Abstract

This report presents a comprehensive analysis of the Holiday Effect trading strategy — a calendar-based momentum anomaly exploiting pre-event price drift in Amazon (AMZN) stock around major shopping holidays. Over 1998–2025, the strategy demonstrates a Sharpe ratio of 0.54 with a 75.8% win rate across 33 trades. The strategy capitalizes on investor sentiment and revenue anticipation preceding Black Friday and Prime Day events. We analyze both equity long and options overlay implementations, evaluate risk-adjusted performance metrics, and provide detailed implementation guidelines. Pre-holiday returns are statistically significant at the 1% level (t-stat 3.96, p=0.0002).

Introduction

Calendar anomalies represent potential market inefficiencies that persist despite widespread knowledge. The Holiday Effect strategy investigates a specific seasonal pattern: pre-holiday price appreciation in Amazon stock before major shopping events. Amazon represents an ideal vehicle given its ~40% U.S. e-commerce market share and its ability to move broader retail sentiment.

The strategy employs a pure calendar-based approach: enter long 10 trading days before the event, exit the day immediately before the event, and repeat for Black Friday (late November) and Prime Day (mid-July).

Motivation

Three fundamental drivers motivate the strategy:

1. Revenue anticipation: Investors bid up Amazon shares ahead of Black Friday and Prime Day, anticipating massive sales figures
2. Sentiment premium: Positive media coverage and consumer excitement create favorable sentiment
3. Information asymmetry: Early demand signals (website traffic, pre-orders) may leak to informed traders before official announcements

The persistence of this anomaly over 27 years suggests structural factors — investor psychology and information diffusion patterns — rather than pure statistical noise.

Key Equations

Black Friday date:

Entry and exit windows:

Transaction cost model:

(20 bps round-trip total)

Hypothesis test (one-sample t-test on 33 event returns):

Kelly criterion position sizing:

Market filters (both must pass for entry):

Algorithm Blueprint

Holiday Effect Trading Strategy — AMZN

Inputs: AMZN daily prices, SPY index, VIX (1998-2025)

Outputs: Trade signals, daily equity curve, options P&L attribution

Algorithm:

1. EVENT CALENDAR CONSTRUCTION

   • For each year from 1998 to 2025:
     • Black Friday: 4th Thursday of November + 1 day
     • Prime Day: historical dates for 2015+, else July 15 (estimated)
   • Trade dates: entry = event date - 10 trading days; exit = event date - 1 trading day

2. ENTRY MARKET FILTERS (both conditions required for trade entry)

   • Trend filter: SPY close (bullish regime)
   • Volatility filter: VIX $< 25$ (low-volatility environment)
   • If NOT (trend_ok AND vol_ok): skip event entirely

3. EQUITY LONG EXECUTION

   • Entry: Buy AMZN at market open on entry date (100% portfolio capital)
   • Exit: Sell AMZN at market open on exit date
   • Costs: Round-trip transaction cost = 20 bps (10 bps per side)
   • Hold duration: 9 trading days (from entry to exit)

4. OPTIONS OVERLAY (parallel component, 2012-2025)

   • On entry date:
     • Sell out-of-money puts: strike (5% OTM)
     • Expiration: 7-14 days to maturity
     • Position size: max 5% portfolio capital at risk
   • On exit date or expiration: close or let expire
   • Provides income cushion and downside protection

5. RISK MANAGEMENT

   • Stop-loss level: -8% on equity position (hit = exit regardless of date)
   • Sentiment monitoring: flag if headline sentiment turns negative during hold period
   • Strategy suspension rule: if rolling 3-year Sharpe $< 0.2$, suspend trading

6. PERFORMANCE TRACKING (per-trade basis)

   • Record entry date, exit date, entry price, exit price, return, hold days
   • Compute rolling metrics: Sharpe ratio (3-year), win rate, maximum drawdown
   • Quarterly anomaly review: assess persistence and adapt thresholds if needed

Results

Equity Strategy (1998–2025):

| Metric | Value | |---|---| | Initial Capital | $1,000,000 | | Final Value | $2,872,486 | | Total Return | 187.25% | | Annualized Return | 3.85% | | Sharpe Ratio | 0.54 | | Maximum Drawdown | -14.26% | | Number of Trades | 33 | | Win Rate | 75.8% | | Avg Days per Trade | 9.3 | | SPY Benchmark Return | 1,044.52% |

Options Overlay (2012–2025):

| Metric | Value | |---|---| | Total Trades | 25 | | Win Rate | 92.0% | | Total Premium Collected | $7,917.64 | | Total P&L | $2,657.14 |

Subperiod robustness: 1998–2010 (discovery): Sharpe 0.61. 2011–2025 (validation): Sharpe 0.48. Modest decay consistent with gradual arbitrage of the anomaly. Window optimization confirms 10 days as optimal (Sharpe 0.54 vs 0.38 at 5 days, 0.42 at 15 days).

Filters eliminated 6 of 39 events (15.4%) due to bearish conditions or elevated volatility. Time in market: only 307 of 7,042 trading days (4.4%), explaining the underperformance vs buy-and-hold SPY in absolute terms while maintaining competitive Sharpe.

Contributions

• Statistical confirmation of pre-holiday momentum in Amazon at the 1% significance level (t=3.96, p=0.0002) over 27 years
• Quantitative window optimization demonstrating 10-day pre-event entry as optimal across 5/7/10/15-day alternatives
• Put-selling overlay achieving 92% win rate by monetizing the pre-holiday upward drift through options premium collection
• Subperiod validation (1998–2010 discovery vs 2011–2025 out-of-sample) confirming anomaly persistence with modest decay

Paper

Author

Frankline Misango Oyolo Quantitative Research Division, Arithmax Research Frankline@arithmax.com — Published: March 2, 2026