(12), §3.2, p. 1331 — the expected delay on every exchange is W/(μβ_i)
ProvedFragLOB.Collapse.eq_12limit-order-bookp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1queueingworkload
Consider the fragmented limit order book market with exchanges, market-order rate and attraction coefficients . Let be a vector of queue lengths with positive workload , and let the market orders be split by the linear attraction model . Then the expected delay on every exchange is
So the one-dimensional workload determines the delays at all exchanges at once. This is the identity (12) of the paper and the first ingredient of state space collapse.
Formalization Note. For the delay is defined by (1) as , and the statement is the identity . At an empty exchange (), where (1) reads , the delay is defined as , so the claim there holds by that convention.
Preamble
import Mathlib import Definitions.Def_FragLOB_Collapse_Model open MeasureTheory
Formal statement
namespace FragLOB.Collapse
theorem eq_12 {N : ℕ} (M : Market N) (hM : M.Standing) (Q : Fin N → ℝ)
(hQ : ∀ i, 0 ≤ Q i) (hW : 0 < workload M Q) :
∀ i, delay M Q i = workload M Q / (M.mu * M.beta i) := by sorry
end FragLOB.Collapse
Source
Maglaras, Moallemi & Zheng, Operations Research 69(4):1324–1348 (2021), p. 1331, §3.2, eq. (12)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.