I still remember the first time a winning long turned into a scratch trade because my stop sat still while price ran away and then snapped back. Fixed stops do that. They protect you, sure, but they also freeze the exit at a number that stops making sense the moment the market cooperates. That is the quiet problem Hyperliquid just tried to fix by launching trailing stops across its perpetual markets.
The idea is not new on traditional desks. What matters here is the timing. Perpetual volume on the platform has been loud for months, more markets keep landing through HIP 3, and traders have been asking for an order type that can walk with the mark price instead of sitting like a brick. Now they have it. A trailing stop follows the best price after activation, then fires a market order when price retraces by a distance or a percentage you choose.
How Hyperliquid Trailing Stops Actually Work
Let me put this in plain language before the jargon takes over. You open a position. You decide how far price is allowed to fade from its best print after tracking starts. If the market keeps moving in your favor, the trigger moves with it. If the market gives back that cushion, the order becomes a market sell for a long, or a market buy for a short.
For a long position, tracking follows the highest mark price reached after activation. For a short position, it follows the lowest mark. That split is the whole mechanic. Everything else is settings: distance versus percentage, optional activation price, size, and the usual reminder that a trigger is not a guaranteed fill at the trigger.
A trailing stop’s trigger price follows the mark price as it moves in favor of the position. When the mark price retraces from its best level by the selected distance or percentage, it triggers a market order.
That last sentence is doing a lot of work. The platform is explicit about using mark price, not last trade, not some mid you invent in your head. If you already use take profit and stop loss on the same venue, this will feel familiar. Those tools already lean on mark price for the trigger. Trailing stops extend that logic instead of inventing a second universe of prices.
Activation Price Versus Immediate Tracking
You can leave the activation field empty. Tracking starts right away from the current mark. Or you can set an activation price so the trail does not even wake up until the market reaches a level you care about. I like that option more than I expected. It keeps a trailing stop from chewing itself up in noisy chop before the trade has even earned the right to be protected.
Picture a long that you only want to trail once price clears a range high. Until that print, you might still want a conventional stop underneath the range. After that print, the trail can take over. The two tools are not rivals. They are sequential. Plenty of traders will still treat the trailing stop as a standalone exit. That is fine. Just know you are choosing a personality for the trade.
- Empty activation field: tracking begins immediately from the current mark price
- Set activation price: tracking waits until mark reaches that level
- Longs trail the highest mark after tracking starts
- Shorts trail the lowest mark after tracking starts
- Retracement can be a fixed distance or a percentage
- Trigger sends a market order for the selected quantity
Distance, Percentage, And Why The Choice Matters
A fixed distance is honest in quiet markets and clumsy in violent ones. A percentage scales with the asset. On a high beta perp, a two percent trail can feel tight. On a slower pair, two percent can feel sloppy. There is no magic number. I have watched people copy a one percent trail from a liquid major onto a thinner market and then act shocked when they get wicked out.
Think in volatility, not in superstition. If the typical noise around your entry is already larger than your trail, you do not have a stop. You have a coin flip with extra steps. Widen the trail, cut size, or skip the product. Those are the adult choices. The order type will not save a thesis that needed more room than the account can fund.
Trigger Price Is Not Fill Price
This is the part that still trips people up, even after years of using conditional orders. The mark price can satisfy the condition. The book can still move. Liquidity can thin out. Existing market take profit and stop loss orders on the venue already carry a stated slippage tolerance on the market versions, while limit versions let you pin a price. Trailing stops, as described at launch, fire as market orders. That means you should assume some gap between the moment of truth and the print you actually get.
In my experience, the traders who complain the loudest about slippage are often the same traders who size as if fills were mathematical constants. They are not. A trailing stop is a risk tool, not a rebate program. If the market is cascading, your trail did its job by getting you out of the dream. It did not promise a fairy-tale price.
Why This Order Type Showed Up Now
Feature drops rarely happen in a vacuum. Perpetual activity on the platform has stayed elevated. Integrations have multiplied. Independent teams have been deploying markets under HIP 3. When more names, more structures, and more participants show up, the demand for better exits rises with them. You can day-trade a handful of majors with blunt tools. You cannot run a wider book that way without turning risk management into a second job.
HIP 3 is the quieter story underneath the headline. Outside teams can deploy and operate perpetual markets using the same infrastructure. A recent testnet path even pointed toward permissioned markets with onchain allowlists, optional rather than mandatory. That sounds dry until you realize it changes who gets to list what. More listings mean more idiosyncratic volatility. More idiosyncratic volatility is exactly when a static stop starts to feel antique.
There has also been a steady drumbeat of attempts to wrap selected perpetual exposure in structures that regulated venues can touch. Energy-linked contracts have been part of that conversation. So have plans involving established firms that want a cleaner path for eligible clients. I will not pretend those processes are simple or guaranteed. I will say this: product design on a decentralized matching engine and product design for a supervised wrapper are different sports. Trailing stops help the first sport today. They do not settle the second.
Where Trailing Stops Sit In The Existing Toolkit
The venue already offered market, limit, stop market, stop limit, take market, take limit, scale, and TWAP. TWAP still chops a large order into smaller tickets on a timer, with slippage caps on the slices. Scale still ladders entries. Conditional exits already existed. Trailing stops fill the gap between “I know my invalidation” and “I want invalidation to travel with the trend.”
| Order style | Best use | Main weakness |
| Fixed stop market | Hard invalidation under structure | Does not harvest trend |
| Stop limit | Control worst fill | Can miss the exit in a gap |
| Take profit | Preplanned target | Caps winners early |
| Trailing stop | Let winners breathe, lock progress | Whipsaw in chop, market fill risk |
| TWAP | Work size without dumping the book | Slow if the thesis breaks fast |
If that table looks obvious, good. Most trading pain is not a missing tool. It is using the right tool in the wrong regime. A trailing stop in a range is a machine for paying the spread. A trailing stop in a clean impulse is a way to stay in the move without babysitting every candle.
A Long Example Without The Fantasy
Say you are long. Mark is 100. You start tracking immediately with a 3 percent trail. Mark rips to 110. The trigger now lives near 106.7, give or take how the venue computes the exact offset. Mark tags 118. The trigger ratchets again. Then mark slumps to 114. You are still in. Mark keeps sliding and kisses the trail. The market order goes out.
Did you sell the top? No. That was never the job. You sold a pullback from the best mark after the position had already paid you. That is the entire point. People who want the exact high should write poetry, not tickets.
Now flip it. Short from 100 with the same 3 percent idea. Mark prints 92, then 88. The trail follows the low. A bounce that is still “only a bounce” leaves you in. A bounce that retraces the full cushion takes you off. Same logic. Opposite direction. If that still feels abstract, paper it for a week. Watching the trail move is more educational than another thread explaining it.
The Subtle Ways Traders Will Misuse This
I have found that new order types attract two crowds. The first crowd wanted the feature for a specific workflow and will use it sparingly. The second crowd treats every new button like a personality upgrade. The second crowd is about to overtrade.
Tight trails on news events are a classic self-own. Spreads widen. Mark can jump. You get clipped, then the candle continues in the original direction while you explain to group chat that the platform “stopped you out wrong.” The platform followed the rules you set. The rules were just too proud.
Another misuse is stacking a tiny trail on top of high leverage and calling it risk management. It is not. It is a confession that the position cannot survive ordinary noise. Reduce leverage. Then decide whether a trail even belongs on the trade. Perhaps the most interesting aspect is how often a wider trail plus smaller size produces a calmer equity curve than a surgical trail plus hero sizing.
- Decide the trade thesis and the hard invalidation first.
- Ask whether the market is trending enough for a trail to help.
- Pick distance or percent from realized volatility, not from a round number you like.
- Choose immediate tracking or an activation price with intent.
- Size as if the fill will be worse than the trigger.
- Review the exit after the fact without rewriting history.
Mark Price Discipline And Why It Exists
Using mark price for triggers is a design choice with a purpose. Last price can be poked. Thin prints happen. A mark that blends index and basis components is harder to bully for sport, though nothing is invincible. If you trade perps, you already live with funding, basis, and the gap between your screen and the cash market. Trailing stops inherit that world instead of pretending it away.
That is also why documentation warnings about trigger versus execution keep showing up. They are not legal wallpaper. They are a description of matching engines under stress. When everyone hits the same side, the book is a hallway, not a warehouse.
Fees, Flow, And The Broader Tape
Order types do not print revenue by themselves. Flow does. Recent industry comparisons have put the platform’s year-to-date revenue in the hundreds of millions, a noticeable slice of a multi-billion pool across crypto projects in that same adjusted set. Take those league tables with the usual grain of salt. Methodologies differ. What you cannot shrug off is that perpetual activity has been large enough to keep product work funded and visible.
Fee design still matters to anyone who actually clicks buttons. Maker rebates for qualifying flow can go negative in the advertised sense for the highest tier names. Staked token thresholds can cut taker costs. An assistance fund that buys the native token on the open market remains part of the loop. None of that changes how a trailing stop calculates a retracement. It does change whether grinding a trail in and out of noisy markets is economically sane.
Collateral experiments have been part of the same season. Manual borrowing that lets users post the native token or bitcoin against stablecoin liquidity is a different product, but it sits in the same household as leverage. A 65 percent loan-to-value figure on the native token versus 50 percent on bitcoin is the kind of detail that looks small until you run the numbers on a book that is already using perps. Trailing stops do not neutralize borrow risk. They only shape how a derivatives position exits.
Integrations Change Who Shows Up In The Book
When a large consumer app pipes hundreds of perpetual markets through an existing wallet flow, the user mix shifts. Some of those users will never touch a trailing stop. Some will discover it on day two and trail every position like it is a video game power-up. Liquidity is not just depth. It is the habits of the people providing and taking that depth.
Geographic exclusions still apply in plenty of those consumer wrappers. That is not a moral lecture. It is a reminder that the same chart can represent different legal realities depending on where the click happens. If you are reading this as a professional, you already know to separate venue access, wrapper access, and the raw matching engine. If you are reading this as someone newer, start there before you start optimizing trail percentages.
HIP 3, Permissioning, And Messier Markets
Permissioned HIP 3 markets, even as an optional design, hint at a future where some perps look more like clubs than public squares. Deployers can hold the allowlist or hand the job to sub-deployers. That can be useful for compliance-minded experiments. It can also create pockets of liquidity that gap harder because fewer participants stand in the hole.
In those pockets, a trailing stop becomes more valuable and more dangerous at the same time. More valuable because you cannot stare at every listing. More dangerous because a percentage that worked on a flagship pair can be fantasy on a thinner contract. I would rather see traders keep a short list of “trail-eligible” markets than spray the feature across the entire board.
Quick filter I use before attaching a trail: Is the market trending or just twitching? Is the book thick enough after hours? Is my trail wider than typical noise? Can I accept a market fill in a flush? Do I still have a hard stop thesis if the trail never activates?
What This Means For Risk Culture
Risk culture sounds like a consultancy slide. Ignore the phrase. Look at behavior. Traders who used to flatten winners by hand at the first red candle now have a structured excuse to stay in. That can raise average win size. It can also raise the number of death-by-a-thousand-wicks sessions. Both outcomes are available. The order type does not pick for you.
I would rather a desk write a one-page policy than invent folklore. Something simple. Trails allowed only after a trade is green by X. No trails inside scheduled event windows unless size is reduced. No trail tighter than the average true range fraction you actually measured. Boring rules beat clever ones because people follow boring rules when they are tired.
Retail flow will not write that page. That is fine. Retail can still treat the feature as a teacher. Watch a trail for ten sessions. Write down every time you hated the exit. You will start to see whether you hate volatility or you hate being wrong. Those are different problems.
A Word On Token Price And Feature Headlines
Native token prints near prior highs tend to travel with product news whether the link is clean or not. Sometimes the market is just in a mood. Sometimes a borrow feature, a listing wave, and an order-type launch land in the same week and the tape tells a story that is neater than reality. I try not to build a trading plan out of that neatness.
If you are here for the token, trailing stops are still relevant. They change how professional flow may hold winners on the venue that generates fees. If you are here for the perps, the token is background music. Keep the layers separate. Mixing them is how people start trailing a coin they meant to invest in and investing in a trade they meant to scalp.
Practical Setup Notes That Rarely Get Written Down
Write the trail as part of the ticket, not as a rescue mission two hours later. Late trails are emotional. They get set after the open profit has already become identity. That is when people pick a distance that protects the ego instead of the account.
Keep a conventional invalidation in mind even when the trail is live. Markets can gap through a trail logic if the mark never prints the path you sketched on a napkin. Rare. Not imaginary. If the trade is large enough to matter, you should know what you will do if the automated exit is late or partial.
Do not trail the full size if the position is really two ideas taped together. Scale the runner. Let a piece come off at a fixed target. Trail the rest. Hybrid exits look less elegant on social media and far better in journals.
The trigger price and the actual execution price can differ because the condition can be true before the order finishes against live liquidity.
– A reality every derivatives trader eventually learns
Who Benefits First
Trend followers who already think in swings. Desk traders who cannot hover. Systematic folks who wanted a native trail instead of a bot that pings the API every few seconds. Maybe also the accidental swing trader who kept turning winners into noise because they clicked flatten too early. That last group might be larger than the marketing copy admits.
Who benefits last? Mean-reversion scalpers in a box. News gamblers. Anyone whose edge is speed through a one-tick window. For them, a trailing stop is a costume. Wear something else.
The Honest Limitation List
- Market-order exits can slip when the book is one-sided
- Choppy markets will harvest tight trails
- Activation mistakes can start tracking too early or too late
- Percentage trails behave differently as price level changes
- A trail is not a substitute for position sizing
- New listings and thinner HIP 3 books can exaggerate every weakness above
None of those limitations make the launch a gimmick. They make it a tool. Tools have grain. You work with the grain or you sand through the table.
What I Would Watch Over The Next Few Weeks
Three things. First, whether traders actually leave activation blank or start using it like a professional switch. Second, whether liquid majors and long-tail HIP 3 names get the same default trail habits, which would be a mistake. Third, whether people treat trailing stops as a reason to hold through events they used to flatten. That last one will show up in liquidation maps before it shows up in recap threads.
I also want to see how often users complain about mark-versus-fill after the first violent session. That complaint is a rite of passage. The quality of the reply from experienced users will tell you if the community is getting sharper or just louder.
A Closing Thought That Is Not A Slogan
Trailing stops will not make a weak thesis strong. They can keep a strong thesis from being sold out of boredom. That is a narrower gift than the headline, and a more useful one. If you trade perps on this venue, spend an afternoon with the settings. Use tiny size. Watch the trail climb and then give something back. Notice how your stomach reacts. That reaction is data.
Then write your own rule in one sentence. Mine is simple. If the market is not clean enough to trail, I do not deserve a fancy exit. I deserve a smaller position and a hard stop. The new button is still worth having. I would just rather see it used like a seatbelt than like a trophy.
The perpetual book will keep changing. Listings will keep arriving. Wrappers and permissioned experiments will keep testing what “access” even means. Through all of that, exits remain the unglamorous half of the craft. Hyperliquid just made that half a little less static. Whether that helps your account depends on whether you let the trail follow the market, or you force the market to follow your impatience.
