Kalman stat-arb
LiveLong-only pairs rotation on JitoSOL/SOL driven by the Kalman filter's standardized innovation. Spot cannot short, so capital rotates into the cheap leg rather than shorting the rich one.
Live vs backtest
Divergence is not computable yet — the live window and the backtest window do not overlap enough. It will appear once the live session accumulates history inside the modelled window; we report it rather than smoothing it.
Live parameters
- capital_usd
- 125
- capital
- $125.00
Metrics
- Return
- 0.00%
- PnL
- $0.00
- Win rate*
- —
- Trades
- 0
- Max DD
- 0.00%
- Fees
- $0.00
- Slippage
- $0.00
- Status
- Live
- Trend
* Win rate counts closed sells vs VWAP cost basis; excludes unrealized PnL on open inventory
Exposure over time
Which leg the filter is holding: 100% = every unit in SOL/USD, 0% = every unit in JITOSOL/SOL. This is a share of position units, not of USD — the legs price differently, so an even value split does not read as 50%.
Not enough position history to draw exposure yet.
Trade history
0 fills · showing 0
| Time UTC | Market | Side | Qty | Price | Notional | Fee | Slip | Tag |
|---|
Data limitation
JitoSOL minute history from GeckoTerminal caps at roughly 75 days, so this strategy's backtest covers a shorter window than the others. Stated because it weakens the evidence, not despite it.