This thread caught my attention, but now I feel baited like much of the bad click-bait articles that overwhelm social media. My question is simply this: Where is this “undisputed proof”?
I was unable to replicate this experiment/example in a way that I felt 2% risk was undoubtedly the best way to go from a losing trader to a winning trader. Allow me to elaborate.
In my opinion, and to keep it simple, there are 2 ways to be unprofitable: either by not having an edge, or by having an edge and using poor money management. The example given in the video seems to be focused on the second, so I did so as well. The video showed profit/losses in arbitrary numbers; risk is risk and in both winning and losing scenarios, the amount doesn’t matter – if it’s negative it’s an x% loss and if it’s positive it’s an x% gain. Therefore when I tested this I just used a binary generator, 1s and 0s, to denote wins and losses.
The “Fixed” tab shows a simulation of fixed risk, and even win rate (10 wins, 10 losses). The “Varied” tab shows a simulation of a random risk generator between 1 and 3. The “Fixed winning” tab shows the same simulation, but with 11 wins and 9 losses instead. The “Varied winning” tab shows this simulation with a varied risk between 1 and 3 as well. The results tab shows two variations of varying risk parameters; 1 for risk between 1-3% (same as fixed 2% in the long run but oscillating) and one for 1.5-4% (averaging higher than 2%). The numbers below show simulations of end equity at the end of 20 trades with 11 wins and 9 losses. Roughly speaking, it seems that using a fixed risk may be better than using a varied risk that averages out to the fixed rate (however, in reality no one uses a random risk generator to determine their risk). In over 1000 trials, the varied approach performs worse than using a simple fixed approach by a margin of about 3% which is negligible in my opinion. However the real evidence for my case lies in the second number, which states the following:
Using a varying risk percentage between 1% and 3% will result in the trader losing money roughly 10% of the time.
An arbitrary buffer of 5% seems adequate, which would lead to the following statement:
Using a fixed risk rate of 2%, rather than a varying risk rate of 1%-3% will change a trader from a losing trader to a winning trader about 5-15% of the time.
The second varied simulation on the “Results” tab really is just there to show that when there is a real edge (win rate>50% & R:R=1), risking more is better.
It seems to me that if you have a legitimate edge, using a fixed risk really doesn’t do that much for you in the long run. IMO DutchTrader1 is spot on; there is an edge in this specific example because the average risk in the losing positions was greater than the average risk in winning positions.
I look forward to learning more about this proof.
Undisputed.zip (117 KB)