港股打新之卖出策略（暗盘和首日）

 

港股新股的，暗盘收盘价，首日开盘价，首日收盘价平均盈利率。

参考：集思录新股暗盘数据

新股：20190501上市以来的新股。

 

结论：

暗盘盘尾卖出预期收益18.3515584416

开盘价卖出预期收益 13.4619749834

收盘价卖出预期收益 11.7666170539

最高价卖出预期收益(不可行) 33.6542119814

最低价卖出预期收益(不可行) -0.714902449712

 

可见暗盘卖出最合适。

 

代码

# 近期上市新股
stockDf=DataAPI.HKEquGet(secID=u"",ticker=u"",listStatusCD=u"L",field=u"",pandas="1")
stockDf=stockDf[stockDf['listDate']>'2019-05-01']
stockDf=stockDf[stockDf['ListSector']=='主板']
print(stockDf.shape)
tickers=stockDf['ticker'].values.tolist()
print(tickers[:5])

import numpy as np
import pandas as pd

#新股上市首日价格信息，开盘价，收盘，最高，最低
stockPriceDf=DataAPI.MktHKEqudGet(secID='',ticker=tickers,tradeDate=u"",beginDate=u"20190101",endDate=u"",field=u"",pandas="1")
stockPriceDf=stockPriceDf.groupby('secShortName').first()

# stockPriceDf
stockPriceDf['openRate']=(stockPriceDf['openPrice']/stockPriceDf['preClosePrice']-1)*100
stockPriceDf['closeRate']=(stockPriceDf['closePrice']/stockPriceDf['preClosePrice']-1)*100
stockPriceDf['highRate']=(stockPriceDf['highestPrice']/stockPriceDf['preClosePrice']-1)*100
stockPriceDf['lowRate']=(stockPriceDf['lowestPrice']/stockPriceDf['preClosePrice']-1)*100
stockPriceDf=stockPriceDf.replace([np.inf, -np.inf], np.nan)
stockPriceDf=stockPriceDf.dropna(subset=['openRate','closeRate'],axis=0)

stockPriceDf=stockPriceDf[['ticker','openRate','closeRate','highRate','lowRate']]
stockPriceDf

hkAnpanDf=pd.read_csv('hk_newstock_anpan.csv')
hkAnpanDf['ticker']=hkAnpanDf['ticker'].apply(lambda x:str(x).rjust(5,'0'))
hkAnpanDf['anpanRate']=hkAnpanDf['anpanRate'].apply(lambda x:float(x.replace('%','').replace('-','0')))

stockPriceDf=stockPriceDf.merge(hkAnpanDf,on='ticker',how='inner')
stockPriceDf
print('暗盘盘尾卖出预期收益 ')
print(stockPriceDf['anpanRate'].mean())
print('开盘价卖出预期收益 ')
print(stockPriceDf['openRate'].mean())
print('收盘价卖出预期收益 ')
print(stockPriceDf['closeRate'].mean())
print('最高价卖出预期收益(不可行) ')
print(stockPriceDf['highRate'].mean())
print('最低价卖出预期收益(不可行) ')
print(stockPriceDf['lowRate'].mean())

#可见，暗盘收盘价卖出最佳
stockPriceDf

# 第二，暗盘收盘买入，次日最高价涨跌率差异
#收益模拟： 暗盘正收益，则暗盘尾部买入，次日卖出价格暗盘+5%,否则持有到尾盘
def score(x):
    if x['anpanRate']<0:
        return 0
    else:
        if x['highRate']-x['anpanRate']>5.0:
            return 5.0
        else:
            return x['closeRate']-x['anpanRate']
stockPriceDf['predictRate']=stockPriceDf.apply(lambda x:score(x),axis=1)
print(stockPriceDf['predictRate'].mean())

#收益模拟： 暗盘正收益，次日止损价格暗盘5%,否则持有到尾盘
def score(x):
    if x['anpanRate']<0:
        return 0
    else:
        if x['lowRate']-x['anpanRate']>5.0:
            return -5.0
        else:
            return x['closeRate']-x['anpanRate']
stockPriceDf['predictRate']=stockPriceDf.apply(lambda x:score(x),axis=1)
print(stockPriceDf['predictRate'].mean())

代码依赖外部数据hk_newstock_anpan.csv

内容如下，从集思录中取得，新股暗盘收盘涨跌百分率。（copy到excel，删除多余列，保存到csv即可）