- Joined
- 12 January 2008
- Posts
- 7,169
- Reactions
- 17,742
Another thought.
I understand that mean reversion systems work best when the mean does not move very much. Abnormal price swings will tend to revert to the mean. If the price swing is too big then it will move the mean in the direction of the swing. There is the distinct possibilty that the historical RIO/BHP relationship is changing and another significantly different valued ratio may be established. This trade will exit when enough time has elapsed for the ratio to catch the price spread. It could be that this trade will never be profitable and you may wait 6 mths or more to find this out.
I understand your decision to not use a technical stop loss, but I can't see the logic in letting a small $ loss become a bigger $ loss especially if there is evidence that the context (historical ratio) of the trade is invalidated.
This trade could be exited if the historical relationship deviates significantly from its norm. The major problem using something like this is the fact that a ratio is a lagging indicator and by the time the ratio signals that the original trade context has changed, price may be miles away and your losses huge.
I understand that mean reversion systems work best when the mean does not move very much. Abnormal price swings will tend to revert to the mean. If the price swing is too big then it will move the mean in the direction of the swing. There is the distinct possibilty that the historical RIO/BHP relationship is changing and another significantly different valued ratio may be established. This trade will exit when enough time has elapsed for the ratio to catch the price spread. It could be that this trade will never be profitable and you may wait 6 mths or more to find this out.
I understand your decision to not use a technical stop loss, but I can't see the logic in letting a small $ loss become a bigger $ loss especially if there is evidence that the context (historical ratio) of the trade is invalidated.
This trade could be exited if the historical relationship deviates significantly from its norm. The major problem using something like this is the fact that a ratio is a lagging indicator and by the time the ratio signals that the original trade context has changed, price may be miles away and your losses huge.