Stats › Intraday and market microstructure › Marking the close
About one per cent of closing prices are manipulated and roughly one episode in 250 is prosecuted
| Quantity | Figure | Basis |
|---|---|---|
| Estimated prevalence, specification one | 1.14% of stock days | Detection model applied out of sample |
| Estimated prevalence, specification two | 1.08% of stock days | Same |
| Prosecuted manipulation | 0.004% of stock days | Hand collected cases |
| Unprosecuted episodes per prosecuted one | About 308 to 326 | The authors' own arithmetic |
| Model discrimination | AUROC 0.902, 95% interval 0.870 to 0.934 | Out of sample |
Built from 184 manipulated stock days drawn from eight independent legal cases across 31 stocks and four exchanges, against a comparison sample of 1,249,748 stock days. Cases were collected for conduct between 1997 and 2009 and the prosecuted cases used run from 1998 to 2005.
A detection model fitted on prosecuted cases and applied to 1.25 million stock days estimates that 1.08 to 1.14 per cent of closing prices are manipulated. Prosecuted cases account for 0.004 per cent of stock days.
Members only
The rest of this page is for members
Below this point there are 4 sections and 1 named source, roughly 400 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.