I'd like to chip in a few things. Though the adjusted fractional system is the basic premise for cash management, the best way to implement it may differ from system to system, also that I feel is dependent very much on the statistical performance of that system. For instance, if you're using an agressive trend following system, that seems for accurate entries into large trends, and aggressively adds to winning transactions you are very likely to have a low probability for entry, (lets say 1 in 8) but because of the compounding effects of winners your average risk reward is an amazing 1:10. Should you use 2 percent total equity on each entry, you will have one debilitating drawdown before that significant winner comes in. And if you miss it? Whereas if you have some sound data regarding the average win dimensions in pips, it is possible to split that 2% equity depending on your scale schedule, and reduce which drawdown significantly. On the flip side, in case you've got a trading system with which indicates are 80% accurate, but come state two times a year, with a good risk reward (based on sound performance data) why risk only 2%, then you can fairly safely utilize 5 percent

On the flip side, lets choose the position of a daytrading scalper, that chooses between 2-8 pip hits but with a 60% strike rate and a risk benefit varying from 1:2 - 1:10. Might it be wise for him to risk two percent equity per transaction? Personally I would say not, since he probably makes 5-10 trades a day, and drawdowns are very likely to happen on a weekly basis. Its easier dividing that two percentage equity over the transactions for the day, realizing that with a good system in place at minimal that he can make a 1 percent return on equity most days, which is 1300 percent annuallised. Why risk two percent per transaction in this circumstance?

Of course all this is assuming you've got a tested winning methodology with strong statistics behind it. Otherwise, the reduction of capital will prevent.

Just my two pips.