On-Chain Update: Small Wallets Are Buying the Dip — Bitcoin's Biggest Whales Are Not (July 12, 2026)

Bitcoin's long-term holders have flipped back to accumulation, but the buying is not coming from where you would expect. Glassnode data reviewed this week shows the strongest accumulation concentrated in the smallest wallets and in mid-sized entities — while the largest whale cohort, wallets holding more than 10,000 BTC, sits close to neutral. As of July 12, 2026, with bitcoin near $64,200, the cohort split is the most interesting thing on-chain.

This matters because 'whales are buying' has been doing heavy narrative work in the July rebound story. The data does not support the strong version of that claim.

What the accumulation trend score shows

Glassnode's Accumulation Trend Score rates cohorts from 0 (distributing) to 1 (aggressively accumulating), weighted by wallet size. The current readings split cleanly:

• Under 1 BTC (shrimp): roughly 0.8–0.9 — near-maximum accumulation.
• 100–1,000 BTC (mid-sized entities): also in the 0.8–0.9 range.
• Above 10,000 BTC (mega-whales): roughly 0.4–0.5 — essentially neutral.

Alongside this, the long-term holder cohort — coins unmoved for 155 days or more — has shifted from net distribution to net accumulation. The 30-day net position change has turned positive, with long-term holders adding somewhere between 50,000 and 100,000 BTC. That reverses the pattern seen through the spring, when long-term holders were net sellers into strength.

So the accurate summary is: retail and mid-tier are buying the dip with conviction; the biggest wallets are watching. It is broad-based, not whale-led.

Why the mega-whale neutrality matters

Wallets above 10,000 BTC are a mixed population — exchange cold storage, ETF custody addresses, sovereign and corporate holdings, and a shrinking number of genuine individual mega-holders. Their neutrality can be read two ways, and honesty requires stating both.

The bearish read: the entities with the best information and the deepest balance sheets are not stepping in at $64,000. If this were an obvious generational entry, they would be. Their absence is a vote of no confidence, or at minimum of patience.

The structural read: this cohort is increasingly polluted by custodial addresses. Coins sitting in an ETF's cold wallet do not 'accumulate' or 'distribute' in any behavioural sense — they move when creations and redemptions require it. As the ETF complex has grown, the mega-whale cohort has become a worse and worse proxy for the intentions of large, informed holders. A neutral reading may be measuring plumbing, not conviction. This is the same measurement problem that has been quietly degrading several on-chain indicators since 2024, and we discuss it at length in our guide to bitcoin valuation models.

The flow cross-check

On-chain accumulation is happening at the same time as the ETF complex has just recorded its first positive week since mid-May — roughly $197.4 million in net inflows for the five sessions ending July 11, ending an eight-week, $8.26 billion outflow streak. Those two facts sit together comfortably: domestic institutional flows stopped bleeding at the same moment that smaller on-chain buyers were already accumulating.

But the confirming signal is still missing. The Coinbase Premium — the spread that proxies U.S. spot demand — has remained negative for a record run of consecutive sessions since May 19, through both the crash and the recovery. Retail accumulation showing up on-chain while the Coinbase Premium stays negative suggests the marginal buying is happening offshore, not on U.S. spot venues. That is a genuine tension in the data, and we are not going to paper over it.

What to watch this week

The on-chain picture is now a secondary input. June CPI on Tuesday, July 14 at 8:30 a.m. ET is the primary one, with PPI on Wednesday and retail sales on Thursday, ahead of the July 28–29 FOMC. If the print runs hot, watch whether the small and mid-sized cohorts keep accumulating through the drawdown — that would be the first genuinely strong-hands signal of this cycle. If they capitulate, the July rebound was a bounce, not a base.

The mega-whale cohort flipping from 0.4–0.5 toward 0.8 would be the single most bullish on-chain development available right now. It has not happened.

How this compares with the spring

The reversal in the long-term holder cohort is the most concrete change. Through the spring, the 155-day-plus cohort was a net distributor — selling into strength, which is the normal and rational behaviour of holders sitting on large unrealized gains. That distribution was one of the quiet sources of supply that made the June collapse as deep as it was: long-term holders were selling at the same moment the ETF complex was redeeming and the basis trade was unwinding. Three sources of supply, no offsetting bid.

Two of those three have now stopped. Long-term holders have flipped to accumulation. ETF redemptions have paused and turned marginally positive. That is a genuine improvement in the supply-demand balance, and it explains a large part of the move from $57,750 on June 30 to roughly $64,200 today without needing to invoke any new buyer at all. Much of the July rebound is simply the absence of the sellers who were there in June.

That framing sets a low bar for disappointment. A rally that exists because selling stopped is fragile in a way that a rally driven by fresh demand is not. It only takes the return of one of those supply sources to reverse it — and a hot CPI print on Tuesday is exactly the kind of event that would bring the basis-trade unwind and the ETF redemptions straight back.

The distribution question

One further point, because it is being widely mis-stated. Bitcoin moving out of a large wallet is not the same as bitcoin being sold. Treasury reorganisations, custodian migrations and internal address consolidation all produce on-chain movement with zero market impact, and they are routinely reported as whale distribution. The reverse is also true: coins can be sold via an exchange's internal ledger with no on-chain footprint at all. Any conclusion drawn from raw wallet-balance changes, without checking whether coins actually reached an exchange deposit address, is unreliable. Exchange netflow is the cross-check, and it is the one most commentary skips.