Polyester Margin Mean Reversion

Code TqSdk

Summary

This example describes a three-leg futures strategy that treats polyester fiber value minus the weighted costs of PTA and ethylene glycol as a production margin. It estimates the margin’s mean and standard deviation from recent daily bars, then calculates a z-score. When the score crosses either side of a stated two-standard-deviation entry threshold, the strategy takes opposing positions across the output and input contracts, scaling leg quantities by production ratios and contract multipliers.

It exits when the score returns within half a standard deviation of its mean and includes a stop condition for further adverse divergence. The document supplies code configured for a dated backtest period, but reports no performance results. Its example does not establish profitability or account for transaction costs, slippage, leg execution risk, or the validity of the assumed production ratios. Position-state handling and risk controls also warrant review before practical use.

Key ideas

The strategy models polyester fiber production margin as output contract value minus weighted PTA and ethylene glycol input values.
It uses a rolling historical mean and standard deviation to express the current margin as a z-score.
Extreme positive and negative scores trigger opposite three-leg futures positions sized using production ratios.
The example closes near the historical mean and includes a stop condition for further adverse divergence.
The document provides sample backtest code but no evidence of realized or simulated profitability.

Tags

Futures Commodities Mean reversion Backtesting Position sizing Risk management

Full text

# pta_spread.py


```py
#!/usr/bin/env python
# coding=utf-8
__author__ = "Chaos"

from datetime import date
from tqsdk import TqApi, TqAuth, TargetPosTask, TqBacktest, BacktestFinished
import numpy as np
import time

# === 用户参数 ===
# 合约参数
PF = "CZCE.PF409"      # 涤纶短纤期货合约
PTA = "CZCE.TA409"     # PTA期货合约
EG = "DCE.eg2409"      # 乙二醇期货合约
START_DATE = date(2024, 2, 1)   # 回测开始日期
END_DATE = date(2024, 4, 30)     # 回测结束日期

# 套利参数
LOOKBACK_DAYS = 30        # 计算历史价差的回溯天数
STD_THRESHOLD = 2.0       # 标准差阈值，超过此阈值视为套利机会
ORDER_VOLUME = 500         # 涤纶短纤的下单手数
CLOSE_THRESHOLD = 0.5     # 平仓阈值(标准差)

# 生产比例（可根据实际工艺调整）
PF_RATIO = 1
PTA_RATIO = 0.86
EG_RATIO = 0.34

# === 初始化API ===
api = TqApi(backtest=TqBacktest(start_dt=START_DATE, end_dt=END_DATE),
            auth=TqAuth("快期账户", "快期密码"))

# 获取合约行情和K线
pf_quote = api.get_quote(PF)
pta_quote = api.get_quote(PTA)
eg_quote = api.get_quote(EG)

pf_klines = api.get_kline_serial(PF, 60*60*24, LOOKBACK_DAYS)
pta_klines = api.get_kline_serial(PTA, 60*60*24, LOOKBACK_DAYS)
eg_klines = api.get_kline_serial(EG, 60*60*24, LOOKBACK_DAYS)

# 创建目标持仓任务
pf_pos = TargetPosTask(api, PF)
pta_pos = TargetPosTask(api, PTA)
eg_pos = TargetPosTask(api, EG)

# 获取合约乘数
pf_volume_multiple = pf_quote.volume_multiple
pta_volume_multiple = pta_quote.volume_multiple
eg_volume_multiple = eg_quote.volume_multiple

# 初始化状态变量
position_time = 0        # 建仓时间
in_position = False      # 是否有持仓
mean_spread = 0          # 历史价差均值
std_spread = 0           # 历史价差标准差

print(f"策略启动，监控合约: {PF}, {PTA}, {EG}")

# === 主循环 ===
try:
    # 初始计算历史统计值
    spreads = []
    for i in range(len(pf_klines) - 1):
        pf_price = pf_klines.close.iloc[i] * pf_volume_multiple * PF_RATIO
        pta_price = pta_klines.close.iloc[i] * pta_volume_multiple * PTA_RATIO
        eg_price = eg_klines.close.iloc[i] * eg_volume_multiple * EG_RATIO
        
        # 涤纶短纤生产利润 = 涤纶短纤价值 - (PTA成本 + EG成本)
        spread = pf_price - (pta_price + eg_price)
        spreads.append(spread)
    
    mean_spread = np.mean(spreads)
    std_spread = np.std(spreads)
    print(f"历史涤纶短纤利润均值: {mean_spread:.2f}, 标准差: {std_spread:.2f}")

    # 主循环
    while True:
        api.wait_update()
        
        # 当K线数据有变化时进行计算
        if api.is_changing(pf_klines) or api.is_changing(pta_klines) or api.is_changing(eg_klines):
            # 重新计算历史价差统计
            spreads = []
            for i in range(len(pf_klines) - 1):
                pf_price = pf_klines.close.iloc[i] * pf_volume_multiple * PF_RATIO
                pta_price = pta_klines.close.iloc[i] * pta_volume_multiple * PTA_RATIO
                eg_price = eg_klines.close.iloc[i] * eg_volume_multiple * EG_RATIO
                
                spread = pf_price - (pta_price + eg_price)
                spreads.append(spread)
            
            mean_spread = np.mean(spreads)
            std_spread = np.std(spreads)
            
            # 计算当前利润价差
            pf_price = pf_klines.close.iloc[-1] * pf_volume_multiple * PF_RATIO
            pta_price = pta_klines.close.iloc[-1] * pta_volume_multiple * PTA_RATIO
            eg_price = eg_klines.close.iloc[-1] * eg_volume_multiple * EG_RATIO
            
            current_spread = pf_price - (pta_price + eg_price)
            
            # 计算z-score (标准化的价差)
            z_score = (current_spread - mean_spread) / std_spread
            
            print(f"当前涤纶短纤利润: {current_spread:.2f}, Z-score: {z_score:.2f}")
            
            # 获取当前持仓
            pf_position = api.get_position(PF)
            pta_position = api.get_position(PTA)
            eg_position = api.get_position(EG)
            
            current_pf_pos = pf_position.pos_long - pf_position.pos_short
            current_pta_pos = pta_position.pos_long - pta_position.pos_short
            current_eg_pos = eg_position.pos_long - eg_position.pos_short
            
            # 计算实际下单手数（依据比例）
            pta_volume = int(ORDER_VOLUME * PTA_RATIO / PF_RATIO)
            eg_volume = int(ORDER_VOLUME * EG_RATIO / PF_RATIO)
            
            # === 交易信号判断 ===
            if not in_position:  # 如果没有持仓
                if z_score > STD_THRESHOLD:  # 利润显著高于均值
                    # 做空利润：卖出PF，买入PTA和EG
                    print(f"做空利润：卖出PF{ORDER_VOLUME}手，买入PTA{pta_volume}手和EG{eg_volume}手")
                    pf_pos.set_target_volume(-ORDER_VOLUME)
                    pta_pos.set_target_volume(pta_volume)
                    eg_pos.set_target_volume(eg_volume)
                    position_time = time.time()
                    in_position = True
                    
                elif z_score < -STD_THRESHOLD:  # 利润显著低于均值
                    # 做多利润：买入PF，卖出PTA和EG
                    print(f"做多利润：买入PF{ORDER_VOLUME}手，卖出PTA{pta_volume}手和EG{eg_volume}手")
                    pf_pos.set_target_volume(ORDER_VOLUME)
                    pta_pos.set_target_volume(-pta_volume)
                    eg_pos.set_target_volume(-eg_volume)
                    position_time = time.time()
                    in_position = True
            
            else:  # 如果已有持仓
                # 检查是否应当平仓
                if abs(z_score) < CLOSE_THRESHOLD:  # 利润回归正常
                    print("利润回归正常，平仓所有头寸")
                    pf_pos.set_target_volume(0)
                    pta_pos.set_target_volume(0)
                    eg_pos.set_target_volume(0)
                    in_position = False

                
                # 止损逻辑
                if (z_score > STD_THRESHOLD * 1.5 and current_pf_pos > 0) or \
                   (z_score < -STD_THRESHOLD * 1.5 and current_pf_pos < 0):
                    print("止损：利润向不利方向进一步偏离")
                    pf_pos.set_target_volume(0)
                    pta_pos.set_target_volume(0)
                    eg_pos.set_target_volume(0)
                    in_position = False

except BacktestFinished as e:
    print("回测结束")
    api.close()


```

Shown in full with attribution under the source's licence. Licence: Apache-2.0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.

Polyester Fiber Margin Mean Reversion with PTA and EG Futures

Summary

Key ideas

Tags