Abstract
We propose a Hybrid Piecewise Intensity model linking point-process econometrics to the physical architecture of electronic exchanges, decomposing trade arrivals in the 100 ms around macroeconomic announcements into four latency-aligned phases. Calibrated on nanosecond Eurex and CME data (2019--2025), modal Phase~1 peaks occur at 1.16 μs (Eurex) and 6.64 μs (CME), tracking each venue's network transmission floor. Comparing private embargo-based (PMI) and government-led (NFP, FOMC) releases, we find reaction-speed differences consistent with protocol-induced latency advantages, though confounds remain. The CME concentrates profitability within 200 μs with deteriorating risk-adjusted returns thereafter; Eurex sustains stable Sharpe ratios across longer horizons.

This publication has 27 references indexed in Scilit: