Garman-Klass Volatility

Garman-Klass volatility was introduced by Garman and Klass (1980) to exploit the full range of intraday price information available in standard OHLC data. Unlike realized volatility estimators that require tick-by-tick data, Garman-Klass provides a closed-form unbiased estimator that leverages the relationship between high-low, high-close, and low-close price ranges to isolate true volatility from measurement error and bid-ask bounce. This makes it particularly valuable in quant scoring systems where computational efficiency and data availability constraints favor daily-frequency OHLC inputs over microsecond-level order book reconstruction.

In insider trading and market abuse detection platforms, Garman-Klass volatility serves as a robust denominator for conviction scoring and signal scaling, since it captures true asset volatility without contamination from transient liquidity shocks. When abnormal insider activity concentration occurs during low-volatility regimes, the signal-to-noise ratio of detection metrics improves significantly. The estimator's insensitivity to overnight gaps and opening auctions also reduces false positives in surveillance systems monitoring pre-announcement trading patterns.