arabianights
Legendary member
- Messages
- 6,721
- Likes
- 1,380
You can't do a statistical test on a story (e.g. 'Central Bank is selling $XXXX').
Although let me give an exception which proves this rule. IBM was buying someone or other (from memory it was buying shares in Intel... or this may have been the other way round) and would buy puts at the same time to hedge its holding against price falls. IBM would go to the less liquid options market first and then buy the underlying. Renaissance Capital noticed that large put buying was almost immediately followed 30 seconds later by someone buying the share. So when they saw the put buying they would buy the share and hopefully sell it to big blue for a couple of cents profit in 30 seconds or so.
Now, you can easily do a test on the transaction pattern. But another possible explanation for the above would be a change in an OTC hedge from the options market to the underlying market (for example). Now, if you had convinced yourself that it was the latter I suspect there's a danger you'd be trying to come up with far more sophisticated strategies to take advantage of this... and you'd get it all horribly wrong because you were wrong in the first place.
All comes back to Rumsfeld, known knowns, known unknowns and unknown unknowns