how much leverage- who knows confidence intervals?

i was analyzing my track record in trading and trying to figure out what amount of leverage i should be using, so i started to do some statistics on my track record, but i was never so hot with confidence intervals.

what i'm wondering is if you could help me figure out X- what % of my account i should risk each day if i'd like a 95% chance of not losing 50% of my account over the next 2 years.

over the past 4 years i would have a track record of an average of +1.61X / month with a sd of 2.63 over the 48 months. each of the 4 years had the following numbers individually 1(.98X/2.10) 2(1.74X/1.99) 3(2.79X/2.24) 4(-.06X/2.77).

what is the equation for risk of ruin over N periods? all i could find were forever
 
Then think about it the other way: what X % should you risk so the probability of losing 50% of C in the next N years is <0.05?

The answer is known and it is 0%.
 
Quote from intradaybill:

interesting....how do you manage basis risk?

i trade value at risk - risk management and if you refer to the first post you'll see the question i'm asking.
 
Quote from dumb_mother:
i was analyzing my track record in trading and trying to figure out what amount of leverage i should be using, so i started to do some statistics on my track record, but i was never so hot with confidence intervals.

what i'm wondering is if you could help me figure out X- what % of my account i should risk each day if i'd like a 95% chance of not losing 50% of my account over the next 2 years.

over the past 4 years i would have a track record of an average of +1.61X / month with a sd of 2.63 over the 48 months. each of the 4 years had the following numbers individually 1(.98X/2.10) 2(1.74X/1.99) 3(2.79X/2.24) 4(-.06X/2.77).

what is the equation for risk of ruin over N periods? all i could find were forever
The meaning of X is a litle bit unclear, because it seems to mean different things in the two places above.
Does "+1.61X/month" mean +61% profit/month?
If you can give the average annual portfolio return (ie. profit) and its standard deviation then it should be calculable...
 
Back
Top