Training window
2026-01-01T00:00:00.000Z → 2026-01-01T03:00:00.000Z
- Strategy
- +834.20 bps
- Benchmark
- +800.00 bps
- Drawdown
- -1176.62 bps
- Modeled costs
- 0.0310
Strategy research
Test assumptions before preparing an action
Compare a bounded rebalancing rule with a hold benchmark, inspect modeled costs and drawdown, and research cross-venue price differences. This workbench cannot build, sign, approve, or submit a transaction.
Interactive simulated research
Choose a hypothesis and rerun the fixture
Every control below reruns the same checked-in synthetic observations in this browser. No AI model, wallet, provider, live price feed, or execution endpoint is contacted.
Simulated run 1 · deterministic result
Evidence fingerprint · balanced:790.77:500:2912.82:3:0.10416659
Research output
Strategy, benchmark, and drawdown
Balanced rebalance: Does a 50% XRP target with a ±5% band outperform holding after modeled costs?
Each evaluation window opens at pre-trade marked equity, so modeled costs at its boundary observation remain inside return and drawdown.
- Full strategy
- +790.77 bps
- Hold benchmark
- +500.00 bps
- Excess vs benchmark
- +290.77 bps
- Maximum drawdown
- -2912.82 bps
Out-of-sample window
2026-01-01T04:00:00.000Z → 2026-01-01T08:00:00.000Z
- Strategy
- +1383.00 bps
- Benchmark
- +1290.32 bps
- Drawdown
- -820.06 bps
- Modeled costs
- 0.0732
Sample comparison
How the three simulated hypotheses differ
These rows are recomputed from the same deterministic fixture. They are software evidence, not real AI recommendations or expected returns.
| Hypothesis | Strategy | Excess | Drawdown | Fills | Modeled costs | Use hypothesis |
|---|---|---|---|---|---|---|
| Balanced rebalanceDoes a 50% XRP target with a ±5% band outperform holding after modeled costs? | +790.77 bps | +290.77 bps | -2912.82 bps | 3 | 0.1042 | |
| Tighter intervention bandDoes intervening at ±2.5% improve tracking enough to justify more modeled fills? | +945.69 bps | +445.69 bps | -2879.19 bps | 7 | 0.2102 | |
| Cost-stress caseDoes the same rule still hold up when assumed slippage rises to 75 bps? | +779.75 bps | +279.75 bps | -2916.81 bps | 3 | 0.2102 |
Deterministic event log
Modeled research fills
3 fills · 0 partial
| Observation | Ledger | Side | Requested gross | Gross notional | Net cash flow | Modeled costs |
|---|---|---|---|---|---|---|
| fixture-2 | 90000002 | sell base | 6.0000 | 6.0000 | 5.9690 | 0.0310 |
| fixture-5 | 90000005 | buy base | 8.5704 | 8.5704 | -8.6104 | 0.0400 |
| fixture-8 | 90000008 | sell base | 6.6201 | 6.6201 | 6.5869 | 0.0332 |
Research alerts only
CLOB ↔ native XRPL AMM comparison
Operational alerts require fresh executable quotes from the same validated ledger. This page uses only the labeled deterministic fixture while applying CLOB costs, native constant-product AMM pricing, pool fees, modeled slippage, and two network-fee assumptions. XRPL AMMs are one pool per pair; this is not concentrated-range liquidity.
No qualifying net edge in this fixture
No route has a positive net edge above the configured threshold after modeled costs.
No automatic arbitrage execution is implemented or implied. Any future action must return to the existing builder, capability, policy, and wallet-approval flow.