Indeed, like driving at 120 mph through NY. You will be the fastest...... or you will be death. Especially if you have to do it several times in a row.
Profitable traders how much do you risk per postion?
- Thread starter Daal
- Start date
Indeed, like driving at 120 mph through NY. You will be the fastest...... or you will be death. Especially if you have to do it several times in a row.
Risk/position should be a function of relative edge per trade. Increase size into a perceived increase in edge; as exhibited in betting and strategy-variation with card-counters using an advanced point count.
In the option market it's more transparent than in share or futures trading due to the addition of a value-parm, implied vols.
In the option market it's more transparent than in share or futures trading due to the addition of a value-parm, implied vols.
I use a fixed percentage risk sizing, based on a fraction (1/5) of the Kelly ratio calculated on my past results / backtesting of my trading method. I trade 2-3 strategies, and I'm shooting to improve my winning percentages as a way of reducing my drawdowns. I'm trading mostly strait options or front/ back month vertical spreads.
Using Larry's competition statistics, that lead to a Kelly ratio of about 10%, he could've increased his account 5 fold after his 308 trades, which would still be a great performance.
Using Larry's competition statistics, that lead to a Kelly ratio of about 10%, he could've increased his account 5 fold after his 308 trades, which would still be a great performance.
Poor Larry. He obviously must not have known any better. In case he did, he perhaps should not have relied on what you recognized to be in fact "variable" MM. For sure, he would have loved to benefit from your insights. In the meantime, a pithy he had, AFAIK, to switch to a sightly different line of business.
FYI, nononsense makes use of statistical pattern recognition techniques to trade commodities. I wouldn't know how to handle multiple strategies - sounds pretty confusing and distracting. During continued development, exhaustive study includes keeping the perceived risk of ruin below adequate levels. As I found out, this never could have been achieved by naive tack-on of probabilistic model gimmickry borrowed from the theory of games of chance.
I don't have the faintest idea of how I could have helped Larry with his predicament, if any. My approach appears to be different.
nononsense
Bud,
Tell us about "variable" MM.
Stay on topic.
Using spike500's Excel file (thanks! 
):
):- trades = 308
wins = 179
average win = 8.2%
max win = 130.9%
losses = 129
average loss = 5.8%
max loss = 44.8%
avg win / avg loss = 1.41
Kelly ratio = 28.4% (LW may have used a fixed 20-30% risk after all)
max drawdown = 78.45%
account:
. initial = 10,000
. final = 1,147,607
. max during contest = 2,048,643
