Marc Cohodes on SBF, Fraud, & The FTX Death Spiral | #456

Episode Summary

Executive Summary: The episode centers on short seller Mark Cohodes explaining how he identified FTX as a likely fraud months before its collapse, emphasizing obvious red flags, poor corporate controls, weak diligence by major investors and media, and the broader failures of regulation and journalism. He also reflects on his fraud-finding method, past shorts, and how skeptical analysis can help investors avoid hype.

Main Topics: How Cohodes spotted red flags in FTX (Priority: 5/5): Cohodes says SBF’s explanation of his wealth, trading strategy, team, and sources of capital never made sense, and that multiple details—especially around Alameda, FTX’s employees, and related-party behavior—suggested a constructed narrative rather than a real business. Institutional and media failure around FTX (Priority: 5/5): He criticizes Sequoia, other venture firms, Bloomberg, CNBC, and the New York Times for failing to independently verify basic facts and for being seduced by prestige, access, and narrative rather than diligence. Fraud detection as a craft (Priority: 5/5): Cohodes explains his process: look for small things that do not add up, ask simple questions, verify claims through competitors and documents, and wait for businesses to break rather than shorting purely on valuation. Role of regulators and enforcement (Priority: 4/5): He argues the SEC has become toothless, that crypto should be regulated through banking on-ramps/off-ramps rather than impossible-to-control tokens, and that enforcement should target the infrastructure enabling abuse. Short selling as purpose-driven work (Priority: 4/5): Cohodes frames short selling as a thankless but socially useful activity that can expose fraud and protect ordinary investors, contrasting committed skeptics with ‘smash and grab’ short sellers. Memorable past shorts and lessons (Priority: 3/5): He revisits Learnout & Hauspie and Canandaigua Wine as formative examples that taught him persistence, patience, and how to identify when a business narrative is false or a fad is unwinding.

Key Arguments: SBF’s story did not cohere: his explanation of how he made money, who trained him, and how FTX/Alameda operated lacked specifics and basic logic. The FTX employee base looked weak; Cohodes describes many hires as effectively glorified interns, inconsistent with a supposedly complex exchange. Red flags multiplied when he learned FTX had a poker-cheating lawyer as chief regulatory officer and that other insiders had questionable backgrounds. Media organizations avoided investigating because doing so risked losing access; prestige and convenience trumped skepticism. Venture firms and other investors likely relied on brand-name investors and FOMO instead of independent due diligence. The SEC and mainstream institutions failed to protect the public; citizen journalists and short sellers played a more effective watchdog role. Crypto is best regulated indirectly through banks and payment rails, since token markets themselves are difficult to police. Short sellers should wait for operational deterioration and business misses rather than simply betting on expensive-looking stocks. Fraud investigations rely on simple logic checks and comparative analysis, not complex models alone. The episode’s broader lesson is that investors often skip basic skepticism when a story is wrapped in celebrity, politics, and social proof.

Data Points: Recorded date: Monday, November 21 - The host notes the show was recorded on this date because the FTX story was evolving rapidly. Initial tweet on FTX: August 1 - Cohodes says he tweeted on August 1 that SBF/FTX was the best short on the board. FTX CEO age: 30 years old - Used to argue that someone reportedly worth billions at that age should have been demonstrably special, which Cohodes says SBF was not. FTX valuation: $32 billion - Referenced as the valuation level reached before the collapse and as evidence of huge investor credulity. FTX token price collapse: From about $35 to a penny or less - Cohodes describes FTT token unwinding after Binance/CZ moved against it. Learnout & Hauspie stock move: 35 to 120 to 0 - Historical short example used to illustrate a major fraud that ultimately went to zero. Learnout & Hauspie market cap: $12 billion - Cohodes says it was the largest fraud in Europe at the time. FBI pages on Cohodes: 1,168 pages - He recounts a FOIA fight where the FBI first claimed 4 pages, then corrected to 1,168. Number of views: 5 million views - He says a Hedgeye video on FTX reached about 5 million views. Creditor claims: 20 creditors claiming nine-figure-plus damages - Host mentions this to underscore the breadth of FTX fallout. Silvergate deposits: Less than $10 billion - Cohodes estimates deposits around this level when discussing leverage and crypto banking risk. Silvergate leverage: North of 100:1 - He argues the bank’s on/off-ramp business model implies extreme leverage and fragility. Canandaigua Wine short: Shorted at 35, covered at 7 - A formative short that later became a long and a major personal win. Canandaigua/Constellation stake: 13% ownership - He says his firm accumulated a large stake after the turnaround began. Canandaigua stock move: 3 to 60 - The long outcome after the turnaround and acquisition path.

Pivotal Quotes: "Never in my career have I seen such a complete failure of corporate controls and such a complete absence of trustworthy financial information as occurred here at FTX." — John Ray: Quoted in the bankruptcy discussion to capture the severity of FTX’s control failures. "I think Sam Bankman Freed will make Bernie Madoff look like Jesus Christ." — Mark Cohodes: Cohodes’s early and stark characterization of what he believed FTX could become as a fraud. "Don’t short something because you think it’s too high or it’s expensive." — Mark Cohodes: His core advice on short selling: wait for operational deterioration and evidence of failure.

Implications: The episode argues that major frauds thrive when media, VCs, and regulators trade diligence for access and narrative. For listeners, the lesson is to slow down, verify basics, and treat “too good to be true” claims as disqualifiers.