Methodology
Basis Methodology

Basis / Cash & Carry Methodology

FYOS supports basis trading (cash and carry) as a first-class strategy family alongside perpetual funding arbitrage.

Both strategy families share the same core methodology principles:

  • Conservative, reality-adjusted metrics
  • Trust qualification before surfacing
  • Score vs deployability separation
  • Bounded interpretation of prescriptive outputs

What Is Basis Trading?

Basis trading captures the spread between spot and delivery futures prices.

When futures trade above spot (contango), you can:

  1. Buy spot
  2. Sell delivery futures
  3. Hold until expiry
  4. Capture the spread as prices converge

This is called cash and carry because you carry the spot position against the futures hedge.

Basis vs Funding: Core Difference

DimensionFunding ArbitrageBasis / Cash & Carry
UnderlyingPerpetual futuresDelivery futures
DurationOpen-endedFixed to expiry
Return sourceFunding paymentsBasis spread
Return behaviorVariableFixed at entry
Primary metricdecay_adjusted_returnmodel_adjusted_basis_apr
Capacity metricedge_capacity_24hdual_leg_capacity_usd

Primary Basis Metric: model_adjusted_basis_apr

The canonical basis return metric is model_adjusted_basis_apr.

This is not the raw spread. It is the spread after:

  • Round-trip trading fees (both legs, entry and exit)
  • Slippage costs from crossing spreads
  • Trust-based haircuts (inverse contracts, low-confidence data)

Why Gross Basis Is Not Deployable

Raw gross_basis_apr overstates realistic returns.

The gap between gross and model-adjusted represents execution reality:

gross_basis_apr: 25%
- fee impact: -3%
- slippage impact: -2%
- trust haircut: -1%
= model_adjusted_basis_apr: 19%

Always use model_adjusted_basis_apr for decision-making.

Dual-Leg Capacity

Basis trades require capital on both legs:

  • Spot leg: Buy the underlying
  • Futures leg: Sell the delivery contract

Capacity is constrained by the limiting leg:

dual_leg_capacity_usd = min(spot_capacity, futures_capacity)

The smaller leg limits the entire trade.

Trust Qualification

Basis opportunities pass through the same trust framework as funding:

QualitySurface Behavior
highDefault surface, full confidence
mediumDefault surface, some caution
lowMay be excluded from default views
excludedNot surfaced in normal flows

Trust qualification is separate from economics — a high-APR opportunity may still be excluded if data quality is poor.

Contract Types

Linear Contracts

Linear (stablecoin-margined) contracts have straightforward P&L.

FYOS applies standard reality adjustments.

Inverse Contracts

Inverse (coin-margined) contracts have non-linear P&L characteristics.

FYOS applies additional haircuts to inverse opportunities to account for this complexity.

Inverse contracts are clearly labeled in all surfaces.

Expiry Behavior

Basis returns are expiry-linked:

  • Shorter expiries have higher annualized APR for the same spread
  • But shorter expiries leave less room for execution error
  • Very short expiries may have outsized fee impact

FYOS categorizes expiry into buckets:

  • short: Near-term expiry
  • medium: Medium-term expiry
  • long: Long-dated expiry

Basis Planner Constraints

When including basis in Planner allocation:

  • Max basis share: Overall portfolio basis limit
  • Max per-expiry bucket: Concentration by expiry
  • Max inverse share: Inverse contract limit
  • Exchange concentration: Per-exchange limits

These constraints prevent over-concentration in specific expiry windows or contract types.

Basis Simulator Semantics

The Basis Simulator follows the same bounded philosophy as the funding simulator:

  • Hold to expiry: Models full basis capture at settlement
  • Early exit: Models partial convergence with uncertainty

Simulator output is scenario output, not validated forecast.

Fee-Tier Overlay Caveat

The Basis Screener supports a user-selected fee tier overlay.

This affects display economics only — it shows personalized effective APR based on your fee tier.

Important: This is a client-side overlay. It does not change:

  • Server-side trust qualification
  • Canonical model_adjusted_basis_apr
  • Underlying opportunity ranking

The overlay helps you see your potential economics but does not replace canonical trust semantics.

Shared Methodology Principles

Basis shares these principles with funding:

  1. Score is not deployability — attractiveness vs executability are separate
  2. Gross is not realistic — reality adjustments matter
  3. Trust gates surfacing — data quality affects eligibility
  4. Planner is bounded — recommendations, not instructions
  5. Simulator is scenarios — exploration, not prediction
  6. Reality is evidence — historical context, not guarantees

Further Reading

For detailed basis documentation:

Cookie preferences
We use essential cookies for docs functionality and optional analytics cookies to improve the beta documentation experience. You can accept or reject non-essential cookies. Learn more in our Privacy Policy.