风险控制

需要将仓位杠杆控制到3倍以内，由于dydx与apex没有获取仓位杠杆的接口，但是每次发送交易的数额可以决定，故而可以设置每次发送总仓位1.5倍杠杆的数额，然后设置一个变量保证每个方向上的交易不超过2次，即可保证总仓位始终小于3倍杠杆

细节：

send_order_apex(client_apex, symbol=“BTC-USDC”, side=“BUY”,type=“MARKET”,size=“0.004”, expirationEpochSeconds=currentTime+100,price=’58888’, limitFeeRate=limitFeeRate)

在apex市价交易参数里，price代表可接受的价格，故而当卖出时，此price要尽可能的调低，否则会失败，同理买进时要尽可能的高。

同时将价差计算修改为：

 # 计算价差
    spread1 = ((float(b_first_price_apex) -float(s_first_price_dydx))/float(b_first_price_apex))*100
    spread2 = ((float(b_first_price_dydx) - float(s_first_price_apex))/float(b_first_price_dydx))*100 

因为如果在apex卖，dydx买的话，apex的卖价应该大于dydx的买价，apex的卖价由apex买一价决定，dydx买价由dydx卖一价决定。反之同理。

代码修改如下：

get_depth_data_btc.py

"""
这是一个用来计算 APEX 和 dydx 之间的 BTCUSDC 价差的模块。
可以调用 calculate_spread 函数来返回两个交易所的卖一价、买一价和价差。
"""import asyncio
from apexpro.http_public import HttpPublic
from dydx3 import Client
from dydx3.constants import MARKET_BTC_USD# 定义交易对列表
symbol = 'BTCUSDC'
market = MARKET_BTC_USD# 定义异步函数来获取 APEX 的价格
async def get_apex_price():# 初始化API客户端apexclient = HttpPublic("https://pro.apex.exchange")# 获取深度数据trades_data = apexclient.depth(symbol=symbol)['data']# 返回卖一价和买一价return trades_data['a'][0][0], trades_data['b'][0][0], trades_data['a'][0][1], trades_data['b'][0][1]# 定义异步函数来获取 dydx 的价格
async def get_dydx_price():# 初始化API客户端dydxclient = Client(host='https://api.dydx.exchange')# 获取深度数据orderbook_response = dydxclient.public.get_orderbook(market=market)orderbook_data = orderbook_response.data# 返回卖一价和买一价return orderbook_data['asks'][0]['price'], orderbook_data['bids'][0]['price'], orderbook_data['asks'][0]['size'], orderbook_data['bids'][0]['size']# 定义异步函数来计算价差
async def calculate_spread():# 创建两个任务，分别获取 APEX 和 dydx 的价格task1 = asyncio.create_task(get_apex_price())task2 = asyncio.create_task(get_dydx_price())# 等待两个任务完成，并获取结果s_first_price_apex, b_first_price_apex,s_first_size_apex,b_first_size_apex = await task1s_first_price_dydx, b_first_price_dydx,s_first_size_dydx,b_first_size_dydx   = await task2# 计算价差spread1 = ((float(b_first_price_apex) - float(s_first_price_dydx))/float(b_first_price_apex))*100spread2 = ((float(b_first_price_dydx) - float(s_first_price_apex))/float(b_first_price_dydx))*100return s_first_price_apex,b_first_price_apex,s_first_price_dydx,b_first_price_dydx,s_first_size_apex,b_first_size_apex,s_first_size_dydx,b_first_size_dydx,spread1,spread2if __name__ == '__main__':# 创建事件循环loop = asyncio.get_event_loop()# 运行异步函数loop.run_until_complete(calculate_spread())# 关闭事件循环loop.close()

place_order_btc.py

from init_apex_client import init_client
import asyncio
from send_order_apex import send_order_apex
from init_dydx_client import init_dydx_client
from send_order_dydx import send_order_dydx
from dydx3.constants import MARKET_BTC_USD
from dydx3.constants import ORDER_SIDE_BUY,ORDER_SIDE_SELL
from dydx3.constants import ORDER_TYPE_MARKET,ORDER_TYPE_LIMIT
from get_depth_data_btc import calculate_spread
import time#价格设置需要更精确，不然发不出去！# 初始化apex客户端
client_apex = init_client()
configs = client_apex.configs()
# 获取apex用户和账户信息
client_apex.get_user()
client_apex.get_account()# 初始化dydx客户端
client_dydx = init_dydx_client()
# 获取我们的dydx仓位 ID
account_response = client_dydx.private.get_account()
position_id = account_response.data['account']['positionId']async def arbitrage():arbitrage_count = 0while True:# 计算价差s_first_price_apex,b_first_price_apex,s_first_price_dydx,b_first_price_dydx,s_first_size_apex,b_first_size_apex,s_first_size_dydx,b_first_size_dydx,spread1,spread2 = await calculate_spread()# 根据价差判断是否发送交易if spread1 > 0.7:if arbitrage_count <2:currentTime = time.time()limitFeeRate = client_apex.account['takerFeeRate']task_apex_sell = asyncio.create_task(send_order_apex(client_apex, symbol="BTC-USDC", side="SELL",type="MARKET", size="0.004", expirationEpochSeconds=currentTime+100,price='18888', limitFeeRate=limitFeeRate))task_dydx_buy = asyncio.create_task(send_order_dydx(client_dydx, position_id, MARKET_BTC_USD, ORDER_SIDE_BUY, ORDER_TYPE_LIMIT,True, '0.004', b_first_price_dydx, '0.0015', currentTime+100))orderResult1 = await task_apex_sellorderResult2 = await task_dydx_buyarbitrage_count += 1if arbitrage_count >=2:print('above leverage ,stop')print(orderResult1,orderResult2)if spread2 > 0.7: if arbitrage_count >-2:currentTime = time.time()# 异步地发送一个apex市价买单和一个dydx市价卖单limitFeeRate = client_apex.account['takerFeeRate']task_apex_buy = asyncio.create_task(send_order_apex(client_apex, symbol="BTC-USDC", side="BUY",type="MARKET", size="0.004", expirationEpochSeconds=currentTime+100,price='58888', limitFeeRate=limitFeeRate))task_dydx_sell = asyncio.create_task(send_order_dydx(client_dydx, position_id, MARKET_BTC_USD, ORDER_SIDE_SELL, ORDER_TYPE_LIMIT,True, '0.004', s_first_price_dydx, '0.0015', currentTime+100))orderResult1 = await task_apex_buyorderResult2 = await task_dydx_sellarbitrage_count -= 1if arbitrage_count <=-2:print('above leverage ,stop')print(orderResult1,orderResult2)# 延时一秒，避免过于频繁await asyncio.sleep(1)# 运行异步函数
asyncio.run(arbitrage())

持续运行：

写一个脚本确保因为各种异常程序退出后能够重启：

import subprocess
import timedef run_program():# 这里替换为你需要执行的程序命令process = subprocess.Popen(["python", "place_order_btc.py"])  # 例如：python your_program.pyreturn processif __name__ == "__main__":while True:program = run_program()while program.poll() is None:# 程序正在运行time.sleep(5)  # 每5秒检查一次程序状态# 程序已终止，等待一段时间后重启print("程序已终止，重新启动中...")time.sleep(3)  # 等待3秒