Risk with Vol. Position Sizing
-
- Roundtable Knight
- Posts: 1436
- Joined: Mon Dec 22, 2003 12:24 pm
- Location: Des Moines, IA
- Contact:
Risk with Vol. Position Sizing
I think most of us here use some sort of volatility adjusted position sizing. Have you explored the implementation of any sort of procedure to limit leverage in the case of extremely low volatility?
-
- Roundtable Knight
- Posts: 2038
- Joined: Wed Oct 06, 2004 10:52 am
- Location: San Marcos, CA
Some vendor systems reduce position size when volatility is low AND they also reduce position size when volatility is high. Different people have different ways of calculating volatility, of course. And different ways of defining "low" and "high" too. However the underlying concept of trading smaller size when volatility is extremely high or extremely low, is not uncommon.
-
- Roundtable Knight
- Posts: 2071
- Joined: Fri Apr 25, 2003 3:33 pm
- Location: London
- Contact:
"Large trading hands" have sophistcated 24 hour trading desks which can minimise the impact of smaller volume markets. In addition to only trading a small percentage of the vol of course. Options for the smaller traders are more limited. 24 hour markets are in some ways a nuisance. So is trading in multiple world markets.
Lots to consider and (in my own case) lots which could do with improvement on the dealing front.
Lots to consider and (in my own case) lots which could do with improvement on the dealing front.
levi-
what i have tried in the past is placing a collar around the values of the short term atr, based on a multiple of the long term atr.
i.e the 20 period atr peaks out at x2 or x0.5 of the 200 period atr. this stops silly fluke trade setups giving you stats that are too awesome, and also limits the chance you will black swan on 5times typical size.
there are some other more complex effects on compounding growth etc and your mileage may vary with differenet collar parameters.
you may also find many markets don't need such a rule, but if you are trading front month STIRs instead of 1year to exp, then this kind of thing is a good idea.
what i have tried in the past is placing a collar around the values of the short term atr, based on a multiple of the long term atr.
i.e the 20 period atr peaks out at x2 or x0.5 of the 200 period atr. this stops silly fluke trade setups giving you stats that are too awesome, and also limits the chance you will black swan on 5times typical size.
there are some other more complex effects on compounding growth etc and your mileage may vary with differenet collar parameters.
you may also find many markets don't need such a rule, but if you are trading front month STIRs instead of 1year to exp, then this kind of thing is a good idea.