Newsletter · 2026-10-07 · Ashutosh Agarwal

Carson Block's AppLovin Short Finally Works but Not for the Reasons He Gave - Track Record - May 2025 to October 2026

A historical scorecard of Muddy Waters founder Carson Block, who defended his AppLovin short on a May 9, 2025 podcast after a March 27, 2025 report, measured through the October 6, 2026 close. AppLovin fell 17.9 percent against the S&P 500's 38.0 percent gain, but first ran 116 percent against the short and the SEC closed its inquiry with no action. Verdict: right on the stock, wrong so far on the why.

Track Record

Carson Block, May 2025 to October 2026

This is a historical review of Carson Block's March 27, 2025 Muddy Waters short report on AppLovin and the call he defended on a May 9, 2025 podcast, with performance measured through the October 6, 2026 close. It was published on October 7, 2026; the prices and grades below are as of the October 6, 2026 cutoff.

Carson Block's AppLovin short is finally working. The stock is down 17.9% since he defended the call on a podcast in May 2025, while the S&P 500 is up 38.0%. But it got there by first going up 116% against him, and for reasons that have little to do with the fraud-style case he made.

The Call

On May 9, 2025, Carson Block, founder of the short-selling firm Muddy Waters, sat down with Josh Brown and Michael Batnick on The Compound and Friends, "The Soul of a Short-Seller" (May 9, 2025). (A short-seller bets a stock will fall. Block is known for publishing detailed research that accuses companies of misconduct.)

Six weeks earlier, on March 27, 2025, Muddy Waters had published a short report on AppLovin (APP), the mobile ad-tech company that was then one of the best-performing stocks in the market. The hosts asked whether he was still short. His answer:

"We are. So we're taking on the chin a little bit today."

The stock was up about 12% that day after strong earnings. Block was clear that he was not saying the numbers were fake:

"We reported on them a couple months ago, and basically the numbers are real. We're not accusing them of fake numbers."

The thesis, in plain English:

• AppLovin pitched its AI ad engine (called Axon) as nearly as good as Google's or Meta's at targeting ads. Block said the real secret was tracking people across apps and websites without permission.
• In his words, Muddy Waters spent its time "actually looking at the code and how AppLovin has created these in violation of all the platforms' terms of services, these persistent identifier graphs that track users from site to site and then hit you with retargeting ads." A "persistent identifier graph" is basically a hidden profile that follows you from app to app.
• He accused management of lying in its denial: "We put out an update yesterday. It was a video that showed, hey, they're still doing this, and the CEO and CTO lied."
• He described two ways AppLovin could lose. Either the platforms "could shut them down," or, if they didn't, copycats would move in: "this doesn't take a genius to do what they're doing... basically their margins are going to be, you know, cut to, you know, razor thin."

What he predicted: AppLovin stock would fall as the cheating either got stopped or got copied. No price target, no timeframe.

Our entry point: the episode was recorded on May 8, 2025, when APP closed at $339.51. For reference, the stock closed at $327.62 the day before the original report came out.

What Happened

The short worked on day one, then stopped working for nine months.

• The report knocked APP down 20% in a single session ($327.62 to $261.70). Tariff panic pushed it to a $219.37 close on April 4, 2025.
• Then it took off. AppLovin was added to the S&P 500, and in Block's words the AI trade became "euphoria." The stock closed September 2025 at $718.54, hit an intraday high of $745.61 on September 29, 2025, and closed at its record of $733.60 on December 22, 2025.
• Anyone who shorted at the podcast price was down more than double at the peak. A short position's loss has no ceiling, so that is the kind of move that forces most funds to cover.

Then the stock collapsed through 2026. Month-end closes: $473 in January, $398 in March, a spring bounce to $613 in May, then $396 in July, $312 in August and $290 in September. On October 2, 2026 it closed at $268.22, its lowest since the podcast.

Why it fell (this matters for the grade):

• Competition fears, not a crackdown. In late January 2026, Google's Project Genie, an AI tool that builds playable game worlds, wiped about $47B off U.S. gaming stocks in one session, per Jefferies. Morgan Stanley's Matthew Cost said on Feb 4, 2026 that APP had sold off "on fresh worries about disruption to the industry." JPMorgan's Cory Carpenter said in February that he expected "Meta will remain an overhang."
• The first miss. In Q2 2026, AppLovin came in under the midpoint of its own revenue and EBITDA guidance for the first time since going public: $1.92B revenue and $1.61B adjusted EBITDA, which Oppenheimer called "the first shortfall in twelve quarters."
• A wave of downgrades. Piper Sandler (to Neutral, target cut to $385 from $665, "more questions than answers"), Wells Fargo (to Equal Weight, $357, saying AppLovin's "market share in mobile games is plateauing"), and BofA (to Neutral, $400, saying the risks to the 30% long-term growth target "have increased"). Tiger Global sold its whole stake in Q2 2026.
• The business is still enormously profitable. Wedbush noted gaming is "still growing north of 20%." Benchmark models adjusted EBITDA margins at 80-84%. That is the opposite of "razor thin."

The legal side cut against Block. The SEC opened an inquiry into AppLovin's data practices in October 2025. On August 5, 2026, the company said the SEC had "concluded its inquiry with no recommended action." On the other side of the ledger, a lawsuit on behalf of "The People" of California was reported filed in San Diego Superior Court on October 5, 2026. It alleges illegal tracking and fingerprinting of users under state privacy law, the exact behavior Block described. It's too early to know what that suit is worth.

The Verdict

Grade: right on the stock, wrong (so far) on the why.

What he got right:

• The scoreboard. Measured from the podcast, the short is up 17.9% in absolute terms and has beaten the market by about 56 points. Measured from the record high, the stock is down 62%. Few shorts on a $100B-plus company pay off this much.
• The "no moat" instinct. His case that AppLovin's edge was easy to copy is close to what the market now worries about: Meta, Google, Unity and slowing share gains in games. Morgan Stanley noted in September that investors fear "success for AppLovin and Unity may be mutually exclusive."
• He was honest about the numbers. He never claimed the revenue was fake, and it wasn't.

What he got wrong, or hasn't proven:

• The path nearly wiped out the trade. A call that goes +116% against you before it works is right on paper and very hard to survive in practice. Block hasn't said how much of the position Muddy Waters held through the peak. The firm's standard disclaimer says it often covers a large part of a short soon after publishing. Most readers who followed the trade at $340 were likely forced out well before $270.
• The mechanism didn't happen. Block's case depended on Google and Meta enforcing their rules, or regulators stepping in. Neither platform has publicly cut AppLovin off, and the SEC walked away. The stock fell because growth disappointed and competition got scarier, which is an ordinary momentum unwind, not a cheater being caught.
• Margins aren't "razor thin." They're 80%-plus.

On timing: he made no promise on timing. That protects him on paper. Grading fairly, the market eventually agreed with the direction while mostly ignoring the argument.

Have They Changed Their Tune?

No. He has doubled down twice.

CNBC's The Exchange, "Nvidia on Deck, Muddy Waters' New Long Position & Read on Retail 11/19/25" (Nov 19, 2025). The host noted the stock had "gone a lot higher" since March. Block didn't budge:

"With AppLovin, our thesis, we fully stand by it that they're cheating and they're doing things that are going to lead to them having to significantly dial back their growth because they're violating terms of service and numerous laws."

He blamed the rally on the environment: "anything that calls itself AI or AI adjacent where they can do no wrong and nobody cares," helped by S&P 500 index buying. APP closed at $530.05 that day. It's down 47.4% since, against the S&P's +17.7%. Doubling down there was a very good call.

Monetary Matters with Jack Farley, "Why Carson Block Won't Short AI Names Until the IPOs Begin & Muddy Waters' Pivot to Long Resources Stocks & S&P 500 Momentum" (Feb 4, 2026). Block admitted the market had beaten him for a while. Muddy Waters published on two new names in the first half of 2025, one of them AppLovin, and "those things just absolutely ripped the second half of the year," so the firm stopped publishing new shorts. He narrowed what the short was really about:

"We weren't attacking the core business, the gaming business. It was the blue sky story."

And he showed some self-awareness:

"I guess maybe this is just me telling myself what I want to hear, but I think it's one of those things where 50 million Elvis fans can't be wrong... eventually those chickens come home to roost, but in this super frothy market where everybody is anesthetized to risk, they're not coming home to roost just now."

APP closed at $387.34 that day, after a 16% drop in a single session. It's down 28.0% since, against the S&P's +13.6%.

One more detail from that interview: Block said his firm's long-only momentum strategy, which buys 20 S&P 500 stocks and rebalances monthly, was "up about 70%" gross in 2025. "It's just stupid," he said, but it's "fantastically easy money compared to... activist short selling."

Running Scorecard

Carson Block

Track Record so far (all investors):

Prices: FactSet daily closes for APP and the S&P 500, March 20, 2025 to October 6, 2026. Analyst and company items: The Fly and MT Newswires, January to September 2026.