CommitTrader Research

Executive Summary

Total Events Analyzed

1,413

Valid Event Studies

1,413

Companies Analyzed

31

Repositories Tracked

47

Metric	Value
Mean AR (Day 0)	0.0773%
Median AR (Day 0)	-0.0175%
Mean CAR (-5, +5)	0.1257%
% Events with Positive AR	49.3%
Standard Deviation	2.1619%

Key Findings

Research Question

Does public GitHub activity from open-source repositories associated with publicly traded companies have measurable impact on stock prices?

Main Results

Mean Abnormal Return (Day 0): 0.0773%
Mean CAR (-5, +5): 0.1257%
Events with Positive Returns: 49.3%

Statistical Significance

Test	P-Value	Significance	N
t_test	0.1794	ns	1413
sign_test	0.6321	ns	N/A
wilcoxon_test	0.8225	ns	N/A
cross_sectional	0.1794	ns	1413
CAR_0_0_test	0.1794	ns	1413
CAR_0_1_test	0.8090	ns	1413
CAR_-1_1_test	0.9373	ns	1413
CAR_0_5_test	0.5458	ns	1413
CAR_-5_5_test	0.5023	ns	1413

Methodology

Event Study Design

This analysis employs standard event study methodology to measure abnormal stock returns around GitHub release events. The market model is used to calculate expected returns:

E(R_it) = α_i + β_i × R_mt

Where R_it is the stock return, R_mt is the market return (S&P 500), and parameters are estimated over a 100-day estimation window (-130 to -31 days before the event).

Statistical Tests

t-test: Parametric test for mean abnormal returns
Sign test: Non-parametric test based on proportion of positive ARs
Wilcoxon test: Non-parametric rank-based test
Cross-sectional test: Tests for abnormal returns using cross-sectional variance

Event Windows

Multiple event windows tested: (0,0), (0,+1), (-1,+1), (0,+5), and (-5,+5) days

CommitTrader Research Report