Reading a Losing Week on Hyperliquid (Full Public Logs)

The week in numbers

Journal issue `/journal/2026-W28` reports: **9 trades closed**, **22.2% win rate**, **−$13.76 net**, **$3.42 fees**, average hold ~50 minutes. Symbols: BTC (4), ETH (5). Best trade +$1.97 BTC; worst −$5.75 ETH.

Past performance is not indicative of future results. This is a journal of one week of live data, not a promise of recovery or future returns. See `/risk`.

Exit mix (what actually closed positions)

Exit reasons in the public stats: TRAIL_STOP 5, SL 2, MANUAL 1, EXTERNAL_CLOSE 1. That mix is the point of the journal: you can see *how* trades ended, not just a headline ROI.

Fees are not optional

Even a quiet notional footprint paid $3.42 in fees. On small accounts, fee drag dominates. This is why we discourage under-capitalized starts — not because of marketing, because of arithmetic.

How to read a losing week without self-deception

1. Separate process quality from short-window P&L. 2. Check whether exits match the published rules (trail / disaster / max hold). 3. Compare to the paper/shadow streams if you follow research variants. 4. Never average up emotional capital after a red week.

Compare approaches honestly

If you want multi-exchange grid/DCA toolkits, see `/compare`. HyperAgent is Hyperliquid-only with one published strategy and a public log. Pick the fit — not the loudest claim.

Apply it

Start from `/register`, explore the console free, then add a payment method only when you are ready for live trading. Trial terms and pricing: `/pricing`.