Research

Statistical research and market analysis projects.

Quantitative Research Projects

Smart Whales publishes statistical research on financial markets with emphasis on correct methodology and transparency. All analyses are open-source and freely available.

Thanksgiving Alpha

Period: 2000-2024 (25 years)
Scope: 8,293 observations, 354 stocks (S&P 500, NASDAQ-100, DJIA)

Analysis of stock seasonality around the American Thanksgiving holiday. Results show positive median returns in 80-90% of stocks, dominated by technology and consumer discretionary sectors.

Key findings: SHOP (+3.36%), ENPH (+3.61%), PANW (+3.05%). Consistency confirmed across all three indices.

Executive Summary (EN) Executive Summary (CZ)

Santa Claus Rally

Period: 2000-2024 (25 years)
Statistical significance: 2 DJIA stocks (DIS, JPM) after Benjamini-Hochberg FDR correction

Quantitative analysis of the 7-day Santa Claus Rally effect. Results show statistically significant returns for DIS (+2.55%, p=0.037) and JPM (+1.97%, p=0.037) with 72% win rate.

Sector trends: Financial services, semiconductors, consumer discretionary. Broad positive effect across 81.7% of S&P 500 stocks.

Executive Summary (EN) Executive Summary (CZ)

WhaleVibe – Production Portfolio

The WhaleVibe portfolio includes 36 strategies deployed in a production environment. Each is optimized on 11+ years of historical data with validation on fresh data. Strategies combine different technical approaches to identify trading opportunities:

Modeling Approaches

  • Volatility & Mean Reversion: Detection of extreme moves and mean reversion (Bollinger Bands, RSI, rolling sigma)
  • Trend Following & Reversals: Identification of trends and breakouts with multi-factor validation (trend flips, momentum confirmation)
  • Regime Detection: Adaptive strategies responding to market regimes (choppy vs trending) with dynamic switching
  • Multi-Factor Alignment: Combined signals (trend + price + momentum) with majority voting for robustness

Market Coverage

  • SPY (US Equity Index): 26 strategies (14 morning, 12 evening) – intraday trading with opening and closing price actions
  • FX (Currency Pairs): 13 strategies (7 morning, 6 evening) – trend breakouts, range reversion, volatility squeeze with global approach

Technical Components

  • Moving averages and Bollinger Bands for dynamic bands
  • RSI and custom "number" indicator for momentum detection
  • Gap reversion, pullback entry and regime-adaptive logic
  • Lagged/unshifted data processing for leak-free backtesting
  • Optimization on large datasets (8k+ observations) with grid search methodology