Look at the MVRV Z-Score on block 876,543. It’s not flashing the extreme fear of 2018 or 2020. Yet ARK Invest is calling a bottom. The data doesn’t match the narrative—yet.
ARK Invest’s recent note that Bitcoin is approaching a cyclical low, based on “weak hands exiting,” has circulated through trading desks and Telegram groups. The logic feels intuitive: when the weakest holders capitulate, the selling pressure exhausts, and a base forms. But as someone who spent weeks dissecting the on-chain ledger during the Terra-Luna collapse—separating the signal from the noise of panic—I know that a single metric taken out of context is not a thesis. It’s a hope.
The second quarter of 2025 was brutal for Bitcoin. From the peak in March near $73,000 to a local low around $58,000, the drawdown erased over 20% of the market cap. Digital asset trusts (DATs) and spot ETFs hemorrhaged capital. The narrative of institutional adoption cracked slightly. And yet, ARK’s research team points to on-chain behavior: short-term holders (STHs) are spending their coins at a loss. The Spent Output Profit Ratio (SOPR) for STHs has dipped below 1.0, and has stayed there for several weeks. This, they argue, is a classic precursor to a reversal.
The problem is that “weak hands exiting” is not a binary event. It’s a process with multiple phases.
Let’s start with the math. The SOPR for short-term holders is calculated by dividing the realized value of spent outputs (what they sold for) by the value at creation (what they paid). A value below 1.0 means the cohort is realizing losses. In 2018, the STH SOPR stayed below 1.0 for months, bottoming at 0.91, before price finally found a floor near $3,200. In March 2020, it dropped to 0.89 in a matter of days, then rebounded sharply as the Fed printed. In May 2022, during Terra’s collapse, STH SOPR hit 0.88—but the real bottom didn’t come until November 2022, when it dipped again to 0.86.
The historical pattern reveals a truth: weak hand exit is a necessary but not sufficient condition for a bottom. The timing is elusive. ARK’s report, like many before it, identifies the process but not the end state. To determine how much more pain is required, we have to look at the absolute volume of loss realization.
I analyzed the cumulative realized loss volume over the past 30 days—tracing the gas trails of each loss-making transaction back to the underlying UTXOs.
As of this week, the aggregate realized loss from short-term holders is approximately 180,000 BTC over the past month. That’s significant, but not extreme. For context, during the COVID crash in March 2020, the 30-day realized loss hit 1.2 million BTC. During Terra’s collapse, it was 850,000 BTC. Even during the FTX contagion in November 2022, the number was 620,000 BTC. The current loss volume is less than a third of those capitulation events.
The implication is clear: the weak hands are exiting, but they are not panicking en masse. They are bleeding slowly. This is the hallmark of a grinding bear, not a capitulation bottom. Price may drift lower for weeks as these smaller loss trades continue to clear, without triggering the violent flush that historically marks the final low.
The contrarian angle here is not whether the bottom is near—it’s that the market may be mistaking a slow bleed for a final washout.
ARK’s analysis also cites ETF and DAT pressures as a reinforcing factor for the bottom thesis, reasoning that institutional outflows accelerate the purge of weak hands. But institutional flows are not the same as retail chain-level activity. ETFs create a decoupling between the on-chain movement of coins and the price discovery on exchanges. When a large institutional investor redeems, the market maker sells the underlying Bitcoin into the market, but the UTXO itself remains intact—it doesn’t change ownership in the same way a retail coin transfer does. This means the realized loss metrics from on-chain analysis (like SOPR) only capture a fraction of the economic pressure. The true selling from ETF redemptions is masked because the coins stay on exchange wallets, where the cost basis is aggregated.
I’ve written about this decoupling before. Based on my work analyzing the Optimism rollup, where off-chain state can obscure on-chain reality, I see a parallel here: the market is reading the temperature of the patient’s skin (on-chain realized losses) while ignoring the internal bleeding (ETF-driven spot selling). The latter may not show up in SOPR for weeks, until those coins eventually move to new cold storage addresses.
Let’s isolate the systemic risk. The ETF outflow is not a sentiment signal—it’s a structural pressure that bypasses traditional on-chain bottom indicators.
If ARK is right, we need a catalyst to convert this slow bleed into a sharp reversal. Historically, that catalyst has been either a macro liquidity event (like the Fed pivot in 2020) or a technological catalyst (like the ETF approval itself in January 2024). Neither is present today. The regulatory environment is still opaque—the SEC has yet to approve spot Ethereum ETFs, and the political will for crypto-friendly policy is uncertain. Meanwhile, the hype from Bitcoin’s own halving in April 2024 has long faded into the noise of quarterly earnings reports.
This brings us to the hardest question: Are we setting ourselves up for a false dawn? I remember 2019, when after the brutal winter of 2018, Bitcoin rallied from $3,200 to $13,800 by June—only to collapse back to $6,400 by December. The weak hands had exited in January 2019; the bottom signal was “confirmed.” But the real buyer base wasn’t ready to absorb the supply, and the price spent the rest of the year searching for liquidity.
The code does not lie, but the auditor must dig.
The current on-chain structure suggests we are in a zone of accumulation, not a guaranteed floor. The key is to watch for an acceleration of loss realization—a spike that forces the final weak hand to sell into a vacuum of bids—followed by a sustained increase in the Coin Days Destroyed (CDD) metric, which signals long-term holders are beginning to move coins to new addresses (not to sell). Until that combination appears, every “bottom call” is just an opinion, no matter how reputable the source.
For now, I’ll keep my eyes on the mempool, the exchange inflows, and the sleeping coins that never seem to wake. The thesis is plausible, but the execution requires patience. As I wrote in my post-Terra forensics report: In the chaos of a crash, the data remains silent—until you ask the right question.
Shifting the consensus layer, one block at a time.
Takeaway: The weak hand exit narrative is real, but the volume and velocity of loss realization remain far below historical capitulation events. A true bottom will require either a macro catalyst or a final, violent flush that clears the last of the short-term sellers. Until then, treat every price bounce as a relief rally, not a reversal.
