First time on the top traders desk?

The desk runs its own numbers and publishes its calls next to the account that trades them. The short route:

DESK — INDICATOR

Funding Stress

Perpetual funding tells you who's crowded. We score every coin's live funding against its own normal (z-score vs history), then read it through the edge we measured on 239,849 past funding prints: after extreme negative funding, price tends to keep falling: capitulation, not a bounce. Fading the crowd doesn't cleanly work.

longs pay to holdshorts pay — capitulation8-hour perp funding against that perp's own history
The idea. Funding is the rent one side pays the other to keep a perp position open. Bars far above a coin's own normal mean crowded longs; far below, crowded shorts. Extreme is a warning about positioning, not a direction call.

updated · live funding from perp markets · scored vs each coin's history · not financial advice

Crowded shorts · below-normal funding

Most-negative z. Historically a downside-continuation flag, not a bounce.
negative funding shorts pay longs — crowded shortz vs the coin's own normalz below −2 extreme, downside flagAPR annualised 8h rate

How to read a row → a coin at −0.045% funding, z −3.1: funding is far below this coin's own normal, shorts are crowded and paying up. The desk reads extreme-negative funding as a downside-continuation flag, not a dip to buy.

Crowded longs · above-normal funding

Most-positive z. Mild fade historically (median −0.19% next 8h).
positive funding longs pay shorts — crowded longz vs the coin's own normalz above +2 extreme, crowded longAPR annualised 8h rate

How to read a row → a coin at +0.08% funding, z +2.6: funding is well above this coin's own normal, longs are crowded and paying up. Historically only a mild fade (median −0.19% next 8h), so read it as caution, not a short trigger.

The edge (backtest, 239,849 funding prints · 76 coins · 2020–26): forward 8-hour spot return after each print, bucketed by how extreme funding was vs the coin's own normal. Extreme negative (z<−3): median −0.58%, 57% closed lower — crowded shorts / capitulation, price keeps sliding. Extreme positive (z>+3): median −0.19%, 52% lower — a mild fade. Net: the naive "fade the funding crowd" trade does not cleanly work; the desk reads extreme-negative funding as a downside-continuation flag, not a long.

Not another funding chart

Funding data is everywhere — CoinGlass, Coinalyze, Laevitas all show it. They hand you the raw number. This does four things they don't.
WHAT'S DIFFERENT

Every funding dashboard shows the rate. Almost none tell you whether it's unusual for that coin, what price did after, or how to trade it. We do:

Feature?Funding Stress?CoinGlass · Coinalyze · Laevitas?
Raw / aggregated funding rates
Funding heatmap, OI, liquidations breadth
Scored vs each coin's own history (z)
Measured forward-price edge published
Desk interpretation / how to trade it
$0 by promo, no login / accountpartial (pro tiers)
Open, spelled-out method
provided not provided / partial = limited or pro-tier only

How to read a row → Scored vs each coin's own history (z): Funding Stress ✓ · others ✗: only this page normalises funding to each coin's own baseline. The mainstream dashboards hand you the raw rate. That normalisation is the core edge this table is drawing out.

Honest read: they win on raw data breadth (heatmaps, order flow, 1000+ assets). We win on turning funding into a decision: normalised, back-tested, read, $0 by promo.

How to read it

Where it breaks: when open interest is falling. The rate prices what it costs to hold the crowded side; open interest is how much of that side exists. A high rate on shrinking OI is positioning being unwound at a high price, not crowding building. The squeeze it looks like has already happened. Read the two together or not at all.

z-score = how far a coin's live 8h funding is from its own average, in standard deviations. z ≈ 0 is normal; z < −2 means funding is unusually negative (shorts paying longs — crowded short); z > +2 means unusually positive (longs paying shorts — crowded long). APR annualises the 8h rate (×3×365). We rank by z, not raw rate, so a large-cap's 0.01% and a memecoin's 0.3% are comparable. When the whole board sits near zero, positioning is calm — no crowded side to fade or ride.

Triggers, not calls. Funding is one lens on positioning; pair it with price structure and liquidity. Backtest is spot forward-return, costs excluded, past behaviour ≠ future. Read what funding rate is.