Abstract
This paper examines the statistical and economic significance of short-term autocorrelation in Australian equities. We document large negative first-order autocorrelation in individual stock returns. Preliminary results suggest this autocorrelation is economically significant, as two simple trading strategies based on the autocorrelation structure appear to yield large risk-adjusted returns. Further analysis, however, shows that these results are driven by the inclusion of smallcapitalisation and low-priced stocks which are vulnerable to a number of market-microstructure-related problems. After revising the dataset to mitigate these problems, little evidence of economic significance remains.
| Original language | English |
|---|---|
| Pages (from-to) | 97-117 |
| Number of pages | 21 |
| Journal | Australian Journal of Management |
| Volume | 28 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 1 Jan 2003 |
Keywords
- AUTOCORRELATION
- ECONOMIC SIGNIFICANCE
- MARKET EFFICIENCY
- RANDOM WALK HYPOTHESIS
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver