Earnings Surprise Drift in Thailand: Both Sides Signal, Miss Side Dominates

Thailand's earnings surprise data tells a two-sided story with the local SET Index as benchmark. Positive surprises produce +1.87% cumulative abnormal return at T+63. Misses produce -1.14%. The Q5-Q1 spread of +4.28% is driven mainly by the bottom quintile collapse (-1.78% at T+63), but the top quintile contributes meaningfully too (+2.49%). Thailand ranks 6th globally in Q5-Q1 spread, behind India, Taiwan, Hong Kong, Japan, and Canada.

Contents

1. Method
2. The Strategy
3. What We Found
4. Positive vs Negative Surprise Drift
5. Quintile Analysis at T+63
6. Why the Benchmark Matters
7. What This Means in Practice
8. Run It Yourself
9. Limitations
10. Part of a Series
11. References

Data: FMP financial data warehouse, 2000–2025. Updated March 2026.

Method

The Strategy

Post-Earnings Announcement Drift (PEAD) documents that stocks continue drifting in the direction of their earnings surprise for weeks after announcement. Ball and Brown (1968) first established this. Bernard and Thomas (1989) confirmed the drift persists for up to 60 trading days.

Thailand's PEAD results shift significantly when measured against the local SET Index rather than a US-listed proxy. The local benchmark eliminates cross-currency noise and captures the actual market environment Thai investors face.

What We Found

Positive vs Negative Surprise Drift

Positive surprises build steadily from T+1 (+0.64%) through T+5 (+1.03%), T+21 (+1.86%), and plateau at +1.87% by T+63. The drift is genuine and sustained. Beats outperform the SET Index throughout the window.

Negative surprises produce moderate negative drift. The miss side is more moderate than markets like Japan (-2.43%) or the US (-2.00%) but still persistent at -1.14% by T+63.

A 54.5% miss rate reflects analyst estimates that are not systematically conservative. Companies can't rely on sandbagging estimates to manufacture beats, which makes genuine beats more informative.

Quintile Analysis at T+63

Q5's t-stat of +5.1 with a 54.0% hit rate is a statistically solid finding. Biggest beats in Thailand produce reliable positive drift over 63 trading days.

Q1 (-1.78%, t=-3.8) is the stronger signal in absolute magnitude, but Q5 (+2.49%) also contributes meaningfully to the spread. This is a two-sided market.

Q5-Q1 spread: +4.28%

Why the Benchmark Matters

Prior analysis using a US-listed proxy benchmark showed all quintiles producing negative absolute CAR. That was an artifact of comparing THB-denominated stock returns to USD-denominated benchmark returns. The cross-currency effect inflated apparent underperformance. Using the SET Index (local currency, same market) removes this distortion and reveals the actual PEAD signal.

What This Means in Practice

Thailand supports both sides of a PEAD strategy. Long Q5 beats (+2.49% at T+63) and short or underweight Q1 misses (-1.78%) both have a statistical basis. The miss side is larger in magnitude, but the beat side is reliable.

For long-only portfolios, focusing on Q4+Q5 events provides a meaningful positive tilt. The drift builds gradually through T+21 and T+63, giving a relatively wide entry window.

SET market cap filter and liquidity constraints apply. Thai stocks outside the top tiers by market cap may face execution challenges for the short side.

Run It Yourself

python3 earnings-surprise/screen.py --preset thailand

-- Recent large earnings surprises on SET (both beats and misses)
WITH deduped AS (
  SELECT
    es.symbol,
    es.date,
    es.epsActual,
    es.epsEstimated,
    (es.epsActual - es.epsEstimated) / NULLIF(ABS(es.epsEstimated), 0) AS std_surprise,
    ROW_NUMBER() OVER (PARTITION BY es.symbol, es.date ORDER BY es.lastUpdated DESC) AS rn
  FROM earnings_surprises es
  JOIN profile p ON es.symbol = p.symbol
  WHERE p.exchange = 'SET'
    AND p.isActivelyTrading = true
    AND es.date >= CURRENT_DATE - INTERVAL '90 days'
    AND ABS(es.epsEstimated) > 0.01
)
SELECT
  symbol,
  date,
  epsActual,
  epsEstimated,
  ROUND(std_surprise * 100, 1) AS surprise_pct,
  CASE WHEN std_surprise > 0.10 THEN 'BEAT' WHEN std_surprise < -0.10 THEN 'MISS' ELSE 'NEAR' END AS signal
FROM deduped
WHERE rn = 1
  AND ABS(std_surprise) > 0.10
ORDER BY date DESC, ABS(std_surprise) DESC
LIMIT 50

Run this query on Ceta Research →

Limitations

4,612 events is a small sample relative to the US (170,150) or China (21,330). Quintile sizes of roughly 900 events each are sufficient for the statistical findings, but individual years will show wider variance. Early years (2000–2010) had sparser coverage, so the 26-year aggregate is weighted toward more recent data.

The Q1 t-stat of -3.8 and Q5 t-stat of +5.1 both meet conventional significance thresholds, but practitioners should use conservative position sizing relative to larger-sample markets.

Short-side implementation on the SET faces practical constraints. Stock borrow availability and cost for smaller Thai names can limit Q1 short strategies. The signal is most actionable for reducing or avoiding Q1 exposure within an existing Thai equity allocation.

Part of a Series

• PEAD Across 14 Global Exchanges: The Drift Is Universal
• Earnings Surprise Drift on US Stocks: 170,000 Events, 26 Years

References

• Ball, R. & Brown, P. (1968). An empirical evaluation of accounting income numbers. Journal of Accounting Research.
• Bernard, V. & Thomas, J. (1989). Post-earnings-announcement drift: Delayed price response or risk premium? Journal of Accounting Research.
• Foster, G., Olsen, C. & Shevlin, T. (1984). Earnings releases, anomalies, and the behavior of security returns. The Accounting Review.

Data: Ceta Research, FMP financial data warehouse