The Abdi-Ranaldo Spread Estimator

Estimating a stock's bid-ask spread historically required intraday quote data — every posted bid and ask through the trading day — which is expensive, large, and often unavailable far back in history or for less liquid markets. The Abdi-Ranaldo estimator gets around this by using only three numbers per day that are almost universally available: the day's high, low, and closing price.

The intuition is that a stock's daily high tends to be recorded on an uptick trading at (or near) the ask, and its daily low on a downtick trading at (or near) the bid, so the high-low range over a couple of days carries information about the width of the bid-ask spread even without ever observing quotes directly. The estimator combines the high-low range with the closing price relative to that range across a two-day window, using a formula designed so that the pure volatility component washes out and what's left is attributable to the bid-ask bounce — prices mechanically bouncing between bid and ask as buy and sell orders alternate.

This makes it especially useful for historical liquidity research going back decades, or for markets and asset classes where intraday quote data was never captured, letting a researcher build a spread time series just from an ordinary daily OHLC dataset that almost every data vendor already provides.