Compound Poisson aggregate claims have independent increments
ProvedKyprianou.Ch01.aggregateClaims_hasIndependentIncrementsLet be i.i.d. exponential interarrival times of rate and i.i.d. marks with law , all jointly independent (a compound Poisson driver). Let be the number of arrival times in , and let
be the aggregate claims (driftless compound Poisson) process. Then has independent increments: for every finite monotone grid , the increments
are mutually independent.
This is the independent-increment half of the statement that the compound Poisson process is a Lévy process. It holds for the concrete path built from an arbitrary rate and mark law on an arbitrary probability space. Combined with a deterministic drift, it gives the Markov and martingale properties of the Cramér–Lundberg and M/G/1 processes.
Formalization Note. is aggregateClaims arrival claims, with arrival the interarrival sequence of IsCompoundPoissonDriver. Independence is HasIndependentIncrements from LeanLevy, stated for the raw paths; no almost-sure modification is used.
import Mathlib import Definitions.Def_Kyprianou_Ch01_RiskQueueCore import Definitions.Def_LeanLevy_Processes_CompoundPoisson import Definitions.Def_LeanLevy_Processes_StochasticProcess open MeasureTheory ProbabilityTheory Kyprianou.Ch01 Set Filter open scoped NNReal ENNReal BigOperators set_option autoImplicit false
theorem Kyprianou.Ch01.aggregateClaims_hasIndependentIncrements
{Ω : Type*} [MeasurableSpace Ω]
(P : Measure Ω) [IsProbabilityMeasure P]
(arrival claims : ℕ → Ω → ℝ)
(rate : ℝ≥0) (hrate : 0 < rate)
(F : Measure ℝ) [IsProbabilityMeasure F]
(hDriver : IsCompoundPoissonDriver arrival claims rate F P) :
HasIndependentIncrements (fun t : ℝ≥0 => aggregateClaims arrival claims t) P := by
sorry