Friday, 27 August 2010

Wild market swings - an ideal testing day

Jerome and I have been thinking lots about stop losses. There is a great temptation to not bother with these as the systems generally trade quite happily without them. But then again - is it not wise to protect against some amazing occurance?

So perhaps we need a really wide stop for each market well beyond any likely move and perhaps hit only very occasionally. But what about things like trailing stops or tight stops based on indicators. After talking about this with Jerome this morning, we are provided with a fine set of test cases this afternoon. Economic data in the USA is not good at 1:30 and the markets slide. But then the Fed chairman comes out with some market-cheering comments and there is a huge rally. The losing positions that we could have been stopped-out of bounce back into profit, so the stops would have cost us money. Trailing stops might have worked, but I prefer things like profit target exits - e.g. where the open profit reaches, say, 100 pts. As we begin to trade above minimum size, perhaps a selection of these will be used?

Alternatively, we should remember that such moves are not a core part of our strategy and a small loss on such days is no big deal. Then we just revert back to normal trading at some future point soon after. This whole subject will be my main research task after I have handed in my dissertation next Wednesday.

Wild swings - the perfect test day for stop loss procedures

No comments: