Double the leverage, not double the money: Growi HF vs. its 2x twin
HyperTwin runs the exact same strategy as Growi HF — at 2x leverage. It's the closest thing to a controlled experiment on Hyperliquid. The result: barely more headline return, 4.5x the drawdown, and a lifetime loss while the 1x version quietly compounded $1.65M.
You almost never get a clean experiment in markets — too many variables move at once. Hyperliquid just handed us one. Growi HF runs a quantitative mean-reversion strategy across every coin on the exchange. HyperTwin runs the exact same strategy — at 2x leverage. Same engine, same trades, one variable changed. So we can answer a question every investor secretly believes they know the answer to: if you double the leverage, do you double the money?
You do not. Here's what actually happened.
The receipts, side by side
| Metric | Growi HF (1x) | HyperTwin (2x) |
|---|---|---|
| Actual portfolio leverage | ~1.0x | ~2.2x |
| Annualized return | ~58% | ~69% |
| Max drawdown | ~14% | ~64% |
| Calmar (return ÷ drawdown) | ~4.1 | ~1.1 |
| Worst single step | −13% | −40% |
| Lifetime profit | +$1.65M | −$45.8k |
| Assets vs. peak | Holding | Collapsed ~83% ($0.7M→$0.12M) |
Double the leverage barely moved the return — and quadrupled the pain
Look at the headline first: the 2x vault's annualized return (~69%) is only a little above the 1x (~58%) — nothing like double. But the drawdown went from 14% to 64%, roughly 4.5x deeper, and the worst single step tripled, from −13% to −40%. The risk-adjusted return — the Calmar — collapsed from 4.1 to 1.1. You paid 4.5x the suffering for a sliver more headline return.
Why: deep holes don't climb back
This is the math that leverage marketing never shows you. Losses and gains aren't symmetric. A 14% drawdown needs a +16% gain to recover. A 40% step needs +67%. A 64% drawdown needs +178% just to get back to even. Doubling leverage doubles every drop — and on a mean-reversion strategy that lives on many small wins, doubling the drops turns ordinary, recoverable dips into craters the compounding can't climb out of. That's "volatility drag," and it's why the same strategy that made $1.65M at 1x is sitting at a loss at 2x.
The market already voted
The clearest verdict isn't in the metrics — it's in the money. HyperTwin's assets collapsed about 83%, from a $0.7M peak to roughly $120k, as depositors who lived through that 64% drawdown left. A single whale now holds 62% of what remains. The 1x engine, meanwhile, still holds over $7M and keeps compounding. Given the identical strategy, depositors chose the un-leveraged version with their feet. They were right to. The drkmttr blowup showed leverage meeting a fast market; this shows the slower, quieter way leverage bleeds a perfectly good strategy to death.
The bottom line
Leverage is not a return multiplier. It's a risk multiplier that only sometimes, in a straight line up, pays like one. The instant the path gets bumpy — which is always — the deeper drawdowns compound against you, and a strategy that wins at 1x can lose at 2x. If you want more return from a good vault, the honest lever is to allocate more capital to the un-leveraged version, not to buy the geared twin. Whenever you see a "2x" or "3x" version of a vault you like, read this table before you click. The full risk checklist is in what can go wrong.
Figures computed from Hyperliquid's public API and our twice-daily snapshots; leverage is the live gross figure (notional ÷ equity). HyperTwin mirrors Growi HF's strategy at 2x but is a separate vault, so returns differ slightly from a literal doubling. Nothing here is financial advice — it's one trader showing his work.