Event study: return dynamics around extreme sentiment

Panel event study using within-ticker z-score of sentiment. Compare average CAR/AAR paths for extreme positive vs negative sentiment days.

Updated: 2026-04-05

Study

Event study • Status: LIVE

Updated: 2026-04-05

z threshold

2.0

window

N pos

13608

N neg

11870

warning

—

date range

2024-10-31..2026-04-03

Sample ticker

—

Tickers (panel)

507

Obs (panel)

261179

R² (TS)

—

R² (FE)

—

Key findings

Event-study profiles help distinguish immediate reaction vs post-event drift.
Asymmetries between positive and negative events can suggest non-linear responses to sentiment extremes.

Specification

AAR(τ) = mean return at event-time τ across events.
CAR(τ) computed relative to event date using log returns.

Data

Universe: S&P 500 tickers in your snapshot pipeline.
Frequency: Daily.
Returns: log(P_t) − log(P_{t−1}) from ticker JSON prices.
Sentiment: score_mean from your sentiment pipeline (ticker JSON).

Limitations

Timing: after-close articles can contaminate same-day results; predictive tests mitigate this but do not fully solve it.
Causality: results are descriptive; omitted variables and measurement error remain.
Trading: no transaction costs / slippage / capacity modeled.

References (minimal)

Newey, W. K., & West, K. D. (1987). A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix.
Petersen, M. A. (2009). Estimating standard errors in finance panel data sets.
MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of Economic Literature.

Methodology

Compute within-ticker z-score of score_mean.
Define events: z≥threshold (positive) and z≤−threshold (negative).
Compute average abnormal return (AAR) and cumulative abnormal return (CAR) paths around events: tau ∈ [−W..W].

Figures

Charts are from the sample ticker (not the full panel).

Series (sample)

No series available.

Sentiment (sample)

score_mean

No sentiment series available. Export results.series.score_mean.

Tables

Event study CAR (z≥2.0 / z≤−2.0): cumulative log return

Event type	CAR(+1)	CAR(+5)
Positive	-4.161e-4	-8.666e-4
Negative	-1.402e-4	-1.636e-4

Note: exported by the Python builder; formatting is intentionally compact.

Models

This study is primarily descriptive (figures/tables). No regression model outputs were exported.

Appendix

Raw exported objects (reproducibility / debugging).

Raw exported JSON (collapsed)

{
  "results": {
    "n_tickers": 507,
    "n_obs_panel": 261179,
    "event_study": {
      "stats": {
        "z_thr": 2,
        "window": 5,
        "n_pos": 13608,
        "n_neg": 11870,
        "dedup_gap": 0,
        "warning": "—"
      },
      "series": {
        "tau": [
          -5,
          -4,
          -3,
          -2,
          -1,
          0,
          1,
          2,
          3,
          4,
          5
        ],
        "aar_pos": [
          0.0005101570815763624,
          -0.00004081955967644973,
          -0.00004107875923609897,
          0.0003275094233872985,
          -0.00008636528108676883,
          0.004189741013740667,
          -0.0004160648819658812,
          -0.00020741016232403753,
          -0.000017514727420027708,
          -0.0004217843557251382,
          0.000196132116268917
        ],
        "aar_neg": [
          -0.0006425097615992088,
          -0.0009691409470284341,
          -0.0008713810881102337,
          -0.0010617182573540776,
          -0.0014193744044491328,
          -0.00849886720839026,
          -0.00014015765620113295,
          0.00017990924370211251,
          -0.00011020367900047485,
          -0.0003186589090092016,
          0.0002255399189789069
        ],
        "car_pos": [
          -0.0006694029049643447,
          -0.00015924582338798176,
          -0.0002000653830644311,
          -0.00024114414230053038,
          0.00008636528108676883,
          0,
          -0.0004160648819658812,
          -0.0006234750442899188,
          -0.0006409897717099483,
          -0.0010627741274350864,
          -0.0008666420111661712
        ],
        "car_neg": [
          0.004964124458541079,
          0.004321614696941868,
          0.003352473749913448,
          0.002481092661803206,
          0.0014193744044491328,
          0,
          -0.00014015765620113295,
          0.00003975158750097958,
          -0.00007045209149949527,
          -0.000389111000508696,
          -0.00016357108152978964
        ],
        "aar_pos_se": [
          0.0001468321754413668,
          0.000146192888293565,
          0.00015294877271477983,
          0.00014787595392344745,
          0.00015308119845865785,
          0.00015297962483160383,
          0.00014840862640122334,
          0.00014054047392110807,
          0.00014945071450468228,
          0.00014408348766943306,
          0.00014332665090023615
        ],
        "aar_neg_se": [
          0.0001700221641626554,
          0.0001630069057743741,
          0.0001670127505861431,
          0.00016621936985360735,
          0.0001847373784331844,
          0.0002025396024126403,
          0.00018240492178625594,
          0.00016470748915606583,
          0.0001616984489260161,
          0.00016567558987369636,
          0.0001681680646573039
        ],
        "car_pos_se": [
          0.00032792359601631956,
          0.0002952377188307074,
          0.0002579434224115485,
          0.00020899241572595073,
          0.00015308119845865785,
          0,
          0.00014840862640122334,
          0.000204968998865221,
          0.0002506378305037249,
          0.0002885579192011415,
          0.0003191947767610057
        ],
        "car_neg_se": [
          0.00038124920468026363,
          0.0003433587651207517,
          0.0002990230729244957,
          0.0002464107376459901,
          0.0001847373784331844,
          0,
          0.00018240492178625594,
          0.00023830240419327057,
          0.0002842008764913188,
          0.00032883993807754884,
          0.00036659307506187297
        ],
        "se_aar_pos": [
          0.0001468321754413668,
          0.000146192888293565,
          0.00015294877271477983,
          0.00014787595392344745,
          0.00015308119845865785,
          0.00015297962483160383,
          0.00014840862640122334,
          0.00014054047392110807,
          0.00014945071450468228,
          0.00014408348766943306,
          0.00014332665090023615
        ],
        "se_aar_neg": [
          0.0001700221641626554,
          0.0001630069057743741,
          0.0001670127505861431,
          0.00016621936985360735,
          0.0001847373784331844,
          0.0002025396024126403,
          0.00018240492178625594,
          0.00016470748915606583,
          0.0001616984489260161,
          0.00016567558987369636,
          0.0001681680646573039
        ],
        "se_car_pos": [
          0.00032792359601631956,
          0.0002952377188307074,
          0.0002579434224115485,
          0.00020899241572595073,
          0.00015308119845865785,
          0,
          0.00014840862640122334,
          0.000204968998865221,
          0.0002506378305037249,
          0.0002885579192011415,
          0.0003191947767610057
        ],
        "se_car_neg": [
          0.00038124920468026363,
          0.0003433587651207517,
          0.0002990230729244957,
          0.0002464107376459901,
          0.0001847373784331844,
          0,
          0.00018240492178625594,
          0.00023830240419327057,
          0.0002842008764913188,
          0.00032883993807754884,
          0.00036659307506187297
        ]
      },
      "table": {
        "title": "Event study CAR (z≥2.0 / z≤−2.0): cumulative log return",
        "columns": [
          "Event type",
          "CAR(+1)",
          "CAR(+5)"
        ],
        "rows": [
          [
            "Positive",
            -0.0004160648819658812,
            -0.0008666420111661712
          ],
          [
            "Negative",
            -0.00014015765620113295,
            -0.00016357108152978964
          ]
        ]
      }
    },
    "tables": [
      {
        "title": "Event study CAR (z≥2.0 / z≤−2.0): cumulative log return",
        "columns": [
          "Event type",
          "CAR(+1)",
          "CAR(+5)"
        ],
        "rows": [
          [
            "Positive",
            -0.0004160648819658812,
            -0.0008666420111661712
          ],
          [
            "Negative",
            -0.00014015765620113295,
            -0.00016357108152978964
          ]
        ]
      }
    ]
  },
  "methodology": [
    "Compute within-ticker z-score of score_mean.",
    "Define events: z≥threshold (positive) and z≤−threshold (negative).",
    "Compute average abnormal return (AAR) and cumulative abnormal return (CAR) paths around events: tau ∈ [−W..W]."
  ],
  "sections": [
    {
      "title": "Specification",
      "bullets": [
        "AAR(τ) = mean return at event-time τ across events.",
        "CAR(τ) computed relative to event date using log returns."
      ]
    },
    {
      "title": "Data",
      "bullets": [
        "Universe: S&P 500 tickers in your snapshot pipeline.",
        "Frequency: Daily.",
        "Returns: log(P_t) − log(P_{t−1}) from ticker JSON prices.",
        "Sentiment: score_mean from your sentiment pipeline (ticker JSON)."
      ]
    },
    {
      "title": "Limitations",
      "bullets": [
        "Timing: after-close articles can contaminate same-day results; predictive tests mitigate this but do not fully solve it.",
        "Causality: results are descriptive; omitted variables and measurement error remain.",
        "Trading: no transaction costs / slippage / capacity modeled."
      ]
    },
    {
      "title": "References (minimal)",
      "bullets": [
        "Newey, W. K., & West, K. D. (1987). A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix.",
        "Petersen, M. A. (2009). Estimating standard errors in finance panel data sets.",
        "MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of Economic Literature."
      ]
    }
  ],
  "conclusions": [
    "Event-study profiles help distinguish immediate reaction vs post-event drift.",
    "Asymmetries between positive and negative events can suggest non-linear responses to sentiment extremes."
  ]
}