We generate new evidence on disagreement among traders
in the S&P 500 options market from high-frequency
intraday price and volume data.
Inference on disagreement is based on a model where investors observe
public information but agree to disagree on its interpretation;
disagreement among investors is captured by
the volume-volatility elasticity. For options, there are two
natural variables related to disagreement:
moneyness and tenor, which we relate to
disagreement about the distribution of the market index at different
quantiles and times. The estimated volume-volatility
elasticity equals unity for options near the money and close to
expiration, which is consistent with the case of no disagreement
among investors. In contrast the elasticity estimates decrease
with increases in the absolute value of moneyness, indicating
investors have a higher disagreement about rare events. Likewise
the elasticity decreases with increases in tenor, implying
in higher investors' disagreement about more distant events.