Files

Abstract

Estimation of liquidity costs in futures markets is challenging because bid-ask spreads are usually not observed. Several estimators of liquidity costs exist that use transaction data, but there is little agreement on their relative accuracy and usefulness, and their performance has been questioned. We use a Bayesian method proposed by Hasbrouck which possesses conceptually desirable properties to estimate liquidity costs of six agricultural future contracts. The method builds on Roll's model and uses Markov Chain Monte Carlo estimation. Our Bayesian estimates are lower than more traditional estimates and as anticipated decrease even more when more realistic assumptions such as discreteness are incorporated. The findings demonstrate the need for further research to clarify the usefulness and accuracy of the procedure.

Details

PDF

Statistics

from
to
Export
Download Full History