Have you ever opened several leveraged trades and watched one bad move quietly eat into the cushion you thought was reserved for something else? That exact tension sits at the heart of Near’s latest move. On October 22 the project flipped on cross margin for its Hyperliquid-powered private perpetuals, letting eligible positions pull from a single shared balance. The promise sounds efficient. The privacy language sounds comforting. Yet the real question that keeps coming back is simpler and sharper: once the margin is pooled, who can actually see the risk?
Near’s Cross Margin Launch And The Privacy Angle
The interface now supports more than fifty markets and advertises leverage up to forty times. Users outside restricted jurisdictions, including United States persons, can access it. Each trader receives a dedicated Hyperliquid trading account. Deposits and withdrawals into that account are described as confidential. That wording is deliberate and limited. It does not claim that every order, every position size, or every risk calculation stays hidden from every observer.
I’ve spent enough time watching derivatives platforms to know that capital efficiency and opacity rarely travel together without trade-offs. Cross margin lets surplus on one position support losses on another. Unrealized profit can become initial margin for a fresh trade. Idle capital shrinks. At the same time the entire account becomes the unit of risk. A sharp move against several correlated longs can cascade faster than isolated mode would allow.
Near’s short launch note did not publish user counts, posted collateral totals, or trading volume tied specifically to the new margin mode. That absence matters. Without those numbers the scale of adoption remains an open variable, and any claim about broad uptake stays speculative.
How Shared Margin Actually Changes The Picture
Hyperliquid’s own documentation draws a clear line. Cross margin shares collateral among all cross positions inside the account. Isolated margin keeps funds ring-fenced to one trade. In the shared model a surplus on the winning side can delay liquidation on the losing side. The reverse is also true: a heavy loss can pull collateral that appeared to sit behind a different strategy.
Consider a simple illustration. An account holds one thousand dollars of collateral and two positions, each carrying two thousand dollars of notional. A three-hundred-dollar mark-to-market loss on the first position and a two-hundred-dollar gain on the second produce a net one-hundred-dollar reduction in equity. Cross margin looks at the combined pool and the maintenance requirements that apply to total exposure. Isolated mode would have limited the damage to the margin assigned to the first trade alone. The practical difference is the path a loss follows through the rest of the account.
Leverage itself is not a permanent shield. It is checked at the moment a position opens. After that, price moves and funding payments reshape effective exposure. Closing a profitable-looking trade does not automatically free every dollar for withdrawal if another cross position is underwater. Collateral that looks attached to one line on the screen may already be spoken for by the broader account.
Spare balance can postpone a single liquidation while the overall account still loses more when several positions deteriorate together.
That dual nature is why the feature is neither automatically safer nor automatically riskier. It simply changes how losses are shared. Traders who switch an existing isolated position into cross mode need to see which pool then supports it and how the change affects withdrawable funds before they confirm.
The Dedicated Account And What Stays Confidential
Near states that each user gets a dedicated Hyperliquid account and that the funding path into and out of that account is confidential. Users sign the relevant transactions on their own devices. Near Intents convert assets from connected chains into the margin currency. Those three elements—the conversion, the transfer, and the subsequent orders—are distinct.
The product page is explicit about the deposit and withdrawal leg. It is quieter about the rest of the journey. Hyperliquid still needs to know position size, direction, collateral, and mark price in order to enforce margin and run liquidations. A privacy layer can obscure the real-world owner or the original funding path while the trading system retains enough information to settle and liquidate.
Visibility can shift across the life of a trade. A deposit may travel through a protected route, an order may hit a public market, and a withdrawal may leave through another protected route. Timing, size, and market choice can still offer clues to an observer who already holds outside information about a person. The product page does not quantify how hard such linkage would be, nor does it claim a formal anonymity set.
In my view the most useful framing is to separate identity privacy from position privacy. A dedicated account can break the obvious link between a user’s broader wallet history and a trading venue. That is valuable. It does not automatically hide the portfolio pattern that forms inside the margin pool itself.
What The Exchange Still Needs To Calculate
A perpetual has no fixed expiry. The venue maintains the position, updates unrealized profit and loss, and settles funding. It must continually refresh the mark price used for margining, liquidation, and unrealized P&L. That mark draws on an oracle component, local order-book data, and other perpetual markets. It can diverge from the last trade visible on a chart. A trader can therefore hit a maintenance threshold even when the most recent printed price looks less alarming.
During fast markets the gap between external prices and the local book can widen. Anyone trying to add margin faces the speed of conversion and transfer while the mark keeps moving. The launch materials did not quantify typical top-up times through the Near routing path. For a leveraged trader that missing performance detail often matters more than a general promise of convenient cross-chain funding.
Hyperliquid’s withdrawal rules add another constraint. Funds can leave a cross account only if the remaining equity still meets both the initial-margin requirement and a floor equal to ten percent of total open notional. A trader may see a healthy-looking balance and still be unable to withdraw the full amount. The rule exists to stop a withdrawal from leaving the surviving positions under-funded at the exact moment of the transfer.
Liquidation Across The Shared Pool
Cross positions become subject to liquidation when account equity, including unrealized profit and loss, falls below maintenance margin for total open exposure. The process first attempts to reduce the position through the order book. If that restores compliance, remaining collateral stays with the trader. If equity drops below two thirds of required maintenance margin and ordinary execution has not succeeded, a backstop can transfer the cross positions and their shared collateral to a liquidator vault.
The two-thirds threshold is not the only trigger. Maintenance breach comes first. Confusing the two levels can make an account appear to have more breathing room than it actually does. Liquidity in the order book also matters; an attempted close during a rapid move may not fill at the hoped-for price.
Funding payments run hourly and transfer value between longs and shorts. They can quietly erode a thin buffer even without a large directional move. When two correlated positions move against the account at the same time, the shared pool absorbs the combined impact. A deliberate decision to use cross margin for capital efficiency therefore requires monitoring every position together, not only the one most recently opened.
- Maintenance breach starts the liquidation sequence
- Order-book reduction is attempted first
- Backstop activates below two thirds of required maintenance if the book fails
- Isolated positions follow a narrower transfer rule that does not touch the separate cross pool
Correlation risk is especially visible in a broad market decline. Long exposure across several tokens can look diversified by ticker while the prices fall in unison. Cross margin aggregates that effect. A hedge on the opposite side may offset some of the loss, yet it is not guaranteed to move at the same speed or at the right moment, and funding can alter the balance further.
The Forty-Times Ceiling Is Not Uniform
The landing page highlights leverage up to forty times across more than fifty markets. That ceiling does not apply equally to every asset. Maximum leverage is set by market and can tighten through margin tiers as position size grows. At forty times, one thousand dollars of initial margin could theoretically support forty thousand dollars of notional before fees under a simplified calculation. A modest percentage move against that exposure can consume a large fraction of the starting balance. Actual requirements depend on the specific market parameters and on every other cross position already open.
The product explicitly excludes United States persons and users in other restricted jurisdictions. Eligibility terms sit alongside the technical design and should not be overlooked. No country-by-country volume breakdown appears in the announcement, so any statement about geographic reach would be guesswork.
Funding from a different chain introduces practical friction. Conversion rates, execution through Near Intents, and arrival of the correct margin currency all take time. Device-side signing confirms authorization; it does not guarantee that a top-up will arrive before a fast liquidation sequence completes.
Separating The Interface From Spot Activity
Near began trading against USDC on Hyperliquid’s spot market in September, and a Near perpetual contract already existed. Coverage of that listing referenced hundreds of millions in open interest at one snapshot. Importing that figure into the October cross-margin launch would be misleading. The earlier number described a token market on the venue, not collateral or activity flowing through Near’s private interface.
The connection is infrastructural. Near provides a route into markets that Hyperliquid operates. A trader can hold Near spot, trade a Near perpetual, or use the Near interface to reach dozens of other perpetual markets. Only the last activity involves the new cross-margin control. Token price moves or venue-wide open-interest totals cannot prove that users adopted the shared-margin feature.
A useful future disclosure would report, for a consistent period, the number of distinct active accounts, the amount of collateral actually routed through the interface, notional trading volume, and the count of cross-margin liquidations, all while preserving individual privacy. Aggregation can reveal scale without naming users. Any such report should also state whether the figures include only traffic originated through Near or the broader venue activity, and how double-counting is avoided when assets move between accounts.
Drawing A Precise Boundary Around Confidentiality
Traders really do leak strategy information when wallet histories, funding transfers, and open positions are easy to link. A dedicated account with confidential funding can break the most obvious connection. That is a genuine improvement. At the same time a person can still reveal themselves by announcing a trade, reusing identifying details, or through records held by an intermediary. Privacy is always a property of a specific data flow and a defined adversary, not a blanket label.
An independent review would need a clear threat model. Can a casual public observer link deposits to the trading account? Can a sophisticated analyst combine timing with external wallet activity? What does the routing service see, what does the frontend see, and what does the underlying venue see? Those are separate questions. An assessment that answers only the first should not be presented as proof of the others.
Perhaps the most interesting aspect is that the product can protect identity while still exposing a portfolio pattern inside the margin pool. The venue needs that linkage to calculate how one loss affects shared collateral. A customer deciding whether the offered confidentiality is sufficient should therefore ask the provider to distinguish identity privacy from position privacy in plain language.
Practical Points Worth Watching
Several concrete items would help anyone evaluating the feature over the coming weeks.
- A product-specific map of which data are public and which remain private across Near Intents, the dedicated account, and Hyperliquid
- Distinct account counts and collateral actually moved through the new cross-margin mode, separated from venue-wide figures
- Clear confirmation of which account modes, collateral types, and markets can share funds inside the interface
- Aggregate statistics on cross-margin liquidations, order-book success rates, and backstop outcomes over a defined window
- Updated restricted-jurisdiction language that leaves no ambiguity about eligibility
Until those pieces appear, the responsible description stays bounded by the product page’s own wording: deposits and withdrawals are confidential, each trader uses a dedicated account, and the underlying venue continues to price and liquidate according to its published rules.
Putting The Risk In Everyday Terms
Think of cross margin as a single household budget that covers every ongoing expense. A surplus in one category can cover a shortfall in another. That flexibility is useful until several categories overrun at once. Isolated margin is closer to separate envelopes: each expense has its own limited cash, and overspending one envelope does not empty the others. Neither approach is universally better. The choice simply decides how far a single setback can travel.
On this platform the shared envelope lives inside a dedicated Hyperliquid account reached through a route marketed as confidential. The exchange still sees the positions and the equity. Outside observers see whatever the public market data and the venue’s account-level disclosures reveal. The funding path itself is the main privacy claim. Everything else remains a matter of inference, timing, and external information.
I’ve found that the traders who handle shared margin best are the ones who treat the whole account as the true risk unit rather than any single line item. They watch funding rates, correlation, and the gap between mark price and last trade. They also treat estimated liquidation prices as living numbers that shift when other positions move or when funding accrues.
The new feature makes that venue model available through Near’s interface. It does not rewrite the underlying risk mathematics. Capital efficiency improves. Loss sharing becomes explicit. Privacy is scoped to the deposit and withdrawal legs. Anyone using the product benefits from keeping those three statements in view at the same time.
Why The Missing Numbers Still Matter
A launch announcement establishes availability. Adoption requires evidence that accounts remain active after the first funding transaction. Open interest measures outstanding exposure. Trading volume measures turnover. Posted collateral shows capital actually at risk. None of those metrics is interchangeable. If figures appear later they should distinguish interface-originated traffic from the broader venue and explain how multi-market traders are counted.
Without those numbers the conversation stays qualitative. The design is clear enough. The privacy boundary is stated. The economic scale is not. That gap is ordinary at the moment of launch; it becomes more noticeable once the feature has been live for weeks or months.
For now the practical takeaway is straightforward. Shared margin offers flexibility and requires broader monitoring. Confidential funding reduces one common form of leakage while leaving the trading system itself with the information it needs to operate. The risk is visible to the account holder, to the exchange that enforces margin, and potentially to any party that can observe the dedicated account’s activity. Identity may stay obscured. The portfolio pattern inside the pool does not disappear.
That distinction is the heart of the story. Near’s private perps now share margin. The question of who can see the risk receives a layered answer rather than a single yes or no. Understanding those layers is the difference between using the feature deliberately and discovering its boundaries only after an unexpected drawdown.
