leverage is the most-discussed and least-understood variable in perp trading. almost every blow-up story is somebody who held a wrong mental model about what leverage actually does. the worst part is that the wrong models sound smart. they get repeated on twitter and discord constantly. they survive contact with thousands of traders before anyone bothers to write down why they are wrong.
we built uponly, an arcade-style perp product where defaults sit between 75x and 500x. we have a more concentrated view of what happens to traders across the leverage band than most. these are the five myths that cost the most money. each one comes with the correct mental model.
1. "higher leverage means higher risk"
this is the myth that does the most damage because it is right enough to feel true and wrong enough to make people size catastrophically. the correct version is: higher leverage on the same collateral means a tighter liquidation band. higher leverage on a much smaller collateral can mean a smaller worst-case loss in absolute dollars.
compare two trades. trade a is a one thousand dollar position at 5x, collateral two hundred dollars. trade b is a one hundred dollar position at 50x, collateral two dollars. the multiplier is ten times higher on trade b. the worst-case dollar loss is one hundred times smaller. risk is about dollars at risk, not about a number on a slider.
2. "low leverage will save you in a crash"
people who say this are usually trying to dunk on degens at 100x. the irony is that low leverage on a huge collateral is one of the most common ways serious traders blow themselves up. if your collateral is your entire stack and a black swan moves the price five percent, you are down a meaningful chunk of your net worth even at 2x leverage. the multiplier is not what saves you. responsible sizing is what saves you. those are different things that get conflated.
3. "you can outsmart liquidation with stop losses"
stop losses are a useful tool. they are not a magic shield. in a violent move, a stop loss can slip past your set price and execute at a much worse fill, or fail to execute entirely if the exchange is overloaded. on a perp dex with sparse liquidity at the moment of the stress event, your stop may functionally not exist. that is not a hypothetical. it happens during every cascade. if your trade only works because your stop loss is going to be perfect, you do not have a trade. you have a wish.
the correct mental model: assume your stop will execute at the worst plausible fill. if the trade still makes sense at that fill, it is a real trade. if it does not, you need to either lower size or accept that the stop is not the actual risk control. for more on how to size against this reality, read max leverage trading 101.
4. "max leverage is for gamblers, not real traders"
this myth is almost always told by people who use leverage themselves, just at a band that lets them feel sophisticated about it. the reality is that there is no platonic "responsible" leverage level. there is only "sized correctly for your goal." somebody using 250x leverage on a five dollar collateral as a fifteen-second arcade trade is making a completely different bet than somebody using 5x leverage on a fifty thousand dollar collateral as a swing trade. neither is automatically a gambler. neither is automatically responsible. the framing collapses the wrong axis.
the correct frame: ask what you are actually trying to do with the trade. directional thesis with multi-day horizon and big collateral, that is one game. entertainment-sized rip with a five-minute horizon, that is a completely different game. they use different leverage profiles for the same reason a sprinter and a marathoner train differently.
5. "the platform sets your max leverage based on safety"
this is the most subtle myth on the list. traders assume that when a perp dex shows a maximum leverage of 50x or 100x, that number was chosen because anything higher would be "unsafe." that is occasionally true, but mostly the number reflects the platform's own risk model for its own pool, not yours. on a peer-to-pool perp dex, the cap protects the pool. on an order-book dex, it protects the matching engine. on a wrapper product, it usually reflects the underlying protocol's parameters.
the implication: do not interpret a platform's leverage cap as advice. interpret it as a fact about the platform. some platforms cap at 50x because their pool needs that buffer. some, like uponly routing to avantis on base, expose the protocol's native band of 75x to 500x because that is what the underlying market supports. neither is a recommendation about how much leverage you should personally use.
the unified mental model
if you only take one thing from this article, take this. risk is a dollar amount. leverage is a multiplier on a thesis. the dollar amount is set by you before you open the trade. the multiplier is set after, based on how tight a band you want around the entry. when you do it in that order, every leverage myth collapses, because you are no longer treating the slider as the risk control. you are treating the collateral as the risk control, which it actually is.
- pick the dollars you would be okay losing this session
- put that exact dollar amount as collateral, no more
- pick leverage based on how much price wiggle room you want
- open the trade
- forget the multiplier exists. focus only on whether the dollar outcome is acceptable.
where to test the model
the easiest way to internalize this is to do twenty rips with a tiny fixed collateral, varying only the leverage. you will feel in your gut what 75x versus 500x means when the collateral is constant. that is the lesson. when you are ready to run that experiment, try uponly with a budget you do not need. the structure where losing trades have zero fee makes it the cheapest possible classroom for this lesson.
and if you want a longer treatment of how to think about sizing under high leverage, the canonical writeup is in max leverage trading 101. everything we said here is consistent with that. they are the same model, told from two angles.