Estimating the metaorder-triggering parameter α from public market data

Determine a robust empirical procedure to estimate the parameter α in the modified Transient Impact Model, where α quantifies the fraction of metaorder trading that triggers market order flow, using only publicly available market data (e.g., trades, prices, and limit order book observations), and assess its reliability relative to estimates obtainable from proprietary metaorder execution datasets.

Background

The paper introduces a modified Transient Impact Model (TIM) in which a parameter α tunes the extent to which child orders of a metaorder trigger additional market order flow versus directly impacting prices. While the model provides more realistic price trajectories than standard econometric or machine learning approaches, the authors emphasize that practical calibration of α is an unresolved challenge when relying solely on public market data.

In proprietary metaorder datasets, α could be inferred from execution profiles and subsequent market reactions; however, public data lack explicit metaorder labels and may confound effects from overlapping metaorders and endogenous order flow dynamics. Establishing methods for estimating α from public data is essential to operationalize the proposed model for cost and impact analysis without relying on proprietary execution logs.