The performance of bid-ask spread estimators under less than ideal conditions
The performance of bid-ask spread estimators is investigated using simulation experiments. All estimators are much more accurate if the data are sampled at high frequency. In high-frequency data, the Huang-Stoll estimator, which requires order flow information, generally outperforms Roll-type estimators based on price information only. The exception is when there is feedback trading (order flows respond to past price movements), when the Huang-Stoll estimator is seriously biased. When only low-frequency (e.g. daily) data are available, the Corwin-Schultz estimator based on daily high and low prices is usually less inaccurate than the Huang-Stoll and Roll estimators. An important and empirically relevant exception is when the spread varies within the day; in this case the Corwin-Schultz estimator significantly overestimates the true spread. For a published version, please see Studies in Economics and Finance, Vol. 32 (2015).
Year of publication: |
2013-09
|
---|---|
Authors: | Bleaney, Michael ; Li, Zhiyong |
Institutions: | School of Economics, University of Nottingham |
Saved in:
Saved in favorites
Similar items by person
-
Decomposing the bid-ask spread in multi-dealer markets
Bleaney, Michael, (2014)
-
Bleaney, Michael, (2014)
-
Excess Volatility and Closed-End Fund Discounts
Bleaney, Michael,
- More ...