Very thoughtful article. I think your 4 fold guardrail is very practical when we apply the Kelly Criterion to stock investing where alot of the inputs are subjective guesstimates. Just curious, on point 3, how did you derive the minimal probability to be 70%?
There’s no shortage of investment ideas on Substack, but less discussion on how to actually construct a portfolio. In my view, the Kelly Formula is a very useful tool for this, and you’ve done a great job outlining its benefits, pitfalls, and practical application. I especially appreciated the simulations showing how different bet sizes compound over time.
P.S. Your point—"Treat the Kelly criterion as a thought experiment rather than a mechanical exercise."—is spot on. Since the odds in investing can’t be precisely known but only estimated, using Kelly as a flexible framework—not a rigid formula—is the right approach.
Great article. Super well structured and such an easy read. Hard to imagine levering up 10x if your max loss is 10% - but the thought does give me comfort in levering up on things like defined-outcome ETFs and the like, but then fees, return, blah...
Very thoughtful article. I think your 4 fold guardrail is very practical when we apply the Kelly Criterion to stock investing where alot of the inputs are subjective guesstimates. Just curious, on point 3, how did you derive the minimal probability to be 70%?
Thanks for reading. The 70% is about sticking to your best ideas.
There’s no shortage of investment ideas on Substack, but less discussion on how to actually construct a portfolio. In my view, the Kelly Formula is a very useful tool for this, and you’ve done a great job outlining its benefits, pitfalls, and practical application. I especially appreciated the simulations showing how different bet sizes compound over time.
P.S. Your point—"Treat the Kelly criterion as a thought experiment rather than a mechanical exercise."—is spot on. Since the odds in investing can’t be precisely known but only estimated, using Kelly as a flexible framework—not a rigid formula—is the right approach.
Thanks a lot, Kenny!
Great article. Super well structured and such an easy read. Hard to imagine levering up 10x if your max loss is 10% - but the thought does give me comfort in levering up on things like defined-outcome ETFs and the like, but then fees, return, blah...
Interesting stuff. Thanks for sharing!
Thanks for reading!