Concept · The RiskModels engine (ERM3)
Explained risk (ER)
The variance share attributable to each level. ER(L1) + ER(L2) + ER(L3) + residual = 1.0 by construction, verified at runtime to ±0.1%.
In depth
The variance share each level explains. Because every level is fit by an honest orthogonal regression, the residual is uncorrelated with every factor — so the shares sum to exactly 1.0, checked at runtime to ±0.1%.
Formula
ER(L1) + ER(L2) + ER(L3) + RR = 1.0
Compute it with the API
GET /api/l3-decomposition
# pip install riskmodels-py
client.get_metrics("NVDA", as_dataframe=True)[["l3_market_er","l3_sector_er","l3_subsector_er","l3_residual_er"]]In the methodology
Explained risk: variance decomposition →Referenced by (0)
No working paper references this concept yet.