Over the past 14 days, Bitcoin’s realized cap has dropped by 3.2%, while the volume of coins moving from exchange wallets to unknown addresses surged to a 6-month high of 128,000 BTC. A single address cluster—which I’ve tracked since my 2022 Terra response—accumulated 4,700 BTC in three days. That’s not retail. That’s a signal. And it coincides with Michael Burry’s recent tweet: "Now is the time to buy the dip."
But Burry’s track record in crypto is mixed. He called the 2021 top but missed the 2022 bottom. The question isn’t whether he’s right—it’s whether the on-chain evidence supports his conviction.
Let me be clear: I don’t trade on Twitter. I query Dune. And over the past week, I’ve built a dashboard that tracks 14 metrics across four layers—realized cap, MVRV Z-score, SOPR, and exchange stablecoin reserves. The data doesn’t lie. But it does require interpretation.
Context: Burry’s Macro Bet
Michael Burry is famous for betting against the housing market in 2008. In crypto, he’s been a bear since early 2021. His recent statement—that "Bitcoin is the best inflation hedge" and "now is the bottom"—is a sharp reversal. The market reacted with a 4% pump, but the real story is under the hood.
Burry’s view rests on three macro assumptions: (1) the Fed will pivot by Q3 2025, (2) the dollar will weaken, and (3) Bitcoin’s hash rate growth will squeeze supply. I’ll test each with on-chain data.
But first, I need to establish my methodology. I use a standardized framework I developed during DeFi Summer: compare current cycle metrics to the 2019-2020 consolidation phase, not the 2021 mania. That’s the only honest baseline.
Core: The On-Chain Evidence Chain
Let’s start with realized cap. This metric sums the price at which each coin last moved. It’s a proxy for aggregate cost basis. Right now, realized cap sits at $380 billion, down 8% from the 2024 peak. That’s similar to the 2019 bottom when it dropped 12% before the 2020 rally. The drawdown is shallower this cycle—bullish on the surface.
But I dug deeper. Using my Dune query (link embedded in the dashboard), I filtered out exchange inflows and whale movements. What I found: the decline is driven by short-term holders (coins held <155 days) selling at a loss. SOPR for short-term holders is 0.97—meaning the average seller is taking a 3% loss. That’s capitulation, but not panic.
In the ashes of Terra, we saw SOPR drop to 0.85. This time, it’s higher. That suggests sellers are disciplined, not fearful. That’s a subtle but important distinction.
Next, MVRV Z-score. This metric compares market cap to realized cap, adjusted for volatility. A Z-score below 1 historically marks a bottom. Right now it’s 1.2. In 2019, it hit 0.8. In 2022, it hit 0.9. So we’re close but not there.
Burry’s bet implies the Z-score will drop further or that this cycle’s floor is higher. Both are possible, but I need to test the second hypothesis.
I calibrated a model using three prior cycles (2015, 2019, 2022) and added a new variable: the percentage of supply held by long-term holders (LTH). Currently, LTH supply is 14.5 million BTC, 74% of circulating supply. That’s an all-time high. In prior bottoms, LTH supply was around 12-13 million. The increase suggests strong hands are accumulating, not distributing.
But correlation isn’t causation. High LTH supply could also mean illiquid coins are stuck, not that new demand is flowing in. That’s the contrarian angle.
Let me show you the SQL snippet I used to isolate exchange outflows:
