Ralph Vince worked close with Larry Williams. They started by using full Kelly to optimize their profits. In the beginning they were very successful, but soon enough they hit a losing streak, a huge drawdown, and lost a lot of their money and customers. Vince eventually came out with a new formula to use a fractional Kelly, and Larry's recommended percentages are based on Vince's later research.
I'm assuming you'd increase size to take advantage of geometric growth of return. If you risked say 1% to achieve 20% returns you're saying I could risk 2% and have 50% returns. The same level of reward:risk can be achieved by using reverse notional funding. All you'd do is figure the fully funded account size for the 2% risk per-trade then divide each trade by 1/2 to keep the risk at 1% and then achieve the reward of 25%. The two propositions are mathamatically equal so you would have to have a preference for a higher drawdown to increase size.
Here there is a table with the probabilities of drawdown between 50% and 80% when using full Kelly , 1/2, 1/3, 1/4, 1/5 and 1/6. It shows how probabilities go up for smaller fractionals.
On scalps and intraday position trades I use .5%-1% (that's 1/2 of 1% to 1%). Most are at the lower range. On longer term pair trades I will use as much as 2%-3%. I always try to set a stop (mental) at a logical point and adjust the share size to fit the total dollar stop. If I've learned one thing in trading, it's that you had better control your losses and you won't be around long.
Larry's figures 5%, 10% ... represent the total risk as percentage of his account: i.e. 5 positions 1% each, or 3 positions 2%+2%+1%.
I don't think Larry Williams has low risk percentages. Larry Williamsâ super performance in 1987 made me curious. So I loaded all his 1987â trades in an excel and came to the following information: Number of trades: 308 Winning trades: 179 (58%) for an amount of 5,991,704.35 $ Losing trades: 129 (42%) for an amount of 4,878,852.65 $ 5 draw downs of over â30% 2 draw downs of over â60% 1 draw downs of over â75% (maximum draw down was â 78.45%) Some remarks: More than 80% of his profits were lost again in his losing trades. The evolution of the returns were frightening, fast up, but with the same speed down again. Chart looks like a roller coaster, so not the ideal investment for someone with a heart disease. When he reached the first significant high around 888,000$ beginning of may, he needed another 4 months and 118 trades before I got above this 888,000$ level. End of august his equity was at 1,040,486.62$. So in the last 4 months he had to do over 100 trades to make an additional 10%. His trading style is completely different from what I had in mind of him. When I see the chart I have the impression he had difficulties to get consistent and stable performances. He had periods that he made money like crazy, but he had also periods that he lost money like crazy. But I must admit that in the end he made a huge amount of money, much more than I will ever be able to make.
Thanks! This is very interesting. I've read that at some point he was up more than $2 mils, and lost about 40% to end up still as the winner. Could you share Larry's trade list? I'd like to see how he was sizing his positions.