How to handle this in practice?

No rebalancing at all also seems nonsensical because the whole point is to buy more when something is supposedly cheap. If it's price rises it gets less cheap if everything else is unchanged so at some point there should be some rebalancing by selling something not that cheap anymore for something else that's supposedly cheaper now. But when to rebalance? How often?

I like this line of reasoning and like to think I do something similar, and of course it is also quite close to the rebalancing that Manlobbi recommends, investing more heavily in stocks that have a higher multiple to the minimal plausible (plausible in your opinion) intrinsic value in 10 years. discussion.fool.com - Welcome revisited

Of course you have better things to do than rebalance by a tiny amount each day, even if transaction costs and tax costs were minimal. But if you DID do that, it wouldn't matter - yes, you would be selling tiny slivers of your holdings in something that had gone up one day, and presumably buying them all back when the share price came back down. But it would give you very similar results. If you wanted to convince yourself of this, try setting up a spreadsheet for a stock that is trading, say, at 5 times your assessment of the IV in 10 years, and which you would sell entirely if it were to go down to 3 times. Now compare 2 strategies, where in one of them you sell 5% every time the Price:IV10 goes down by 5% of the distance between 5 and 3, i.e. by 0.1, from 5.0 to 4.9 for example, and from 4.9 to 4.8. And for the other strategy, you sell 20% every time the Price:IV10 goes down 20% from 5, i.e. only at 4.6, 4.2, etc. from its original value. I've not yet succeeded in getting Excel to simulate this for me, but I'm pretty sure both strategies will give me the same gain. The more active one will register smaller gains often; the less active one will register big gains less frequently.

For instance, going from a Price:IV10 of 5 to 4, the active strategy will give me a small gain at 4.9 (as the price is higher), another small gain at 4.8, again at 4.7, etc.. Once it gets to 4.6, it is true that you would have obtained a larger gain by waiting, getting the full benefit of that price change instead of only most of it. But the price may not go down directly to 4.6, either; it may bounce from 4.8 back up to 5.0, then to 5.3, then back to 4.9, then back to 5.1, ... Here, the fine-toothed strategy is giving you little gains all the time, which I expect will fully compensate you for the fact that you have been reducing your stake a bit after the small price gains.

There's probably a mathematical way of proving (or, I suppose, disproving) this, but I can't think of it right now. That's not to say that it is a good idea to be making all these little transactions with small price movements; you are mostly wasting time and effort. I'm just saying that it shouldn't actually hurt your performance.

dtb