Pricing Model
Comparing real option prices to Black-Scholes and SABR theoretical values across 11 underlyings using live chain snapshots taken every 5 minutes during market hours. The residual (market price minus theoretical) shows where and by how much the models misprice. Where the residuals converge across underlyings, the mispricing is structural to the model itself. Where they diverge, it is specific to the underlying's volatility dynamics. Full writeup: Measuring What the Pricing Models Get Wrong.
Collecting chain snapshots every 5 min since Jun 28, 2026
Residual by Underlying
Mean BS mispricing per symbol. If these converge, BS has a structural bias. If they spread, the mispricing depends on the underlying.
Residual by Days to Expiration
How much the market price deviates from Black-Scholes at each DTE. Positive means the market prices higher than theoretical.
Residual by Moneyness
Strike / underlying price. Below 1.0 is in-the-money calls (OTM puts), above 1.0 is OTM calls (ITM puts). Skew shows up here.
Residual Distribution
Distribution of (market price - theoretical price) across all observations.
Daily Mean Residual
Average mispricing per day. Tracks whether the gap is stable or drifting.
About This Model
Data source: brokerage options chain API, snapshotted every 5 minutes during market hours for 11 symbols (SPX, SPY, QQQ, IWM, AAPL, TSLA, NVDA, AMZN, MSFT, META, GOOGL). Collection started June 28, 2026.
The theoretical value is the broker-provided Black-Scholes calculation. The market price is the mark (midpoint of bid/ask). SABR is calibrated fresh from the chain data using the Hagan et al. (2002) approximation. Observations with mark or theoretical below $0.10 are excluded.
The fidelity of standard pricing models is something most retail traders take on faith. This page measures it directly against millions of real market observations. Where the models agree with the market, synthetic backtesting is trustworthy. Where they diverge, the backtest results are lying to you by exactly the amount shown.
This model updates daily. As data accumulates, the patterns in mispricing by DTE, moneyness, and volatility regime will sharpen. The goal is a correction model that makes the backtester's synthetic pricing closer to what you'd actually trade at.