okx_spot_future_arbitrage

中文 | English

Introduction

Spot-Future Arbitrage: Spot-future arbitrage refers to the process of buying the lower-priced asset and selling the higher-priced asset when there is a significant price difference between futures and spot markets for the same cryptocurrency. Profits are realized by closing positions when the price gap narrows.

The market uses OKX spot leverage and coin-margined contracts

Spot leverage orders
Coin-margined delivery contracts for hedging

Annual yield = (future_price - spot_price) / spot_price / future_expire_days * 365 * 100%

Yield reference: https://www.okx.com/zh-hans/markets/arbitrage/spread-usd

Before You Start

1. DYOR (DO Your Own Research)
2. Before formally using the strategy, it is recommended to test it in a demo account first to confirm the stability and yield of the strategy

Preparation：

• Change the account to cross-margin (Upgrade to cross-margin mode requires the equity in the trading account to be not less than 10,000.00 USD)

Core Strategy Logic

The strategy module obtains real-time positions and fund balances through WebSocket
The data processing module obtains real-time yield rankings
Opening position logic:
    Process trading pairs in the yield ranking sequentially
    Check if the data is timed out (greater than a specified time, e.g., 10s)
        Skip if timed out
    Check if opening a position is currently allowed
        Is the strategy enabled
        Is spot USDT sufficient (balance > per_order_usd)
        Has the position limit been reached
        Does the yield meet the criteria
    Two-leg order placement
        Simple logic uses limit orders above market price, e.g., spot buys at asks[5], future sells at bids[5]
        Complex logic:
            Spot maker, place contract taker order after receiving WebSocket trade message. Price same as above. Risk point is rapid price movement
            Precision handling
                Handle spot and contract price precision based on market data
                Handle spot size precision based on market data
Closing position logic
    Iterate through current positions, obtain corresponding yields, close positions with yields below the closing threshold
    Closing operation also involves two-leg order placement, executed simultaneously

Project Deployment

dependent components:

• redis = 6.x (e.g. 6.2.6)
• mysql = 8.0.x (e.g. 8.0.31)

env up

create mysql database or using make setenv auto create db

DROP DATABASE basis_alpha_db;
create database basis_alpha_db charset=utf8mb4;

install requirements

pip install -r requirements.txt
pip install -r requirements-dev.txt

initial default data

• Create admin user

python manage.py createsuperuser --role=admin

• Initialize database tables

python manage.py migrate

Configure account

Visit http://localhost:8000/admin/strategy/account/add/ to add an account Note: The Api Secret field needs to use encrypted text, use tools/aes_encrypt.py for encryption

Add strategy configuration

Start service

• Configure environment variables TEST environment:

export PROFILE=dev

Production environment

export PROFILE=production

• Admin panel (account management, strategy configuration, order management)

python manage.py runserver

• Market data module

python manage.py start_okx_future_spot_spider

• Strategy module

python manage.py start_strategy --strategy_name test --account_name okx_test_account

Observe order execution status

FAQ

how to install mysqlclient

Exception: Can not find valid pkg-config name.

Install MySQL client and pkg-config using Homebrew: brew install mysql-client pkg-config

Set the PKG_CONFIG_PATH environment variable: export PKG_CONFIG_PATH="$(brew --prefix)/opt/mysql-client/lib/pkgconfig"

Install the mysqlclient package using pip: pip install mysqlclient==2.2.4