A popular 0DTE rule is the percent stop — “cut it if the option loses 50% of its value.” It sounds disciplined, but it's a worse framework than a dollar stop sized from your risk budget, for two reasons.
A percent stop takes random risk
50% of a $0.40 option is $20 per contract; 50% of a $2.00 option is $100 per contract. Same “50% stop,” 5× the dollar risk — because the risk floats with the premium you happened to pay. And it's disconnected from your trade thesis: a 50% option loss might correspond to a trivial SPY move on a cheap far-OTM contract or a large one on a near-the-money contract. The percent isn't tied to where you're actually wrong.
A dollar stop is consistent and thesis-based
A dollar stop starts from the SPY level where your idea is invalid, translates it to a premium via delta, and sizes so that loss equals your fixed dollar risk. Now the stop is where the trade is actually wrong, and the loss is the amount you decided to risk — both anchored to real things (the structure and your account), not to a percent of an arbitrary premium.
A percent stop asks “how much of the option am I willing to lose?” A dollar stop asks “where is my idea wrong, and what will I risk on it?” Only the second is risk management.
When percent-thinking sneaks back in
There's one honest use: a hard cap. Even with a dollar/level stop, some traders add “also never let any single option lose more than X%” as a backstop against a gap. That's fine as a secondary rail. But the primary stop should be a dollar amount tied to a level and your risk budget — which is exactly how NoVo sizes and places it: a level-based, dollar-risked, protective stop, not a naive percent of premium.