简单低波动率指数
来源:https://uqer.io/community/share/5566a9b8f9f06c6641e97aea
金融市场的波动性加剧,为了提供更好的下行保护,低波动率的Smart Beta策略受到了广泛的欢迎
代表指数
目标指数
HS300
选股
计算目标指数股票池中样本股过去100个交易日中的历史波动率,并挑选其中波动率最低的50只股票作为指数的成分股
加权
与传统指数市值加权不同,本指数根据股票波动率倒数为个股权重
实现细节
通过DataAPI.EquRetudGet
获取不考虑现金红利再投资情况下的每日收益率,波动率为调仓前100个交易日的日收益率标准差
import numpy as np
import pandas as pd
start = '2012-01-01' # 回测起始时间
end = '2015-05-01' # 回测结束时间
benchmark = 'HS300' # 策略参考标准
universe = set_universe('HS300') # 证券池,回测支持股票和基金
capital_base = 10000000 # 起始资金
refresh_rate = 100 # 调仓频率,即每 refresh_rate 个交易日执行一次 handle_data() 函数
cal = Calendar('China.SSE')
def initialize(account): # 初始化虚拟账户状态
pass
def handle_data(account): # 每个交易日的买入卖出指令
volatility_res = {}
cal_today = Date.fromDateTime(account.current_date)
start_day = cal.advanceDate(cal_today, '-101B', BizDayConvention.Following)
yesterday = cal.advanceDate(cal_today, '-1B', BizDayConvention.Following)
for stk in universe:
try:
data = DataAPI.EquRetudGet(ticker=stk[:6], beginDate=Date.toDateTime(start_day).strftime('%Y%m%d'), endDate=Date.toDateTime(yesterday).strftime('%Y%m%d'), field=['ticker',"dailyReturnNoReinv"])
revenue = data['dailyReturnNoReinv']
volatility_res[stk] = np.std(revenue)
except:
universe.remove(stk)
res = pd.Series(volatility_res).order()[:50]
temp = np.ones(50)
res = np.divide(temp, res)
weight_sum = res.values.sum()
order_list = dict(res/weight_sum)
for stk in account.valid_secpos:
order_to(stk, 0)
for s, weight in order_list.iteritems():
if account.referencePrice[s] == 0:
continue
order(s, capital_base*weight/account.referencePrice[s])
print "Benchmark Volatility : ", perf['benchmark_volatility']
print "Index Volatility : ", perf['volatility']
Benchmark Volatility : 0.213927304422
Index Volatility : 0.156413355501
结果分析
通过以上结果我们可以看到,该策略alpha极小,beta较大,并显著减小了波动率