Whatever you do, ppnewbie, don't try this scheme with options. The trading cost, slippage, spread and blowing by your limits would destroy any chance at all of coming out ahead. It's a huge mistake to not count those into your modeling program.Tortoise wrote: ↑Wed Apr 01, 2020 1:20 pmNot to discourage learning and experimentation with using the Kelly criterion in options trading, but aren’t any expected gains already baked into the option prices (on average)? Otherwise it seems like hedge funds and big quant groups could swoop in and make a killing using very simple math.
If I were going to give this a shot, I would use some highly liquid, highly volatile ETF that you could trade for free. Maybe one of KBG's triple leveraged Nasdaq ETF's or something like that. One more issue that you would need to factor in: The distribution of wins/losses are guaranteed not to be gaussian. So don't use a random number generator for your simulation. You would be surprised how quickly you can go broke if you assume a normal distribution. Model this with historical tick by tick prices if you can get them.
Should be a fun experiment. If you're careful, it won't be completely stupid and you might do okay.
Keep us updated if you jump in.