Long share fell from 64% to 16% overnight, and Bitcoin moved only 0.2%. What happened to top traders?
Columns · · Satoshi Note Editorial
Read in: 한국어
Long share fell from 64% to 16% overnight, and Bitcoin moved only 0.2%. What happened to top traders?
Ten snapshots of the 24-hour group versus the 7-day and 30-day groups
As of yesterday evening, the top 24-hour traders held more money on the long side than on the short side.
In the snapshot taken at 3:51 p.m. KST on November 1, this group held about $774 million in longs and about $428 million in shorts. Longs were 64.4% of the total.
About 15 hours later, at 6:47 a.m. KST on November 2, longs were about $363 million and shorts about $1.97 billion. The long share was 15.6%.
Over the same stretch, Bitcoin went from $78,137 to $78,296, a gain of 0.2%.
Did top traders all switch sides while the price barely moved?
What is interesting is that the 7-day and 30-day groups, read at the same moments, hardly moved at all.
The numbers show less about how quickly top traders change their minds, and more about how quickly a group called "top 24-hour performers" can change its membership.
1. Across ten snapshots, the 24-hour group's long share ranged from 7% to 82%
Here are ten Satoshi Index snapshots between October 28 and November 2, side by side.
| Snapshot (KST) | 24-hour long share | 7-day long share | 30-day long share | Bitcoin price |
|---|---|---|---|---|
| Oct 28, 07:38 | 73.8% | 10.4% | 9.0% | $80,912 |
| Oct 28, 17:21 | 81.6% | 11.4% | 9.3% | $81,606 |
| Oct 29, 15:31 | 7.1% | 8.5% | 8.7% | $80,003 |
| Oct 30, 05:54 | 18.6% | 17.6% | 11.5% | $78,520 |
| Oct 30, 15:28 | 20.6% | 15.9% | 12.4% | $77,950 |
| Oct 31, 06:48 | 11.5% | 13.6% | 7.6% | $77,286 |
| Oct 31, 13:58 | 9.8% | 14.8% | 7.3% | $77,255 |
| Nov 1, 05:23 | 68.4% | 16.2% | 7.9% | $77,491 |
| Nov 1, 15:51 | 64.4% | 16.2% | 8.7% | $78,137 |
| Nov 2, 06:47 | 15.6% | 15.5% | 8.7% | $78,296 |
Long share is the long notional divided by long plus short notional. The Bitcoin price is the Hyperliquid mid price read at 8:41 a.m. or 5:41 p.m. KST for each edition, so it can be a few hours away from the snapshot time.
The 24-hour group went from 81.6% on the evening of October 28 to 7.1% in less than two days. It climbed back to 68.4% early on November 1 and stood at 15.6% this morning.
Across the same ten snapshots, the 7-day group stayed between 8.5% and 17.6%, and the 30-day group between 7.3% and 12.4%.
While the 24-hour group swung between 7% and 82%, the other two groups were more than 80% short the whole time.
One number is missing from the table.
The total position size of the 24-hour group rose from about $1.2 billion yesterday evening to about $2.3 billion this morning, nearly double. The mix changed, and so did the amount of money the snapshot captured.
2. Did top traders drop their longs and open shorts?
The first explanation that comes to mind is that the same traders flipped their positions within 15 hours.
This data cannot support that conclusion.
The Satoshi Index's 24-hour group is not a fixed list of traders. At each snapshot it takes the 100 accounts per exchange with the best performance over the last 24 hours.
The accounts on the list yesterday afternoon and the accounts on the list this morning can differ substantially.
So the move from 64% to 16% could be the same people changing sides, or different people entering the ranking, or a mix of both.
The current records do not follow accounts' past positions, so these three cases cannot be told apart.
The near-flat 7-day and 30-day groups fit this point. The longer the window, the harder it is for the ranked accounts to be replaced wholesale within a day.
A change in the 24-hour group has to be read as a question of who entered the ranking as much as who changed a position.
3. Bitcoin stayed inside a 5.6% range
Across the ten snapshots, Bitcoin traded between $77,255 and $81,606. The gap between the high and the low was about 5.6%.
The highest price, on the evening of October 28, coincided with the highest long share of the 24-hour group, 81.6%.
At the next snapshot on October 29, the price was down 2.0% at $80,003, and the long share was 7.1%.
That fits a common reading: whoever guessed the last day's direction rises in a 24-hour performance ranking.
But the latest 15 hours do not fit that reading well.
At 8:41 this morning, Bitcoin was up 1.1% over 24 hours and Ether up 0.9%, yet the top 24-hour group was 84% short.
A group's performance is a blended return across many assets and different leverage, not a bet on Bitcoin alone, so price direction by itself does not explain the ranking.
That is why this record cannot say whether the shift led the price, followed it, or had nothing to do with it.
4. In 15 hours, money moved much more than headcount
Splitting the 24-hour group by asset, yesterday evening versus this morning, shows one more thing.
| Asset | Long share by notional, yesterday evening | Long share by notional, this morning | Long / short accounts, yesterday evening | Long / short accounts, this morning |
|---|---|---|---|---|
| Bitcoin | 77.0% | 21.3% | 134 / 63 | 113 / 99 |
| Ether | 82.5% | 16.8% | 115 / 53 | 96 / 95 |
| HYPE | 26.1% | 3.5% | 61 / 57 | 41 / 83 |
| ZEC | 27.4% | 2.3% | 38 / 41 | 27 / 87 |
| NEAR | 86.2% | 6.5% | 35 / 26 | 22 / 71 |
Snapshots taken at 3:51 p.m. KST on November 1 and 6:47 a.m. KST on November 2.
By account count, Bitcoin went from 134 longs and 63 shorts to 113 longs and 99 shorts. The share of long accounts moved from 68% to 53%, a drop of 15 percentage points.
By notional, longs fell from about $264 million to $127 million, less than half, while shorts grew from about $79 million to $469 million, nearly six times.
Ether looks similar. The share of long accounts fell from 68% to 50%, but the long share by notional fell from 82.5% to 16.8%.
The hidden figure is this: the gap in account counts shifted gently, while the direction of the money flipped completely.
That points to a small number of large positions setting the notional share. If a few newly ranked accounts hold large shorts, notional share can change a lot even when account counts barely move.
This snapshot data cannot show which accounts, holding how much, produced the change.
5. What this number does not separate
It is easy to read these numbers as "top traders turned from long to short."
That does not make the indicator meaningless. The speed of change in the 24-hour group is information in its own right.
It shows how much the current ranking leans on short-term performance.
Three things can be kept apart when looking at it. First, which window the top group comes from. Second, whether the share is by notional or by account count. Third, whether the change is the same people or a change in who is ranked.
The third cannot be checked with the current records. Checking it would require account-level history to be stored over time.
Wrapping up: a group with the same name may not be the same people
- The 24-hour group's long share was 64.4% yesterday evening and 15.6% this morning.
- Over the same hours the 7-day group went from 16.2% to 15.5%, and the 30-day group stayed at 8.7%.
- Across ten snapshots the 24-hour group ranged from 7.1% to 81.6%, while Bitcoin's price range was about 5.6%.
This record cannot say whether the change led the price, followed it, or was the choice of the same people.
Whose positions changed between long 64% and long 16% is a question this number cannot yet answer.
Check it yourself in the Satoshi Index
The Satoshi Index ranks public accounts on Hyperliquid, Lighter, dYdX and Pacifica by 24-hour, 7-day and 30-day performance, then shows how much the top 100 traders hold long and short in each asset.
No login is needed to view it.
https://satoshinote.io/traders
FAQ
Why can the 24-hour group's long share change so much in a day?
The 24-hour group is not a fixed list of traders. It is the 100 best-performing accounts per exchange over the last 24 hours at the time of each snapshot. When the accounts in the ranking change, the share can move sharply even if no one changed their position.
Why did the 7-day and 30-day groups barely move?
One possibility is that the longer the window, the harder it is for the ranked accounts to be replaced wholesale in a short time. In these ten snapshots the 7-day group stayed between 8.5% and 17.6%, and the 30-day group between 7.3% and 12.4%.
Can this share tell us where the market is heading?
No. It adds up only the top 100 positions on four exchanges, and other exchanges and spot holdings are not captured, so hedging cannot be seen either. This piece does not forecast direction.
Data basis: ten snapshots from 7:38 a.m. KST on October 28 to 6:47 a.m. KST on November 2. Accounts ranked in the top 100 by 24-hour, 7-day and 30-day performance on Hyperliquid, Lighter, dYdX and Pacifica were counted per exchange and summed by asset; rankings are recomputed every hour. Bitcoin and Ether prices are Hyperliquid mid prices. The 24-hour total of about $1.2 billion and $2.3 billion is long plus short notional. This is an analysis of market data, not investment advice.