The Hype Bot trades a single market on a single signal. It goes long when RSI crosses up through 50 on a completed four-hour candle and short when it crosses down through 40 — then gets out of the way and lets a trailing stop decide when the trade is over.
An RSI crossing 50 is the most widely known signal in retail trading, and on any single chart it is close to worthless — it fires constantly and most fires are noise. What changes its character is refusing to act on anything else: no intra-candle entries, no averaging down, no second position, and no override when the chart looks tempting.
Every position carries two exits. A 10% hard stop is the floor. Above it sits a 0.55% trailing stop — it arms once the trade is 0.55% in profit, then follows the best price reached, 0.55% behind, and never moves backwards.
The detail that makes it work is when the trail is allowed to start. It cannot arm during the candle the position opened in. A trade has to survive one full four-hour candle before the trailing stop exists at all.
Without it, a trade that ticks 0.55% up in its first minutes arms the trail immediately and gets stopped out on the first wobble — booking a rounding error and, after fees, often a small loss. Tested across the full HYPE history at one-minute resolution, arming immediately turns the strategy from clearly profitable into a net loser. Same entries, same trail width, same hard stop. The only difference is the wait.
Every closed trade since December 2024, with the point where live execution began marked on the curve. Entry, exit, side, net result and exit reason for each one — nothing summarised away.
Astrabit copy trading will let you mirror the Hype Bot from your own account, with custody under your own keys. Every entry and exit would hit your account as it fires. Join the Discord waitlist to be notified the moment it goes live.
The curve comes from the strategy tester, on both sides of the go-live marker. It is charged a 0.2% round-trip commission but does not model slippage beyond that, and the simulator prices roughly 32 points per four-hour candle. At full one-minute resolution the same rules return closer to +0.5% per trade than the +0.8% the tester reports. Treat the published figures as the optimistic end of the range, not as realised fills.
HYPE rose roughly 450% across the period covered here. A long-biased engine on a trending asset flatters itself, and this record has never been tested through a sustained HYPE downtrend. There is no diversification: when HYPE is untradeable, the strategy has nothing else to do.
A tight trail wins often and small, and loses rarely but large — the average loss is several times the average win. That shape depends entirely on the win rate holding. If it slips a few points, expectancy inverts quickly. A high win rate is not the same thing as a safe strategy.
The hard stop sits well inside the liquidation buffer at the operating leverage, and it rests on the exchange so it fires whether the bot is running or not. It is not a guarantee: a gap through the level fills below it. Do not commit capital you cannot afford to lose.
Every trade is published, including the losing ones and the period before real capital was following the signals. Decide from the record, not the pitch.
The same swing engine run across a basket of volatile perpetuals from one shared Hyperliquid vault, one position per coin.
HYPE-dominant directional vault on Hyperliquid with AVAX diversification. 10% profit share, self-custody, on-chain.
Liquidation-anchored mean reversion across a curated crypto universe on Binance USD-M futures.