Skip to content
All library documents

Comparing Power-Law and Logarithmic Models of Stock and Warrant Price Impact

Article arXiv papers · Author: Hai-Chuan Xu et al.
Read the original

Summary

This study estimates the immediate price impact of market orders using order flow data for a stock and its warrant. It compares a power-law specification with a logarithmic specification, focusing on how reliably each model estimates its parameters and how well it forecasts out of sample.

The authors find the power-law model performs better on both measures. For filled trades, they also observe significant positive correlations between the estimated impact parameters for ask-side and bid-side orders, indicating that the two sides' impacts move together in their sample. These findings may help inform optimal execution decisions. The summary does not state the securities studied, sample period, effect sizes, or how the forecasts were validated, so it offers limited guidance on how broadly the model comparison applies.

Key ideas

  • The study compares power-law and logarithmic models for immediate market-order price impact.
  • Order flow data from a stock and its warrant supports the estimation.
  • The power-law model has more robust parameter estimates and better out-of-sample forecasts in the reported analysis.
  • For filled trades, ask-side and bid-side impact parameters show significant positive correlation.
  • The results may inform execution analysis, though the document gives few details about sample scope or validation.

Tags

EquitiesOptionsMarket microstructureExecutionStatistics
Full text
# Immediate price impact of a stock and its warrant: Power-law or logarithmic model?


# Immediate price impact of a stock and its warrant: Power-law or logarithmic model?









Based on the order flow data of a stock and its warrant, the immediate price impacts of market orders are estimated by two competitive models, the power-law model (PL model) and the logarithmic model (LG model). We find that the PL model is overwhelmingly superior to the LG model, regarding the robustness of the estimated parameters and the accuracy of out-of-sample forecasting. We also find that the price impacts of ask and bid orders are consistent with each other for filled trades, since significant positive correlations are observed between the model parameters of both types of orders. Our findings may provide valuable insights for optimal trade execution.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.