97% of day traders who stick with it lose money anyway.1 Usually the plan was fine. They just couldn't leave it alone. HyperForge runs hard-tested strategies on your own Hyperliquid account and lets nobody touch them mid-trade, including you. You set a dollar cap and a daily loss limit. The kill switch is always one click away.
No card. Non-custodial. Your keys, your account, your kill switch.
Every strategy trades inside limits you set once and cannot talk yourself out of at 2am. The cap, the halt, and the kill switch sit next to the live curve where you can see them work.
Every closed trade of our longest-running strategy, in order, after fees, from the trade database.5 The drawdown stays on the chart because it happened.
Read the shape, not the endpoint. One cluster of trades in late May produced a large share of this curve, and the flat stretch after it is real. A record drawn to look smooth has been drawn wrong.
| Strategy | Start capital | Net P&L | Return | Fills | Days | Sharpe | Max DD |
|---|---|---|---|---|---|---|---|
| S-01 · Divergence |
$272.91 | +$81.42 | +29.8% | 253 | 132 | 1.98 | -13.5% |
| S-02 · Reversal |
$711.13 | +$65.33 | +9.2% | 28 | 66 | 3.34 | -0.8% |
Each strategy is quoted on its own capital. Start capital is the account value when the strategy took its first trade, and Return is simply Net P&L divided by it: $81.42 on $272.91 is +29.8%, $65.33 on $711.13 is +9.2%. Sharpe and drawdown come from each strategy's own trade sequence, after fees. Nothing is averaged, blended or cherry-picked across strategies.
This is a small sample. 134 days and 281 fills is not enough for certainty. Annualised Sharpe runs high at low trade counts and should fall as the sample grows, and 28 fills on Reversal is far too few to conclude anything.
The record above updates as it trades. Watching is free.
A good backtest is cheap to manufacture, so a backtest alone funds nothing here. Before a strategy touches an account it survives six gates, and any one of them can kill it. Most die there. That attrition is the product.
Walk-forward windows, a deflated Sharpe that discounts every parameter we tried, an overfit probability score, and a look-ahead audit that voids any strategy caught reading the future.
The strategies trade your own Hyperliquid account through an API key you generate and can revoke in seconds. The key can trade. It cannot withdraw. Your money never passes through us, on any tier, ever.
A dollar cap no position can size past, a daily loss halt that ends the day early, and a kill switch that flattens the book in one click. Every fill and every fee lands in a log you can audit against the chain.
Of everyone who began day trading Brazilian equity futures over three years and persisted past 300 sessions, 97% lost money. The authors found no evidence that practice improved results.
On 10 October 2025 one tariff headline liquidated $19 billion of leveraged crypto positions in a day, across 1.6 million traders. The largest wipeout ever recorded, roughly nine times the previous record.
Regulators make brokers print their own client outcomes. Plus500's current disclosure says 76% of retail accounts lose money on leveraged products, and across brokers the figure clusters near 80%.
Every trader can list the failure modes. Each one meets a hard stop that ships with the product, and the two that matter most are enforced somewhere we cannot reach.
The API key you issue can open and close positions. It cannot withdraw, transfer, or touch anything else, and that restriction lives at the exchange, outside our reach.
A dollar ceiling on total margin in use. No strategy can size past it, alone or in combination, whatever the signal says.
Hit your threshold and trading stops until the next session. A bad day stays exactly one day long.
One click flattens every position and cancels every order. It works from your phone, mid-trade, at any hour.
Every fill, fee and halt lands in a log you can check against the chain. If we misbehaved, you could prove it.
The standing commitments, in writing: no custody, no return promises, no buried records. The FAQ repeats each one.
Automation platforms are not new. The differences that matter are what gets proven before money moves, and what happens when a strategy stops working.
| Trading it yourself | Bot marketplaces | HyperForge | |
|---|---|---|---|
| Who executes | You, at every hour the market is open, which is all of them | Any strategy anyone chose to upload | Strategies that cleared six statistical gates |
| Proof before money | Your memory of it working last month | Self-reported backtests and screenshots | Walk-forward, deflated Sharpe, overfit score and leak audit, published |
| When it stops working | You notice eventually, at a price | Quietly delisted | Retired in the open, and the FAQ says so in advance |
| Worst-case guard | Willpower at 3am | Per-bot settings, if you remember each one | Account-level dollar cap, daily halt, one-click kill switch |
| Your funds | Your exchange account | Varies with each bot's permission scope | Your account only, trade-only key, revocable in seconds |
No code, no servers, no deposit to us. You connect, you cap, you switch a strategy on. From there the discipline is mechanical.
Generate a trade-only API key on Hyperliquid and paste it in. The key cannot withdraw, and you can revoke it in one click without asking us.
A dollar ceiling on margin and a daily loss halt. Hard limits, not suggestions: no strategy can size past the cap, and a bad day ends early.
Pick from the published strategies and watch every fill, fee and halt land in your log as it happens. The kill switch stays one click away.
Every tier sees the same dashboard and the same record. Founding rates hold for the life of the account, and you can cancel from inside the app in two clicks.
Watch the live account with real numbers before risking anything.
No card required
The published strategies trade your own Hyperliquid account.
Founding rate, normally $129
Your own idea, through the same six gates, then live.
Founding rate, normally $399
Hyperliquid charges its own maker and taker fees on every fill. Those are the exchange's, not ours, and they are already deducted from every figure on this page.
No, at no point. Strategies trade your own Hyperliquid account through an API key that you generate and can revoke in seconds. The key permits trading, never withdrawals. If you closed your HyperForge account tomorrow, your funds would not move.
You should not trust it completely, and we would rather write that than let you discover it. 134 days and 281 fills is a small sample, and annualised Sharpe runs high at low trade counts.
What the record does establish is narrower and still worth something: these strategies place real orders on a real book, the fills and fees are real rather than modelled, and the numbers are read straight from the trade database rather than assembled for marketing.
It gets retired. Every strategy is monitored against its own validation baseline, and a persistent gap between live and expected behaviour ends it. Retirement has happened before and will happen again. Strategies are not defended because they were expensive to build.
Real-time fills are part of the paid product. The 24 hour delay is short enough that you can audit everything we publish against the chain, and long enough that the free tier is not a zero-cost signal feed.
Technically a few hundred dollars, since Hyperliquid's own minimums are low. Practically we suggest $1,000 to $2,000 or more, so position sizing is not dominated by minimum tick and lot rounding. Below that, rounding error eats a meaningful share of the edge.
Two families are live today: open interest divergence, where open interest climbs while price stalls and late entrants get trapped, and engulfing reversals, a specific candle structure at a level with a volatility filter. Both are public knowledge. The work is in proving which parameter settings survive the gates, and in rejecting the far larger number that do not.
No. HyperForge is execution software. It does not know your circumstances, it does not recommend that you trade, and nobody here is a licensed adviser. Trading perpetual futures with leverage can cost you more than you deposited.
The strategy does not get tired at three in the morning, does not need to win it back, and does not widen a stop because this one feels different. That was always the hard part. It was never the chart.