Market Stats
Stock market odds in plain numbers: seasonality, the presidential cycle, option expiration, Fed days and volatility, calculated from S&P 500 and VIX history.
Market statistics
Williams' VIX Fix: Formula, Worked Example and How It Tracks the VIXThe Williams VIX Fix rebuilds the VIX from price alone. Since 1990 it has a 0.75 correlation with the real VIX. Formula, worked example, spike stats.
When the VIX Surges 20% in a Day: What Came Next, 1990-2026The VIX has jumped 20%+ in one day 133 times since 1990. The S&P 500 fell on every one; the next day it rose 63% of the time. Stats by decade.
The Third Year of the Presidential Cycle: S&P 500 Returns Since 1928In year 3 of a US presidential term the S&P 500 averaged a 14.0% gain from 1928 to 2025 and rose in 19 of 24 years. Every year, the full table, and the caveats.
How the S&P 500 Does in December: Seasonality Since 1928The S&P 500 rose in 72% of Decembers from 1928 to 2025, averaging 1.27%, its most reliable month. Stats by decade, by half-month, and the Santa Claus rally.
January Option Expiration Week: S&P 500 Returns Since 1974January option expiration week rose in only 47% of years since 1974 (n=53) and in 36% since 1994, versus 58% for other expiration weeks. The data, by era.
FOMC Day Highs and Lows: How the S&P 500 Trades on Fed Announcement DaysOn 232 FOMC statement days since 1994, SPY's daily range averaged 1.57% vs 1.29% on other days, and it traded above the prior day's high 76% of the time.
Level of Bullishness: What Followed the S&P 500's Rarest Winning StreaksSince 1928 the S&P 500 has gone 25 sessions without two down closes in a row only 36 times. Here is what followed, plus up-close streaks and VIX lows.
SPY Holiday Effects: Returns Before and After US Market Holidays (1993-2026)SPY's session before a market holiday averaged 0.11% and rose 57% of the time (n=297, 1993-2026), about twice a normal day. The data by holiday.
Volatility Risk Premium (Part I): How Often the VIX Beats Realized VolatilitySince 1990 the VIX closed above the S&P 500's realized volatility over the next month on 86.8% of days, by 4.43 points on average. The data by year.
Volatility Risk Premium (Part II): Measuring It in Real Time, VIX Term Structure, XIV and VXXThe real-time volatility risk premium (VIX minus past realized vol) matched the true premium's sign on 80% of days since 1990, with a 0.09 correlation.
Volatility Risk Premium (Part III): The Risks of Selling VolatilityA stylized short-volatility trade won in 85.5% of months since 1990, but 64 losing months erased 36% of the gains. Drawdowns in 2008, 2018, 2020 and 2024.
Volatility Risk Premium Strategy: A Hypothetical Test, 1993-2026A hypothetical VRP rule (hold SPY only when VIX tops realized vol) returned 14.75% a year in 2015-2026 vs 13.20% buy-and-hold, before costs. Rules and caveats.
Fed POMO Days: Do Stocks Rise When the Fed Buys Treasuries?On 1,218 Fed Treasury purchase days in the QE era, SPY averaged 0.10% vs 0.12% on the other days of the same programs. The POMO-day edge doesn't show up.
September Triple Witching: The Week, the Day and the Week After (S&P 500 Since 1990)The S&P 500 fell in the week after September triple witching in 28 of 37 years (1990-2026, average -0.93%). Here's the week, the day and how it compares.
52-Week Highs in December: What the S&P 500 Did Next36 of 97 Decembers from 1929 to 2025 saw an S&P 500 52-week closing high. The next year averaged 8.59%, barely above all Decembers (8.00%). Full tables.