I was reading Perry Kaufman's Smarter Trading and thinking about his chapters on price shocks.
A few weeks ago I was in Orange Juice when it dropped 5 cents. In about 5 minutes. It dropped 12 cents in two days. I was also in Live Cattle a month or so ago when I foolishly held onto a position over the weekend and there was a mad cow scare in the US.
So I've been involved in two price shocks albeit run-of-the-mill.
The theory is that price shocks are more common than people think. And that they are not accounted for in many systems. Systems do not handle them well, and people do not backtest for them adequately.
And then there are the black swans which are the mega price shocks such as October 1987 or summer/fall of 1998, where things happen that are not "supposed" to happen.
How do you handle these in your system, and what has happened to you in the past that has made you adjust your system from price shocks?
how your system handles price shocks, outliers, black swans?
By definition, I don't think there is anything in particular that can guard against a true price shock, black swan type of event. Stops will help in an orderly market, but a true shock will likely gap beyond your stop. Then you need to have thought about how to handle this scenario...out at the open/first chance, wait for some order to restore itself and then get out, it's an individual choice.
Back testing over a sufficient length of time should incorporate a number of these events, and thus in a sense, you will have accounted for these in your testing. Look back at these and see how your system would have handled them and then see if there is anything that would improve on it.
These types of events are in a sense the cost of doing business. There is always that risk that can't be fully accounted for. This is where risk % of capital, money management, etc. comes into play as another way to help guard against these events.
Nothing that you already don't know I'm sure,
J
Back testing over a sufficient length of time should incorporate a number of these events, and thus in a sense, you will have accounted for these in your testing. Look back at these and see how your system would have handled them and then see if there is anything that would improve on it.
These types of events are in a sense the cost of doing business. There is always that risk that can't be fully accounted for. This is where risk % of capital, money management, etc. comes into play as another way to help guard against these events.
Nothing that you already don't know I'm sure,
J