3.94 million BTC. That's what short-term holders are carrying right now — roughly 20% of circulating supply. Their average cost basis sits at $74,100. And their unrealized profit just slid from 19% to 15% in under two weeks.
Run the arithmetic: 74,100 × 1.15 = $85,215. That's the implied spot price the CryptoQuant data quietly assumes. Nobody printed that number. I derived it. And it matters, because it tells you the entire short-term holder cohort is sitting on a ~13% buffer above their collective break-even.
Cost basis breach detected. Watch.
This is not a buy signal. It's a tripwire map.
Short-term holders — STH, in on-chain shorthand — are addresses that have held BTC for less than 155 days. That threshold isn't arbitrary. It's the historical inflection point where the market stops treating a coin as new supply and starts treating it as conviction supply. Long-term holders, LTH, sit on the other side of that line.
The 155-day cut-off matters because it's not a clean line — it's a rolling window. Coins age out of STH status every single day, migrating into the LTH bucket. That mechanical drift means the STH cohort is constantly being diluted by exits, which makes any single-day reading a moving target.
The STH cost basis is the average price at which this cohort bought. It's the closest thing crypto has to a live, per-cohort break-even line. When spot trades above it, the group is in profit. When spot drops below it, the group flips to unrealized loss — and historically, that's when stop-losses cluster, when panic selling accelerates, and when a healthy pullback becomes a structural break.
The data comes from CryptoQuant, attributed to analyst Axel Adler Jr., with a snapshot dated October 4. Three things stand out. First, STH holdings climbed 87,000 BTC over 30 days — about 2.26%, a measured, non-explosive build. Second, that growth has been positive for seven consecutive weeks and has stayed above the prior month since August 18. Third, unrealized profit peaked at 19% on September 22 and has bled down to 15%.
Let me be blunt about the source before we go further. One platform. One analyst. No raw data links. No methodology disclosure. That's not a dealbreaker — CryptoQuant is a top-tier provider — but it's a caveat you carry into every conclusion below.
The cost basis moved the needle. Here's how. Let's do the math the original write-up skipped.
If STH hold 3.94 million BTC against a circulating supply near 19.7 million, that's almost exactly 20% of all liquid coins concentrated in the hands of the most price-sensitive cohort in the market.
A cohort holding 20% of supply at a single average price is a structurally fragile arrangement. The tighter the cost distribution around $74,100, the more violent the reaction when price tests it. If buyers stacked in a narrow band near $74k, you get a stampede. If they're spread across $60k–$90k, you get a cushion. The data doesn't tell us which. That's a hole.
Now the growth number. 87,000 BTC added in 30 days. In dollar terms, at an $80k+ spot range, that's roughly $7 billion of net new demand. Not parabolic. Not noise either. Seven straight weeks of accumulation means this is trend behavior, not a single-day impulse — exactly what you want to see if you're separating real bids from reflex buying.
Watch the profit compression math. Peak was 19% on September 22. Now 15%. That's a four-point drop in roughly two weeks while holdings kept climbing. New money kept entering, but the cushion it bought got thinner. Rising entry costs with falling unrealized gains means the price is no longer outrunning the basis — it's losing that race.
Here's where my audit background kicks in. I've spent two weeks inside Terraform Labs' transaction logs tracing the UST decoupling. I've dissected Parity's multisig before the press caught the reentrancy risk. The lesson from both: the headline number is never the story. The parameters are.
And this dataset has a black box. The 155-day STH threshold isn't stated. Neither is whether the figures are entity-adjusted — whether multiple addresses under one owner get collapsed into a single holder. Neither is whether exchange-internal transfers were stripped. Change any one of those settings and 3.94 million becomes a different number. The $74,100 moves too.
Without the methodology, you don't have a precise level. You have a directional claim wearing a precise costume.
Still — the logic chain holds. Cost basis is the cohort's average entry. Price below it means unrealized losses. Unrealized losses mean stops trigger. That sequence has been validated across cycles.
The part everyone's skipping: the 30-day change is the real leading indicator, not the price. Watch that number. If it rolls over from +87,000 to negative, the cohort has flipped from accumulation to distribution. That's the hard confirmation. Price is the lagging echo.
One more layer. The average cost has been climbing roughly $1,000 a week. That's high-price turnover accelerating. Early low-cost STH are exiting, likely rotating into LTH status, and new buyers are absorbing their coins at a higher basis. The cohort's break-even line is being dragged upward by its own composition.
Then there's the self-fulfilling trap. Once a level like $74,100 gets quoted widely enough, it stops being a statistic and becomes an order-book fact. Desks place stops there. Bots watch it. The most dangerous level in any market is the one everybody agrees on.
There's a DeFi overhang too. BTC is collateral across lending protocols. If spot slides toward $74,100, some of those positions hit liquidation thresholds before the STH cohort even flips to loss — and liquidations are forced sells, not optional ones. The cohort's break-even line and DeFi's liquidation cliffs don't sit at the same price. Whichever triggers first sets the tempo.
Cost up, profit down, holdings flat-to-up: that combination is the signature of a market that needs new money just to stand still.
Now the angle nobody's pricing. This is where it gets uncomfortable.
The data has no time anchor. The source is stamped October 4 — but the publication date is unclear. This isn't a footnote. It's the whole ballgame. The exact same data means opposite things depending on where in the cycle you are.
If this snapshot landed near a local top, then rising cost basis plus shrinking profit is a textbook distribution signature — new buyers paying up while momentum fades. Bearish.
If it landed after a pullback, the same numbers read as healthy handoff — old low-cost holders taking profit into LTH status, fresh demand stepping in. Bullish.
Same data. Opposite verdicts. The missing timestamp isn't a small oversight; it's a methodology-level defect that makes the conclusion unfalsifiable.
Then there's the crowded-trade problem. $74,100 is an oddly specific number, and that's precisely the danger. When thousands of desks read the same analyst and set stops at the same level, you don't get clean support — you get a liquidity magnet. Market makers see the cluster. A round-number stop shelf is a target, not a floor. The wick that pierces $74,100 and snaps back is the likelier outcome than a clean break.
Leverage rush vibes. Proceed with caution.
The number to watch isn't $74,100. It's the 30-day holding change. Accumulation turning to distribution is the signal that precedes the price move, not the one that follows it.
Until someone republishes this with the parameters attached and a date stamped on it, treat the level as a hypothesis. The cohort is 13% from a collective loss. Whether that 13% is a cushion or a cliff depends on a data point the source never gave us.