Risk Desk exposure, stress and tail report

As of 2026-09-12 · base currency USD · data mode synthetic · riskdesk 0.2.0

not degraded synthetic inputs

Input source: Synthetic energy book, not market data. Values are already in USD; the TTF leg was converted from EUR at 1.08 USD per EUR. Two hand-chosen shapes, neither fitted to a dated episode, neither carrying a probability, neither a forecast. Supply dislocation: European and Asian gas rise far more than North American gas, crude rises moderately, power follows gas by less. Demand collapse: every benchmark falls, crude and European gas furthest, power least.

Stress input SHA-256: 4d35aa34c68625b675ff6b6fd63ca684669c180c960a3f9dcc8092a7f0ba9fe2

Sign convention. Positive VaR and Expected Shortfall denote loss; a negative value denotes a gain and is not floored at zero. Stress P&L is signed, and loss is its negation, so a negative loss is a scenario the book gains in. P&L observations on the tail chart are profit-positive, which is why the VaR and ES markers sit on the negative side of that axis. Exposure market values are signed: a negative net value is a short position and is drawn on the short side of the zero line, while gross counts absolute values before any offsetting, so a share of gross is never negative.

Linear portfolio exposure

Method: signed base-currency linear position exposure. Gross counts absolute position values before offsetting; net does not. Leverage divides by the supplied NAV.

Measure	Value	Against NAV
NAV	12,000,000.00 USD
Gross exposure	14,100,000.00 USD	1.175x
Net exposure	6,000,000.00 USD	0.5x

Net exposure by asset

Share of gross exposure

Exposure input source: Synthetic energy book, not market data. Signed market values are already translated into USD: crude and Henry Hub legs are quoted in USD; the TTF leg was converted from EUR at 1.08 USD per EUR; the JKM-linked cargo and the power forward are stated in USD. No position is an option or a spread instrument.

Exposure input SHA-256: b9255219f647f09b25284129c5320f70d11b4e6cc748e7f041912b3decac8179

Explicit stress scenarios

Method: linear full-position scenario revaluation. No probability is attached to any scenario. The stressed NAV column is the supplied NAV moved by the scenario P&L.

Scenario	P&L	Loss	Stressed NAV
european_gas_supply_dislocation_shape	1,843,000.00 USD	-1,843,000.00 USD	13,843,000.00 USD
global_energy_demand_collapse_shape	-2,058,500.00 USD	2,058,500.00 USD	9,941,500.00 USD

european_gas_supply_dislocation_shape

global_energy_demand_collapse_shape

Historical tail risk

1000 observations, 97.5% confidence, 1 day horizon, tail probability mass 25 observations. Portfolio value 12,000,000.00 USD.

Measure	Value (loss positive)
VaR	511,773.00 USD
Expected Shortfall	617,885.16 USD

Tail input source: Synthetic daily P&L for the example energy book, not market data. Reproduce with numpy.random.default_rng(7).standard_t(4, 1000) * 180000, rounded to whole USD. It is a drawn sample from a heavy-tailed distribution, not an observed history, and it carries no claim about any real book. Each sample already spans the one-day horizon.

Assumptions carried from the inputs

Shocks are simple asset returns; all assets must have explicit shocks.
No probability is attached to a scenario. Derivatives and financing costs are outside this model.
Market values are signed and already translated into the stated currency.
Gross concentration uses absolute position values before offsetting; leverage divides by supplied NAV.
Positions are linear assets; derivative Greeks and counterparty netting are separate calculations.
Each P&L sample already spans the stated horizon; no square-root-of-time scaling.
Equally weighted empirical distribution; sample dependence and forecast validity are not tested.
Loss is positive; negative VaR or ES denotes a gain. Values are not floored at zero.
VaR uses skfolio's empirical order statistic; ES integrates fractional tail mass.