$2.35 billion, $1.28 billion, $1.21 billion: Wells Fargo, Goldman Sachs, and Millennium just disclosed their Q3 13Fs, while Bitcoin ETFs saw $2.35 billion in outflows over two weeks—the same set of numbers, telling two opposite stories.

#高盛与Millennium各持超12亿美元比特币ETF (No. 5 trending topic, captured 11-17 at 14:58)

First, what happened: The 13F filings released by the SEC on Monday show that Wells Fargo held 49.58 million shares of IBIT, worth about $2.35 billion—more than double its 24.4 million shares in Q2. Goldman Sachs held about $1.28 billion in spot Bitcoin ETFs, and Millennium about $1.21 billion. Across the market, only Wells Fargo and Jane Street held more than those two. The three together held $4.84 billion.

Now for the other side of those numbers: this is a snapshot from September 30, when Bitcoin was at $83,700. Today (OKX, 15:00), it’s at $64,665—the same positions have shrunk by about 23% in seven weeks. What stings even more: according to SoSoValue, Bitcoin ETFs saw a combined $2.35 billion in outflows in the first two weeks of November, exactly equal to Wells Fargo’s holdings. A 13F is a snapshot, not a record of what happened in between. Whether they sold in October or November won’t be clear until their Q4 filings, due by February 16 next year.

Millennium’s trajectory is the most interesting: it held $1.7 billion last December, cut that to $741 million by March this year, trimmed again in June, then bought back up to $1.21 billion by the end of September—it sold, cut its position, then bought back in. Galaxy Research says nine major banks and hedge funds collectively cut their Bitcoin ETF exposure by 68,000 BTC in the first half of the year. Goldman Sachs and Millennium were among them, and both bought back in during Q3. As for Wells Fargo’s $2.35 billion, most of that is clients’ money, not the bank’s own trading capital—don’t mistake it for the bank going all in.

My take: A 13F tells you who was in the market at the end of September, not who’s still there today. Wells Fargo doubling its holdings shows that client demand is real, but the filing date means it can’t explain November’s outflows. I’ll revise my view based on two things: whether these three are still in their Q4 13Fs next February, and whether ETF outflows ease this week. If either changes, I’ll run the numbers again.

Do you think these major institutions that added to their positions at the end of September held on through November’s drop—or have they already bailed?

When the Q4 13Fs come out next February, I’ll immediately check whether these three sold. Follow me so you don’t have to wait for the trending headlines.

—Bao Bao