Skip to content

CASE 06 · DELTA-NEUTRAL CARRY

Funding-Rate Carry Framework

Bybit spot/perpetual pairs · Funding z-scores · Depth-aware position sizing

PAIR UNIVERSE5

Spot/perpetual pairs on Bybit.

HISTORY1 YEAR

Second-level market data.

BOOK INPUTL1 / L2

Depth-aware execution gating.

HANDOFFLIVE

Sold to an independent operator.

FRAMEWORK

Carry survives only after execution frictions.

The research separated signal selection, executable depth and settlement accounting so an attractive funding observation could not automatically become a position.

01 · SIGNALFunding z-score

Ranked dislocations within each pair's own history.

02 · LIQUIDITYL1/L2 depth gate

Reduced size when visible depth could not support both legs.

03 · HEDGELong perp / short spot

Paired legs to constrain directional exposure.

04 · LEDGERNet carry

Applied fees, funding settlements and slippage.

RESEARCH PIPELINE

From raw books to an executable decision.

The implementation used Python and CCXT, while the public case study deliberately excludes operator-specific thresholds, position sizes and performance.

01

Normalize

Aligned second-level funding, spot, perpetual and order-book observations.

02

Standardize

Converted funding behavior into pair-level z-score states.

03

Gate

Conditioned entries and size on executable L1/L2 depth.

04

Backtest

Replayed one year with fees, funding settlements and slippage.

05

Transfer

Delivered the framework to an independent operator for live use.

EXECUTION PRINCIPLE

Depth determines capacity.

Signal strength alone did not set position size. Both legs had to remain executable under the displayed order book and modeled slippage.

DISCLOSURE BOUNDARY

No unsupported performance claim.

The code, thresholds and operator results remain private. This page documents the research architecture and verified scope only.

Public BTCUSDT stream · display only · auto-stop after 10 min