Execution and Rebalancing Flow
Surf moves capital safely, deliberately and only when it is worth it.
Surf is not a system that constantly trades or churns funds. It is designed to behave like a disciplined treasury operator:
Patient
Risk-aware
Cost-sensitive
And fully controlled by deterministic rules
The system is designed to answer one question repeatedly:
Is moving capital right now materially better than doing nothing, after risk, cost, and safety are considered? Only if the answer is yes, execution is even allowed to begin.
This page explains how capital actually moves inside Surf.
High-Level Flow
Every reallocation follows the same sequence:
Observe
Analyse
Simulate
Propose
Verify
Execute
Monitor
No step is skipped. No action bypasses control.
Step 1: Continuous State Monitoring
Surf continuously observes:
Lending and liquidity venues
Utilisation and borrow demand
Incentive emissions and decay curves
Liquidity depth and withdrawal capacity
Volatility and correlation shifts
Oracle deviations and price integrity
Gas, slippage, and execution cost
Cross-chain latency and bridge health
The data is aggregated across all allowlisted chains and venues. This creates a real-time state model of the opportunity and risk landscape.
2. Risk-Adjusted Analysis
The AI agent does not optimise for headline APY.
It scores opportunities based on:
Net yield after gas, fees, and slippage
Liquidity available for exit
Protocol maturity and historical behaviour
Correlation with existing positions
Concentration impact
Cross-chain execution complexity
The result is a risk-adjusted return profile, not a raw rate comparison.
3. Pre-Execution Simulation
Before proposing any move, Surf simulates:
The full execution path
All intermediate states
Slippage impact
Liquidity drawdown
Post-move portfolio composition
Updated exposure and concentration metrics
Worst-case unwind scenarios
If the simulation shows marginal improvement or elevated risk, the move is discarded. Most potential actions never reach execution.
4. Proposal Generation
Only after passing the simulation does the AI generate a proposal:
Which assets to move
From which venues
To which venues
Through which routes
In what size
Under what timing assumptions
At this stage, nothing has moved. This is only a structured plan.
5. Guardian Layer Verification
The proposal is then evaluated by the Guardian Layer.
It enforces:
Protocol allowlists
Exposure caps per asset and venue
Concentration limits
Slippage bounds
Liquidity exit thresholds
Cross-chain routing constraints
Portfolio health invariants
Stress and anomaly checks
If any rule is violated:
The action is rejected
No funds move
A rejection reason is logged
This is what converts AI from an operator into a constrained optimiser.
6. Deterministic Execution
If and only if all rules pass:
Transactions are constructed
Signed via MPC (Multi-Party Computation)
Batched where possible
Executed atomically when supported
Execution paths are predefined and auditable. There is no discretionary runtime behaviour.
Funds never leave user custody. The system only holds execution authority within pre-approved bounds.
7. Post-Execution Monitoring
After execution:
Portfolio health is re-evaluated
Risk metrics are recalculated
Drift from targets is measured
Performance is recorded
New baselines are established
The system returns to observation mode.
When Rebalancing Does Not Happen
Most of the time, the best action is no action.
Surf explicitly avoids:
Chasing small yield deltas
Over-optimising for short-term incentives
Moving capital when costs exceed benefits
Increasing risk to marginally improve APY
Rebalancing only occurs when:
The expected improvement is meaningful
Safety margins remain intact
Exit liquidity is preserved
Guardian constraints are fully satisfied
Why This Matters
Traditional yield optimisers focus on rate chasing. Surf focuses on state transitions under control.
This is the difference between:
“Where is the highest APY right now?”
and
“Is it safe and worth changing the entire system state to get it?”
Execution is not about being fast. It is about being correct, verifiable, and reversible.
This flow is intentionally conservative.
Surf optimises for:
Capital preservation first
Predictable behaviour
Long-term compounding
And user trust
Not for:
Maximum short-term yield
High turnover
Or speculative positioning
Key Properties
AI never touches funds directly
All moves are simulated before being proposed
All proposals are rule-checked before execution
All executions are deterministic and auditable
All outcomes are continuously monitored
In Simple Terms
Surf behaves like this:
“Only move money when the benefit is clear, the risk is bounded, and the rules allow it.”
That is how automation becomes safe for savings. That is the execution standard Surf is built on.
Last updated