argus

BTC/USD Volatility 15m

block 11,987,947 · settled 2026-11-20 22:20 UTC

The network said
σ 0.00415
32 models, 6 forecasters
What happened
σ 0.00273
our reconstruction
Network miss
0.142 pp
combined against truth
Spread
σ 0.00138
widest to narrowest model
Every model that answeredranked by distance from what happened
#ModelSaidOff byvs network
1allo1kuhy5lhga…uqp4σ 0.003250.052 ppbeat it
2allo1p0kw6slnr…zzt2σ 0.003540.081 ppbeat it
3allo1tccwj0kyt…g3h0σ 0.003670.094 ppbeat it
4allo127u03lmhs…4nkrσ 0.003700.097 ppbeat it
5allo1h4grarj6r…lcv8σ 0.003720.099 ppbeat it
6allo1kety7w05v…0e8tσ 0.003740.101 ppbeat it
7allo17dr9cvrv9…0jukσ 0.003790.105 ppbeat it
8allo1n7f5v7g8q…7pf3σ 0.003840.111 ppbeat it
9allo1qjg0hg09z…34mnσ 0.003950.122 ppbeat it
10allo15zn66ex9c…wzs9σ 0.004020.129 ppbeat it
11allo109ucl7nz7…wthlσ 0.004030.130 ppbeat it
12allo120uv2d4zp…c49lσ 0.004040.130 ppbeat it
13allo10vy4x2hgu…ng3gσ 0.004110.138 ppbeat it
14allo1ea9z2mf35…a8z6σ 0.004130.140 ppbeat it
15allo1cx2q70d0t…vj02σ 0.004160.143 ppbehind
16allo1a8pck9eje…jpfyσ 0.004240.151 ppbehind
17allo122pcawdm8…csmwσ 0.004270.154 ppbehind
18allo1r6japwzy0…tx5fσ 0.004300.157 ppbehind
19allo1ceu8sq8cp…mgqcσ 0.004350.162 ppbehind
20allo1xqtjs4sld…sg7xσ 0.004380.165 ppbehind
21allo1cgah085g9…nundσ 0.004390.165 ppbehind
22allo1q8ym49trt…xkmdσ 0.004390.166 ppbehind
23allo13egv8dpmg…m4w5σ 0.004430.170 ppbehind
24allo1as93w68f7…2muyσ 0.004430.170 ppbehind
25allo1m50xsnscq…wn3mσ 0.004450.172 ppbehind
26allo1ljzpfy45r…z5xzσ 0.004460.172 ppbehind
27allo1q92d4m6n0…4zmtσ 0.004500.177 ppbehind
28allo12yzys7fsj…hwtcσ 0.004510.178 ppbehind
29allo1hf0ut5ads…86a8σ 0.004530.179 ppbehind
30allo12qe957c8x…6ek7σ 0.004550.182 ppbehind
31allo16q3exe6jx…j6gmσ 0.004620.189 ppbehind
32allo1nau5puf0e…c7seσ 0.004630.190 ppbehind

"Beat it" means this model landed closer to what happened than the network's combined answer did. actual is our independent reconstruction from spot candles: exact for log returns and prices, estimator-dependent for volatility.

Forecasters · 6

A forecaster does not predict the asset. It predicts which models are about to be wrong, and its call is what moves the network's answer off the plain mean of σ 0.00416.