How to Trade 10Y (US 10-Year Treasury Yield) on Hyperliquid

What the 10Y Perp Actually Tracks

The 10Y contract references the yield on the on-the-run 10-year US Treasury note — the most recently auctioned one, which is also the most liquid — expressed in percent. The quote is the yield. At the time of writing the perp prints around 4.70, and that is not a dollar price of anything: it is 4.70% annualized.

This inversion trips up traders coming from equity or crypto perps. Bond prices and bond yields move opposite each other. Going long 10Y is a bearish bond position. You are betting that Treasuries sell off, that long-end borrowing costs rise, that duration gets punished. Going short 10Y is the classic bond-bull, duration-long trade — you profit when yields fall and note prices rally.

The unit that matters here is the basis point: 0.01 in quoted price. Every rates desk on earth speaks in bps, and so should you when sizing this market. On August 14, 2026 the 10-year yield rose to 4.70%, a five basis point move from the prior session, after testing a 19-month high of 4.75% earlier that same week. Those are the increments you are trading.

Why this specific tenor? Because the 10-year is the reference rate for long-term US dollar borrowing. Mortgage rates key off it. Investment-grade corporate spreads are quoted over it. Equity analysts plug it into discounted cash flow models as the risk-free rate. When people say "the discount rate under every risk asset," this is the number they mean. The Fed sets the front end; the market sets this.

The Federal Reserve publishes the official daily series in its H.15 Selected Interest Rates release, and the full history lives in FRED as DGS10 — worth bookmarking both if you plan to trade this contract seriously.

Why 4.70% Is the Number Macro Traders Are Watching

Context matters more than the level. Over the previous decade the 10-year yield averaged 2.8%. Today's print is more than 60% above that baseline. This is not a normal rates environment, and the reasons are structural rather than cyclical.

The supply story. The United States borrowed $1.8 trillion in the first ten months of fiscal year 2026, including $432 billion in July alone — more in ten months than in all of fiscal 2025. That paper has to clear somewhere, and the clearing price is yield. Treasury has signaled in its quarterly refunding statement that it expects to maintain nominal coupon and FRN auction sizes for at least the next several quarters, so the supply is not letting up. The most recent 10-year auction was held August 12, 2026 and settled August 17.

The inflation story. University of Michigan year-ahead inflation expectations rose again in August, a fifth consecutive month above 4%, with energy prices pushed higher by conflict in the Middle East. Inflation expectations are what long-end yields are ultimately compensating for.

The policy story. At the July 28–29 FOMC meeting — Chair Kevin Warsh's second — the committee held the target range at 3.50%–3.75%, with three members dissenting in favor of a hike. Markets that entered 2026 pricing cuts have been repricing toward the possibility of tightening. That repricing shows up in this contract before it shows up almost anywhere else.

The fiscal feedback loop. Yields at these levels sit roughly 45 basis points above CBO projections. If sustained through the decade, the Committee for a Responsible Federal Budget estimates that adds $1.7 trillion to the national debt, pushing annual interest costs to $2.4 trillion by 2036 and debt to 124% of GDP. Higher yields make the deficit worse, and a worse deficit argues for higher yields. That reflexivity is the fundamental thesis behind a structural long position in this market.

How the HIP-3 Perpetual Works

10Y exists on Hyperliquid because of HIP-3, the builder-deployed perpetuals framework that went live on mainnet in October 2025. HIP-3 turned Hyperliquid from a single exchange into a market factory: any party staking 500,000 HYPE, with a 183-day minimum holding period, can deploy an independent perp DEX on HyperCore.

This market runs under the para deployer. The deployer defines the contract — it selects the oracle source, sets the leverage cap, and controls settlement behavior. The protocol's requirement is that the underlying be a "well-defined underlying asset or data feed which is difficult to manipulate and has underlying economic significance." A published sovereign benchmark yield is about as clean a fit for that standard as exists.

What the deployer does *not* rebuild is the plumbing. Matching, margining, order books, and liquidation all inherit from HyperCore. The same engine that clears BTC perps clears this one, and positions are margined in USDC. Fees on HIP-3 markets split 50/50 between the deployer and the protocol.

Two mechanical points worth internalizing:

Funding, not carry. In a cash bond position you earn coupon and pay repo. Here there is none of that. The funding rate exists purely to pull the mark price toward the oracle yield. Your carry is a mechanical anchoring payment between longs and shorts, not an economic yield. Do not model this like owning the note.

Deployer settlement power. The deployer holds a haltTrading action that cancels open orders and settles outstanding positions at the current mark price. Trading can later resume. This is a real, documented counterparty consideration that does not exist on validator-operated crypto perps, and it should factor into how much size you leave on overnight.

One structural quirk specific to this contract: the underlying is the *on-the-run* note, which rolls to a newly issued security through the quarterly refunding cycle. The benchmark being referenced changes identity over time even as the ticker stays constant.

Key Trading Considerations

Leverage is far more aggressive than it looks. Rates move in basis points, and the quoted price is a small number, so percentage moves compound fast. A five basis point session — an entirely ordinary day in Treasuries — is a 1.06% move from 4.70. At 20x that is roughly 21% on margin. The 4.63 to 4.75 range traded in a single week of mid-August is about 2.6% in price terms, or better than 50% on a 20x position. Conversely, under 25 basis points of adverse movement is enough to wipe out margin at maximum leverage before fees. Size for basis points, not for percent.

Liquidity is thin. Roughly $682,000 in 24-hour volume is a fraction of what flows through major crypto perps, and a rounding error against the cash Treasury market. Expect wider spreads, meaningful slippage on size, and the possibility that stops fill worse than you modeled. Use limit orders. Scale in.

Session mismatch is a genuine edge and a genuine hazard. The cash Treasury market keeps hours; this perp does not. Weekend and overnight geopolitical headlines — the Middle East energy shock being the live example — can move the fair value of the underlying while the reference market is closed. That creates gap risk on Monday and opportunity for traders who are actually awake and reading the tape.

Your calendar is a macro calendar. CPI, payrolls, PCE, FOMC decisions, quarterly refunding announcements, and coupon auction results are the event risk. A weak auction has historically been enough to move the long end several basis points in minutes. If you are carrying leverage into a 1:00 p.m. ET auction, you are taking a position on the bid-to-cover ratio whether you meant to or not.

Understand what you are hedging. Crypto portfolios are long duration in disguise: risk assets get repriced by the discount rate. A long 10Y position rises when the discount rate rises, which is often exactly when your spot book is bleeding. That correlation is the most defensible reason for a crypto trader to hold this market at all — not directional macro conviction, but a hedge against the rate that prices everything else you own.

Counterpoints to the structural long. Elevated yields are not guaranteed to persist. Forecasts cluster toward moderation into the 4.4–4.6% range depending on inflation and growth. Sustained high borrowing costs eventually slow the economy, which pulls yields back down. And if the Fed does hike into a slowdown, the long end can rally on growth fears even as the front end sells off. Curve dynamics are not one-directional, and a single-tenor perp gives you no way to express a curve view.