422 SMB and HML represent size and value factors, respectively. Abnormal returns are computed as: , =( , − , )−( , − , )̂ Cumulative abnormal returns (CAR) are then calculated over the event window: ,[ 1, 2] =∑ 2 = 1 , Statistical significance is evaluated using standard t-tests. 2.2 Event windows To capture both immediate and delayed market reactions, four symmetric event windows are analyzed; [-3, +3], [-7, +7], [-14, +14], [-21, +21]. This allows assessing whether the impact is instantaneous or evolves over time as additional information becomes available. 2.3 Sample selection The study considers ten major environmental or environmental-regulatory events involving large mining companies: VALE S.A. – Brumadinho dam collapse (2019), BHP Group – Samarco/Mariana tailings failure (2015), Rio Tinto – Juukan Gorge destruction (2020), Glencore plc – Konkola acid spill (2006), Anglo American – Minas-Rio tailings spill (2018), Newmont Corporation – Yanacocha mercury spill (2000), Imperial Metals – Mount Polley tailings breach (2014), Freeport-McMoRan – Grasberg contamination ruling (2018), Southern Copper – Río Sonora spill (2014), Teck Resources – Elk Valley selenium contamination (2021). For each event, one responsible firm and a set of peer firms were defined based on sectoral similarity. 2.4 Spillover analysis Spillover effects are evaluated using aggregated indicators across peer firms: Mean CAR, Median CAR, Fraction of peers with negative CAR and Fraction of peers with statistically significant CAR This approach allows identifying sectoral patterns even when individual firms do not exhibit statistical significance. 3. RESULTS
RkJQdWJsaXNoZXIy MTM0Mzk2