Asset Pricing in the Presence of Market Microstructure Noise

Battle Count: 84

Highly relevant for quantitative traders involved in options market-making, volatility trading, and derivative pricing. The paper provides a framework for separating microstructure noise from fundamental price dynamics, which is critical for accurate option pricing and hedging. The implied epsilon surfaces and calibrated noise parameters offer practical tools for adjusting BSM pricing. The binomial tree approach preserving drift information is particularly useful for discrete-time trading strategies. However, the model's practical implementation requires careful calibration and may be most relevant for index options and liquid markets.

The mathematical framework involves advanced stochastic calculus (Tanaka's formula, local time, Feynman-Kac solutions). The binomial tree calibration requires solving constrained optimization problems over multiple parameters simultaneously. The ARMA-GARCH fitting adds computational complexity. However, the final pricing formulas (modified BSM with sigma+epsilon) are straightforward to implement. The main challenge lies in the calibration procedure requiring option market data and the joint estimation of drift and volatility noise components.

The paper provides detailed mathematical formulations, calibration procedures, and parameter tables. However, the empirical data (S&P 500 option chain from Cboe, US Treasury rates) is referenced but not provided as a downloadable dataset. The ARMA(3,3)-GARCH(1,1) specification is fully described. The binomial tree calibration method is detailed with explicit formulas. Reproduction would require access to the same Cboe option data for April 21, 2025 and the historical return series.