Comments
1) You believe casinos should be illegal due to societal harm caused.
Here’s an important factor: if you really believe this, you’ll probably do a bad job running a casino. You’ll probably be more successful if you try to run something else.
One might naively think that this isn't relevant to an 'EA' perspective, but I think it's a pretty straightforward application of an EA principle that the ends don't justify the means (https://www.lesswrong.com/posts/K9ZaZXDnL3SEmYZqB/ends-don-t-justify-means-among-humans).
In this hypothetical it sounds like you think the harm done will occur regardless, and so the tradeoff is between the additional benefits you can accrue from your ROI, VS the moral discomfort you will have in buying in? So there's both a moral uncertainty based on how strongly you personally think casinos should be illegal, how strongly you think the means justifies the ends (i.e. how important is it that you aren't the one responsible for the harm here, and an empirical Q of how much additional ROI you are actually getting compared to the next best option.
It also depends if this is something done in a personal capacity VS something that an EA organization is doing, because then there are other considerations that are harder to measure. E.g., even if you fully take the "ends justifies means" view, if CEA buys out a bunch of casinos, what negative effect does that have on movement building, and is that worth the marginal extra ROI compared to the next best investment?