Traders do certainly not belong to the people you should listen to when it comes to trading.

Drawdown depth increases over time. You can make this yourself plausible by considering that the more you trade, the higher is the probability of a long loss streak. That's why a system, tested over 10 years, produces a higher max drawdown than the same system tested over 5 years.

When modeling drawdown mathematically, you'll find that with a break even system, the drawdown depth increases with the square root of the number of trades, which is in turn proportional to the trading time.

Drawdown also increases with the invested amount: When investing twice the lots you'll get twice the drawdown. When you reinvest a percentage of your profit, the drawdown depth increases with the profit, which is also proportional to the trading time.

When you sum up both effects, you'll get an overproportional drawdown increase with a time exponent of 1.5. The 1 comes from the reinvested profit, the 0.5 from the square root of the number of trades. The exponent 1.5 means that your drawdown grows faster than your profit and account balance. It is inevitable that at some point the drawdown will be higher than the account balance, causing a margin call.

That's in a nutshell why you should not reinvest a percentage of your profit, regardless whether the percentage is calculated with Kelly factors or - better - OptimalF factors.

The same, by the way, applies to a system of which you regularly withdraw all your profits. This method will also eventually cause a margin call with 100% certainty.