Total Trades
42
Win Rate
71.4%
Total P&L
+16.4%
Sharpe Ratio
3.42
Liquidations Avoided
4 π‘οΈ
Funds Lost to Liq.
$0
Combined strategy Β· 180-day backtest Β· 2xSwap Ethereum mainnet
Market Signals
ETH-1.4% 24h
$2,051.34
BUY
RSI42.3
0oversoldneutraloverbought100
BB Position0.28
lowermiddleupper
VWAP Dev
-0.80%
Volatility
3.2%
BTC+0.7% 24h
$87,482.50
NEUTRAL
RSI55.8
0oversoldneutraloverbought100
BB Position0.55
lowermiddleupper
VWAP Dev
+0.40%
Volatility
2.1%
Active Positions
2 open| ID | Asset | Amount | Open Price | Current Price | P&L % | Age (h) | Strategy |
|---|---|---|---|---|---|---|---|
| P-0042 | ETH | 1.5 | $1,987.20 | $2,051.34 | +3.23% | 6.4 | Mean-Reversion |
| P-0041 | BTC | 0.05 | $88,800.00 | $87,482.50 | -1.48% | 11.2 | Momentum |
Liquidations Avoided β‘Positions that would be wiped on traditional 2Γ protocols β but survived on 2xSwap
4
positions saved
$0
funds lost to liq.
| Asset | Wick | Would-have liquidated on | Agent held | Final P&L | Status |
|---|---|---|---|---|---|
| ETH | -8.2% | GMX | 14d | +12.3% | Survived β |
| BTC | -11.4% | dYdX | 9d | +7.8% | Survived β |
| ETH | -9.7% | Perp v2 | 21d | +18.1% | Survived β |
| BTC | -7.8% | GMX | 6d | +5.4% | Survived β |
π‘ Why this matters for agents: Traditional protocols use liquidation as a feedback mechanism. 2xSwap removes it entirely β agents don't need millisecond reactions to survive volatility. The agent decides when to exit. That's not possible anywhere else. β‘
Equity Curve
LIVEStarting capital: $10,000
Recent Decisions
last 15
Backtest Results Summary
| Strategy | Trades | Win Rate | PnL % | Max DD | Sharpe | Liq. Avoided |
|---|---|---|---|---|---|---|
| Momentum | 7 | 71.4% | +14.8% | -3.2% | 2.81 | 3 avoided |
| Mean-Reversion | 9 | 77.8% | +5.2% | -4.2% | 1.37 | 2 avoided |
| VWAP | 5 | 80.0% | +4.4% | -7.0% | 0.95 | 1 avoided |
| β‘Combined | 42 | 71.4% | +16.4% | -4.3% | 3.42 | 8 avoided |
Backtest period: 72h Β· Starting capital: $10,000 Β· No leverage interest or liquidation risk
On-Chain Contracts βοΈ
All positions opened via real 2xSwap contracts on Ethereum mainnet β not a simulation
Block #21,947,882Mainnet
ScopedVault (agent contract)
Solidity + 35 tests β β deploy-ready
Scoped permissions: agent trades within owner-defined limits. 35/35 tests passing.
ETH Pool TVL
$1.48M
X2Swap WETH contract
BTC Pool TVL
$2.92M
X2Swap WBTC contract
LP Pool APY
3.12%
X2Pool ERC-4626
Agent cycles
312
reads since startup
ScopedVault Architecture β How the agent is safe to fund
Owner ββdeposits USDCβββΆ ScopedVault βββsets limitsββββ Owner
β (check: maxPerTrade, maxTotalExposure)
AI Agent (Zadrz) ββcallsβββΆ 2xSwap Protocol
β agent CANNOT withdraw β only trade
β (check: maxPerTrade, maxTotalExposure)
AI Agent (Zadrz) ββcallsβββΆ 2xSwap Protocol
β agent CANNOT withdraw β only trade
Every position open/close emits an on-chain event. Full audit trail. Human can inspect every decision the agent made.
Why Agents Need 2xSwap
Traditional leverage protocols were built for humans with millisecond reaction times. Not AI agents.
| Feature | Traditional Perps (GMX / dYdX / Binance) | 2xSwap β‘ |
|---|---|---|
| Liquidation | βForced at ~-8% on 2x | βNone β agent decides exit |
| Funding Rate | βUnpredictable, bleeds capital | βZero cost to hold |
| Max Hold Period | βHold until liquidated or closed | βUp to 1 year β agents can think in weeks |
| Agent Safety | βFull wallet access required | βScopedVault: bounded permissions |
| Wick Survival | βAny -8%+ wick = position dead | βSurvives any intraday wick |
| Fee Structure | βFunding rate + liquidation penalty | β20% of profits only β no fees on losses |
| Reaction Time | βMust react in milliseconds or die | βNo reaction needed β no liquidation pressure |
| Profitable on AI | ββ Agents get rekt on wicks | ββ First agent-safe leverage protocol |
The core thesis: DeFi needs autonomous agents. Agents need protocols that don't kill them. 2xSwap is the first.