d %	365d %	Outcome
EQIX Equinix	Hindenburg	Mar-24-2024	−10.4%	+8.1%	+22.4%	+34.8%	Allegations refuted by audit committee; stock recovered
CVNA Carvana	Hindenburg	Jan-04-2024	−4.1%	+38.2%	+71.5%	+202.4%	Liquidity claims overstated; structural turnaround
NIO NIO Inc.	Grizzly Research	Jun-28-2022	−2.2%	−18.4%	−36.1%	−62.4%	Battery-swap accounting issues genuine; under SEC review
DJT Trump Media	Kerrisdale	Apr-08-2024	−4.8%	−42.1%	−58.4%	−72.8%	Most claims directionally correct; valuation collapsed
FFIE Faraday Future	J Capital	Oct-22-2021	−9.1%	−21.4%	−54.2%	−94.8%	Going concern materialized; substantiated
FUTU Futu Holdings	Citron	May-14-2024	−6.2%	+12.1%	+24.8%	+48.4%	Chinese broker concerns dismissed by regulator
MBLY Mobileye	Spruce Point	Aug-12-2023	−3.4%	+4.2%	+14.1%	−18.7%	Mixed; design-win softness independent of report
WBA Walgreens	Muddy Waters	Sep-18-2024	−6.8%	−12.4%	−28.1%	−42.4%	MW thesis on store-write-down materialized in Q4
SDGR Schrödinger	Bonitas	Feb-08-2025	−4.2%	+8.4%	+18.1%	–	AI-drug discovery claims survived scrutiny
APP AppLovin	Muddy Waters	Mar-12-2026	−18.2%	TBD	TBD	TBD	SUBJECT \| ABI base case: ~$281 (flat from spot)

Base-rate read: Across 9 prior comparable short reports – 5 went up, 4 went down at the 365-day horizon. The key fork is whether allegations materialize in subsequent quarterly prints. Of MW's prior 6 named-target reports since 2022, average 365-day return was −24% (vs ABI universe baseline +9.2%). MW has a real hit rate on operational claims; this is the single most important reason to keep some short / put exposure even if we are skeptical of the headline figures.