Skin in the game, tested: when the manager's own money still isn't enough
Three Hyperliquid vaults where the manager holds 42–62% of their own fund — far above the 5% floor — and results from −30% to +56% a year. The cleanest test there is of whether “skin in the game” means anything. It proves conviction, not competence.
"The leader has skin in the game" is the line that's supposed to make you relax. A manager betting their own money alongside yours won't gamble recklessly — right? I went looking for the cleanest test of that idea on Hyperliquid: vaults where the leader holds an enormous share of their own fund, far above the 5% floor. I found three, each with the manager holding 42–62% of the pot. Their results could not be more different — and that's the whole lesson.
The one-line verdict: all three managers are massively invested in their own vaults. One is bleeding badly, one is giving its gains back, one is genuinely good but won't tell you how. Skin in the game proves conviction, not competence.
What each one is
+convexity (leader 42%) discloses its strategy in full as: "+convexity." That's the entire description. Yet its record is the best of the three — about 56% a year at a shallow 12% drawdown, lightly deployed today. Strong, and a black box.
Edge & Hedge (leader 62%, the highest skin of all) is the honest one: a quantitative delta-neutral long/short book at about 3x leverage — and on-chain it's running exactly that, 54 positions at 3.1x gross. But it peaked near $226k in profit and has bled back to $42k.
FC Genesis — Quantum (leader 43%) sells the dream: "fully systematic, powered by ML on high-frequency data." The reality is a lifetime loss of $361k, positive in only 28% of periods, and after the damage it has retreated almost entirely to cash. High alignment, machine-learning buzzwords, losing money.
The receipts
| Metric | +convexity | Edge & Hedge | FC Genesis |
|---|---|---|---|
| Leader stake | 42% | 62% | 43% |
| Annualized return | ~+56% | ~+14% | ~−30% |
| Max drawdown | ~12% | ~24% | ~24% |
| Calmar | ~4.6 | ~0.6 | ~−1.3 |
| Positive periods | 38% | 46% | 28% |
| Current leverage | 0.2x | 3.1x | 0.06x (idle) |
| Strategy disclosure | None | Clear | Buzzwords |
Reading the risk
High stake, low crowd. All three are mostly the manager's own money — outside depositors have barely shown up. The alignment is real partly because almost nobody else is in.
Conviction cuts both ways. FC Genesis's manager has 43% of their own money in a strategy down $361k. They are losing right alongside depositors — exactly as "skin in the game" promises. Alignment worked; the trading didn't. That's the trap: alignment guarantees shared pain, not shared profit — the same lesson drkmttr taught, where a leader with 34% of his own money still rode the vault into the ground.
Disclosure doesn't predict results either. The most transparent vault here (Edge & Hedge) is fading; the most opaque (+convexity) has the best numbers. Neither honesty nor secrecy tells you the outcome — only the track record does.
Sonuç
Zoom out and the point lands harder. These managers staked 42–62% of their own money — and they span from −30% to +56% a year. Meanwhile Growi HF runs beautifully with its leader at the bare 5% floor, and elsewhere on the leaderboard a vault with the leader at 98% (Long LINK / Short XRP) blew up to a total loss. Across the entire spectrum, leader stake barely predicts the result. Skin in the game is a check on intentions — it tells you the manager will hurt when you hurt. It is not a check on skill. Treat it as one box on the list (one of many), never the reason you deposit.
Figures from Hyperliquid's public API and our twice-daily snapshots; leverage is the live gross figure. Annualized returns extrapolate each vault's own history. Nothing here is financial advice — one trader showing his work.