Let's simply assume the trading stats given here is reasonable and reliable.
In fact, It's a live, leveraged account, mainly for currencies, the $ amount has been proportionally modified, the stats is based on one currency pair for simplicity. However it shows the essence of the performance.
The holding period is a range from a few hours to a few days, some positions could be held for a few weeks. my trading stats shows a winning rate of 55% - 65%, the return could be avg 15% when it won, or avg -10% when it lost. And there is small chance (<10%) that it could lose 30% (due to unexpected events).
In the past, I always allocate a small % of available fund, based on 2/3 of classical Kelly, and further divide it 50/50 for each currency pair (I only trade two pairs). And I have a feeling that the money was somehow under utilized in most time.
That's why I'm looking for opinions on how people calc and use Kelly Criterion in their trading.
With some recent research, now, I'm OK with numerically optimized solution for Kelly Criterion, which means I could always find an optimal ratio using past trading stats.
Now, the problem is that the optimal value of Kelly Criterion given my trading stats is 4.6241, based on numerical optimization.
How could I interpret this ratio and apply it in real trading? I guess kut2k2 may suggest me borrowing more (3.6 times of my current), HAHAHA, but it's already a leveraged account...
Given your holding period and using 2 pairs I'd suggest around 3% (https://www.docdroid.net/ldv7Aiq/londontradersexporobertcarver.pdf.html#page=20)
This is far below what any Kelly formula will produce based on the numbers you have given. They are based on Kelly calculations using realistic performance figures that are closer to what most traders achieve.
All these arguments about the right Kelly are completely academic in the face of real trading records where we have a huge amount of uncertainty about what the real reward:risk profile is like. There is uncertainty about whether the historical performance is representative in a statistical sense (the less data, the more unreliable the track record).
This at least is a known known; for example after 10 years of trading a trader with a sharpe ratio of 0.5 will have a 95% confidence that their true sharpe ratio (of which their track record is just a random sample) is between -0.1 and 1.1. The relevant Kelly criteria figures are 0% and 110%.
There are also unknown unknowns: There is uncertainty about the future repeating like the past. With a human discretionary trader there is further uncertainty about performance repeatability.
The problem is that the numbers you have given translate into very high reward:risk figures and thus a very high Kelly criteria. I personally think it's unrealistic and irresponsible to use such high figures. I'm not suggesting you are fudging the numbers, merely that you don't have an appreciation of the uncertainty involved in extrapolating into the future.
GAT