Who is holding the contracts — and what happens when they let go.
HC measures the breadth of participation on each side of the market — specifically, whether the long side or the short side has more accounts (participants). When the long side has very few participants relative to the short side, those few long accounts tend to be larger, more informed players while the crowd is positioned against them. The inverse is also true.
This is normalized against recent history, so the reading reflects how unusual the current imbalance is — not an absolute number of accounts.
| HC Value | Meaning |
|---|---|
| Positive / Green | The long side has unusually few participants — concentrated among fewer, likely larger accounts. The crowd is predominantly short. |
| Near zero | Participation is balanced — neither side is notably more concentrated than the other |
| Negative / Red | The short side has unusually few participants — concentrated among fewer, likely larger accounts. The crowd is predominantly long. |
The absolute value tells you how extreme the imbalance is. BlaveClaw labels these levels as: Concentrated → Highly Concentrated → Extremely Concentrated. Extremely concentrated readings are the most actionable — in either direction.
HC supports two opposing strategies, and which one works depends entirely on the concentration level.
When HC is positive and concentration is moderate-to-high, follow the large players: go long. When HC is negative, go short. The thesis is that informed money leads the market.
signal = 1.0 if HC > threshold else 0.0
When HC reaches extremely concentrated, the large position is already fully built. The next move is unwinding — which creates a sharp move against the concentrated side.
signal = 0.0 if HC > extreme_th else NaN
A common mistake is applying the contrarian approach at moderate concentration levels. Moderate concentration just means institutions are in position — the unwind hasn't started. The contrarian setup only applies at extreme readings where the concentration is unsustainable.
The single most reliable HC signal is not the absolute level, but the direction of change. A concentration that is decreasing (shrinking bars) tells you the large position is being unwound.
| Pattern | What it signals |
|---|---|
| Green bars growing | Institutions adding longs — bullish |
| Green bars shrinking | Institutions taking profits on longs — potential bearish reversal ahead |
| Red bars growing | Institutions adding shorts — bearish |
| Red bars shrinking | Institutions covering shorts — potential bullish reversal ahead |
The shrinkage signal typically leads price by one to several periods. It's most valuable as an exit signal for trend-following positions: if you're long because HC was positive, the moment HC starts shrinking is your earliest warning that the institutional support is leaving.
HC tells you who holds the position, but not what's driving current price action. TI tells you the immediate flow. The combination is more powerful than either alone:
| HC | TI | Interpretation |
|---|---|---|
| Positive (longs concentrated) | Positive (net buying) | Institutional longs being backed by active buying — strong bull signal |
| Positive (longs concentrated) | Negative (net selling) | Institutions hold longs but retail is selling — watch for HC shrinkage as confirmation of reversal |
| Negative (shorts concentrated) | Positive (net buying) | May be short liquidations — TI spike is forced, not genuine. Check if HC is also shrinking. |
| Negative (shorts concentrated) | Negative (net selling) | Institutional shorts with active selling pressure — strong bear signal |
The third row is the key edge: when you see a TI spike during a downmove, you might assume buyers are stepping in. But if HC is deeply negative and not shrinking, the TI spike is probably short liquidations — forced market buys that will end when the liquidations are done, not genuine accumulation.
Every HC reading comes with a stat object that gives historical context for that reading level:
| Field | What it tells you |
|---|---|
up_prob | Historical probability of price being higher 24h later when HC was at this level |
exp_value | Expected 24h return (%) based on historical outcomes at this HC level |
return_ratio | Average upside / average downside at this level. >1 = historically favored longs |
is_data_sufficient | Whether there's at least 1 year of history at this level. If false, treat stats as illustrative only. |
up_prob of 0.68 means that in 68% of historical cases with a similar HC reading, price was higher 24h later. It doesn't mean 68% of your next 10 trades will win.
Treating HC as a directional signal in isolation. HC = +2.5 does not mean "buy now." It means large longs are concentrated. If price has already moved 20% in their favor, the trade opportunity may have passed. Always contextualize with price action and TI.
Going contrarian at moderate concentration. The contrarian setup requires extreme concentration, not just elevated concentration. Going short because HC is "a bit high" while institutions are still building their position is fighting informed money.
Ignoring shrinkage speed. Slow, gradual shrinkage over many periods is normal position rotation. Rapid shrinkage in 1–2 periods is a forced unwind — the signal urgency is very different.
