Yen Intervention Fade

Thesis

Bets on a stronger yen by shorting USD/JPY near Japan's intervention levels, staying in USDC if the dollar never climbs high enough.

Strategy

• Entries: One campaign: 40% after an hourly close at 158.50, 60% at 159.50, only in live PT weekday hours through October 23, with liquidity checks and clips. • Sizing: Keeps the allocation-times-leverage cap and four-hour clips; the 60% step remains eligible after a partly filled 40% step expires. • Exits: Takes half at 155.50 and half at 153.50; exits on a 161.20 hourly close, a 162 stop, weak-yen tolerance, or October 29 at 18:00 PT.

Exposure

• Holds: JPY short and USDC — Shorts xyz:JPY or waits entirely in cash; never starts a second campaign. With both steps filled near their thresholds, the hourly stop is about 1.3% above average entry; the native stop also protects weekends. • Venue: Hyperliquid xyz — Starts with 50 Core USDC minimum in the isolated-only JPY market. Uses PT trading hours because Circuit exposes no live pricing-mode flag. Choose leverage yourself when starting. • Leverage: You choose, 1–50x — Allocation times your choice, capped by the live market maximum. Planning losses including funding: ~1.5% at 1x, ~7.5% at 5x, ~15% at 10x before gaps. Held-position runs warn at roughly 30x+, where liquidation may precede the stop. • Costs: Trading fees, funding — Pays fees and variable funding on position size. Your assumptions: shorts pay 7–10% annually and typical hourly volume is $23,000. Actual liquidity sizes clips. Weekend or intervention gaps and slippage can increase losses. • Data: Circuit and native news — Uses candles, prices, liquidity, holdings and orders. Credible intervention tightens a profitable stop to entry and takes half at a 2% gain; warnings and early rate hikes leave exits unchanged. Stops never widen. • On unwind: Core USDC — Exits October 29, 2026 at 18:00 PT before the BOJ decision; cancels resting orders. Native protection is reduce-only, but active closes use exact-quantity non-reduce-only orders, unlike your requested policy.