vildX
LAYER 1 · ENGINE 03← Back to architecture

Allocation

Sizes every position by risk-adjusted yield.

Allocation is the brain of vildX. Given the risk-vetted universe from Engine 02, it solves for the portfolio that maximises net yield per unit of risk — subject to vault tier, concentration limits, and liquidity constraints. The output is a target weight for every position in every vildX vault.

What it does

The capabilities this engine contributes to the vildX system.

Risk-adjusted yield model

Computes Sharpe-style score per position — net APY divided by composite risk score from Engine 02.

Constraint solving

Solves for optimal allocation subject to per-position caps, per-protocol caps, per-chain caps, and tier-specific liquidity floors.

Diversification

Forces minimum N positions per vault to prevent concentration — typically 5 for Prime, 7 for Frontier, 4 for Boost.

Tier-aware mixing

Prime gets only stablecoin-only positions; Frontier blends stables with RWA; Boost mixes token-denominated yield with fixed-rate PT.

How it works

Step-by-step flow inside this engine — from input signals to handoff.

  1. 1Receive cleared positions from Engine 02 with risk metadata.
  2. 2Compute Sharpe-style score for each position: net APY / risk score.
  3. 3Solve a constrained optimisation: maximise Σ (weight × score) subject to all caps.
  4. 4Produce target weights for each vault (Prime, Frontier, Boost).
  5. 5Pass the target allocation to Engine 04 (Rebalance) to execute.

Inputs & outputs

Data feeds and downstream consumers.

INPUTS
Risk-cleared universe

From Engine 02, with caps, scores, and constraints attached

Vault parameters

Per-tier targets, liquidity floors, diversification requirements

Current allocation

Live on-chain positions and pending rebalance actions

Yield forecasts

Forward-looking APY estimates blending live and historical data

OUTPUTS
Target portfolio

Per-vault, per-position target weights

Allocation deltas

Difference between current and target — input for Engine 04

Rationale logs

Per-rebalance reasoning published on the public dashboard

Other engines in vildX

The full Layer 1 stack — five engines working together.

⚠ Risk disclosures

vildX is a non-custodial software platform. Yields are variable and not guaranteed. Past performance does not predict future returns. Smart contract, protocol, and asset risks may result in partial or total loss of principal. Boost is not principal-protected. Not financial advice. Not for U.S. persons. Read full risk disclosures →