Hook
Forty times leverage. Two attempts. Two failures. On August 23, 2024, Maji Fund leader Huang Licheng tried to open a 40x long BTC position—first quietly, then with a $24.3 million face-value push. Both got liquidated before they could settle. The collective loss: $165,000. Not catastrophic for a fund of that size, but the signal is deafening. Math doesn't care about conviction. The fund then rotated into a $75 million ETH long at $2,370, currently up $1.96 million. The question isn't whether ETH will rally—it's whether the same high-leverage strategy that failed on BTC will now succeed on a different asset, or if the structural fragility of 40x leverage will eventually catch up.
Context: The Fund and the Man
Maji Fund operates in the opaque layer of crypto capital allocation. Huang Licheng is a known figure in Chinese-speaking crypto circles, often associated with the moniker "Liangxi"—a trader famous for high-risk, high-leverage positions. The fund's strategy appears centralized: a single decision-maker with a high risk appetite. They hold longs in BTC (failed), ETH ($75M), HYPE (~$19.85M), and PUMP (~$4.87M). The exact trading venue is unspecified, but 40x leverage suggests a centralized exchange (Binance, OKX) or a permissionless derivatives protocol like dYdX or Hyperliquid. The latter is plausible given the HYPE holding—a token tied to the Hyperliquid ecosystem, a decentralized perpetual exchange. This is not a passive index fund; it's an active, aggressive trading desk betting on directional moves.
Core: Code-Level Analysis of the Trade Mechanics
Let's dissect the numbers. A 40x leverage means a 2.5% adverse move wipes out the entire position. The initial BTC longs were attempted twice—first presumably at a lower amount, then at $24.3M notional. The $165K loss implies a roughly 0.68% adverse move on the second attempt. That's a tight margin: BTC's daily volatility often exceeds 2%. The fund got stopped out by a minor swing.
Now the ETH long: $75M notional at $2,370. With 40x leverage, the required margin is $1.875M. A 2.5% drop to $2,310.75 triggers liquidation. The current price is presumably above $2,370 (since they show $1.96M unrealized profit), but that's a 2.6% move from entry. So they are already close to liquidation if the market reverses. The math is brutal: $75M * 2.5% = $1.875M loss. That's the entire margin. The fund's net asset value is unknown, but a single 2.5% ETH drop could wipe out this trade entirely.
Smart contracts execute. They don't negotiate. If the liquidation engine fires, it's instant. There's no room for a margin call if the position is on a decentralized platform like Hyperliquid—the contract will close the position at the best available price, potentially causing slippage if the liquidity pool is thin. The $75M size is significant for ETH perpetuals. On Binance, the open interest for ETHUSDT is around $5-6 billion. A $75M long is about 1.25% of that. Enough to cause a local cascade if the price dips and the liquidation engine starts selling.
I've audited liquidation logic in Aave V2 and observed how concentrated positions can amplify market moves. The same principle applies here: a single large position with high leverage creates a feedback loop. If ETH drops below $2,310, the stop-loss triggers, and the market order to sell equivalent ETH perpetuals pushes the price further down, potentially liquidating other leveraged longs. This is the classic "liquidation cascade" pattern. The community governance of the protocol (if it's a DEX) might have parameters to prevent this—like partial liquidations—but many venues still use full liquidation.
Contrarian Angle: The Illusion of Smart Money
The market narrative will likely interpret Maji's shift as "smart money rotating from BTC to ETH." I disagree. This is a desperate gambler switching tables after losing two hands. The fund attempted 40x on BTC and failed. Instead of reducing leverage, they doubled down on ETH with the same aggression. This is not a signal of conviction; it's a signal of overconfidence compounded by loss aversion. Liquidity is an illusion until it isn't. The $75M ETH position provides a temporary bid, but it's a fragile one. If the fund is forced to unwind, the liquidity they provided will evaporate just as fast.
Moreover, the fund's holdings in HYPE and PUMP are small by comparison. HYPE is illiquid—its market cap is likely under $200M, with daily volume less than $10M. A $19.85M long position in HYPE represents a massive percentage of the token's free float. If HYPE drops 10%, the fund's unrealized loss could be $2M. That's a concentrated risk vector. The same applies to PUMP, a meme coin from Solana's Pump.fun ecosystem. These are not hedges; they are speculative bets on top of speculative bets.
The real blind spot is the lack of any hedging strategy. The fund is net long BTC, ETH, HYPE, and PUMP. No shorts, no options, no stablecoin reserves. This is a portfolio with zero downside protection. In a bear market or a sudden crash, the fund could be wiped out entirely. The 40x leverage on ETH is the most dangerous part. A 2.5% drop is normal for ETH. One bad news event—like a regulatory crackdown or a hack on a major exchange—could trigger that drop. The fund's entire $75M position would be liquidated, and the loss would be $1.875M plus any slippage. That's a 100% loss on the margin.
Takeaway: Vulnerability Forecast
Watch ETH price action around $2,310. That's the liquidation threshold. If ETH dips below that, expect a rapid sell-off as Maji's position is unwound, potentially dragging ETH down 3-5% in a short period. The same goes for HYPE and PUMP: any sudden drop could trigger a cascade. The fund's risk management is a ticking time bomb. The market should not view this as a bullish signal for ETH. It's a warning about the fragility of leveraged capital. The question is not whether Maji will be right about ETH, but whether they will survive the inevitable volatility. Based on my experience auditing ZK-rollup state transitions and DeFi liquidation engines, I can tell you: the math always wins. And 40x leverage is a math problem that ends badly 90% of the time.
