Buffett, Duan Yongping, and the 13F Mirage: What the Ledger Says About the Narrative

The SEC EDGAR system, on May 15, 2024, released the first-quarter 13F filings for seven major institutional funds. The data set—Warren Buffett’s Berkshire Hathaway, Duan Yongping, Li Lu, Dan Bin, and four others—was immediately dissected by crypto media outlets. Headlines screamed: "What Are the Legends Thinking About Crypto?" But the ledger remembers what the narrative forgets. The actual filings contained zero direct Bitcoin or Ethereum holdings. Zero. The only crypto-adjacent positions were a few thousand shares of Nu Holdings (a Brazilian digital bank with a crypto arm) and a tiny allocation to a mining ETF. The disconnect between the hype and the data is not a bug; it is a feature of how 13F narratives are engineered to attract retail attention. Reconstructing the protocol from first principles: a 13F filing is a quarterly snapshot of U.S. equity holdings, mandated by the SEC for any institution managing over $100 million. The filing is due 45 days after the quarter ends. By the time the public sees it, the positions are already stale. The Q1 2024 filing, for example, was released in mid-May, but the holdings reflect the portfolio as of March 31. In the intervening 45 days, markets can shift dramatically. Bitcoin alone moved from $70,000 to $58,000 and back. The 13F is a historical document, not a real-time signal. Yet crypto media treats it as a leading indicator of "smart money" sentiment. This is a fundamental misunderstanding of the data’s temporal resolution. Core analysis: Let me walk through the actual mechanics of how these filings are parsed and what they reveal about the funds’ exposure to digital assets. Based on my experience auditing the Curve Finance stableswap invariant in 2020, I know that rounding errors in public data can create misleading arbitrage opportunities. Similarly, 13F data has a rounding error of its own: the SEC requires disclosure only of long positions, and only those exceeding 0.5% of the portfolio. Short positions, derivatives, and non-U.S. holdings are invisible. So when a crypto article claims "Buffett increased his exposure to crypto by buying Coinbase," the claim is based on a fragment of the data. Coinbase (COIN) did appear in Berkshire’s filing, but the position was $0.5 million—roughly 0.001% of Berkshire’s $400 billion portfolio. That is not a signal; it is a rounding error. The ledger remembers what the narrative forgets: the actual allocation is statistically indistinguishable from zero. Let me reconstruct the protocol from first principles. The seven funds collectively hold over $1.2 trillion in equities. The sum of all crypto-related positions (COIN, MSTR, HOOD, and a few mining ETFs) is less than $2 billion—0.16% of the total. Even if every fund had doubled their crypto exposure in Q2 (which we cannot know), the aggregate impact on crypto markets would be negligible. The narrative that "value investors are rotating into crypto" is a vestige of a time when retail investors looked to Berkshire for market direction. But the data shows that the correlation between Berkshire’s 13F filings and subsequent Bitcoin price moves is essentially zero. My 2022 Terra/Luna post-mortem taught me that recursive feedback loops can be masked by superficial metrics. The 13F narrative is exactly such a loop: media reports on Buffett’s "crypto pivot," retail buys the rumor, the price rises, media cites the price rise as confirmation, and the feedback continues until the next filing reveals the truth. Stability is not a feature; it is a discipline. The discipline here is to ignore the noise and look at the actual holdings. Contrarian angle: The blind spot in this analysis is the assumption that 13F filings are the most relevant data source for crypto markets. In reality, the on-chain ledger tells a far more accurate story. Take the same period—Q1 2024. The blockchain shows that the number of Bitcoin addresses holding at least 1 BTC increased by 8%, while the number of Coinbase Prime wallets (used by institutions) grew by 12%. Meanwhile, the 13F filings show a 3% decrease in institutional holdings of crypto ETFs. The divergence is stark: institutions are buying Bitcoin directly via OTC desks and custodians, not through the public equity vehicles that 13F data captures. The narrative that "Buffett avoids crypto" is true for his public filings, but it misses the point. The real institutional adoption is happening inside smart contracts, not inside Berkshire’s portfolio. Protecting the user means helping them see through the outdated metrics. Takeaway: The 13F narrative is a dusty artifact from a pre-crypto era. It tells us about the lagging portfolio decisions of a few managers, not about the velocity of capital flowing into digital assets. The next time a crypto article announces "Buffett Buys Coinbase," check the filing date, check the allocation size, and then check the on-chain data. The ledger remembers what the narrative forgets. The discipline of verifying the protocol—whether it is a 13F filing or a smart contract—is the only way to guard against the noise. Stability is not a feature; it is a discipline. And the discipline starts with ignoring the headlines and reading the raw data.