企业🤖AI Agent构建引擎,智能编排和调试,一键部署,支持私有化部署方案 广告
# 市值最小300指数 > 来源:https://uqer.io/community/share/5604fbe6f9f06c597665ef37 刷爆沪深300 策略名称: 市值最小300指数 回测时间:2013-01-01 到 2015-09-24 调仓期 :20交易日 策略思想:找A股市场市值最小的300只股票,等权重构建最小300指数 注意 : + 内存不够请自行缩短回测时间或universe + 此贴有4个!!! ```py import pandas as pd import numpy as np from pandas import Series, DataFrame start = '2013-01-01' # 回测起始时间 end = '2015-09-24' # 回测结束时间 benchmark = 'HS300' # 策略参考标准 universe0 = set_universe('A') # 证券池,支持股票和基金 universe1 = set_universe('HS300') universe2 = set_universe('ZZ500') universe = list(set(universe0).difference(set(universe1+universe2))) #最小市值股一定不在中证500和沪深300 pass capital_base = 100000000 # 起始资金 freq = 'd' # 策略类型,'d'表示日间策略使用日线回测,'m'表示日内策略使用分钟线回测 refresh_rate = 20 # 调仓频率,表示执行handle_data的时间间隔,若freq = 'd'时间间隔的单位为交易日,若freq = 'm'时间间隔为分钟 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 total_money = account.referencePortfolioValue prices = account.referencePrice buylist = [] marketValue = DataFrame() today = account.current_date.strftime('%Y%m%d') for s in range(len(account.universe)/40 + 1): if s == len(account.universe)/40: temp_list = account.universe[s*40:] else : temp_list = account.universe[s*40:(s+1)*40] #MktEqudGet接口一次最多选50个 try: #排除最后一次temp_list为零的可能 marketValue_temp = DataAPI.MktEqudGet(secID = temp_list,tradeDate= today, field=u"secID,marketValue",pandas="1") except : pass marketValue = pd.concat([marketValue,marketValue_temp]) marketValue = marketValue.sort('marketValue',ascending=True).drop_duplicates('secID') marketValue.set_index('secID',inplace=True) marketValue = marketValue.dropna() #排除新股发行日 for s in list(marketValue.index) : if not (np.isnan(prices[s]) or prices[s] == 0) : buylist.append(s) if len(buylist) >= 300 : break sell_list = [x for x in account.valid_secpos if x not in buylist] for stk in sell_list: order_to(stk, 0) for stk in buylist: order_to(stk, int(total_money/300/prices[stk]/100)*100) ``` ![](https://box.kancloud.cn/2016-07-30_579cbdb0b0220.jpg) 更改权重比例 :加权流通市值倒数(越小越买) 买买买! ```py import pandas as pd import numpy as np from pandas import Series, DataFrame start = '2013-01-01' # 回测起始时间 end = '2015-09-24' # 回测结束时间 benchmark = 'HS300' # 策略参考标准 universe0 = set_universe('A') # 证券池,支持股票和基金 universe1 = set_universe('HS300') universe2 = set_universe('ZZ500') universe = list(set(universe0).difference(set(universe1+universe2))) capital_base = 100000000 # 起始资金 freq = 'd' # 策略类型,'d'表示日间策略使用日线回测,'m'表示日内策略使用分钟线回测 refresh_rate = 20 # 调仓频率,表示执行handle_data的时间间隔,若freq = 'd'时间间隔的单位为交易日,若freq = 'm'时间间隔为分钟 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 total_money = account.referencePortfolioValue prices = account.referencePrice buylist = [] marketValue = DataFrame() today = account.current_date.strftime('%Y%m%d') for s in range(len(account.universe)/40 + 1): if s == len(account.universe)/40: temp_list = account.universe[s*40:] else : temp_list = account.universe[s*40:(s+1)*40] #MktEqudGet接口一次最多选50个 try: #排除最后一次temp_list为零的可能 marketValue_temp = DataAPI.MktEqudGet(secID = temp_list,tradeDate= today, field=u"secID,marketValue,negMarketValue",pandas="1") except : pass marketValue = pd.concat([marketValue,marketValue_temp]) marketValue = marketValue.sort('marketValue',ascending=True).drop_duplicates('secID') marketValue.set_index('secID',inplace=True) marketValue = marketValue.dropna() #排除新股发行日 for s in list(marketValue.index) : if not (np.isnan(prices[s]) or prices[s] == 0) : buylist.append(s) if len(buylist) >= 300 : break sell_list = [x for x in account.valid_secpos if x not in buylist] for stk in sell_list: order_to(stk, 0) #加权流通市值倒数购买 weight_list = [] for stk in buylist: weight_list.append(1.0/marketValue['negMarketValue'][stk]) temp_sum = 0 for temp in weight_list: temp_sum += temp weight_list = [x/temp_sum for x in weight_list] i = 0 for stk in buylist: order_to(stk, int(total_money*weight_list[i]/prices[stk]/100)*100) i += 1 ``` ![](https://box.kancloud.cn/2016-07-30_579cbdb0cc785.jpg) 是不是在想賺钱分分钟了? 您有买300只个股的毛爷爷吗,啊! 〇_〇- . . 木有?我会告诉你买十只也很叼吗??? ```py import pandas as pd import numpy as np from pandas import Series, DataFrame start = '2013-01-01' # 回测起始时间 end = '2015-09-24' # 回测结束时间 benchmark = 'HS300' # 策略参考标准 universe0 = set_universe('A') # 证券池,支持股票和基金 universe1 = set_universe('HS300') universe2 = set_universe('ZZ500') universe = list(set(universe0).difference(set(universe1+universe2))) capital_base = 100000000 # 起始资金 freq = 'd' # 策略类型,'d'表示日间策略使用日线回测,'m'表示日内策略使用分钟线回测 refresh_rate = 20 # 调仓频率,表示执行handle_data的时间间隔,若freq = 'd'时间间隔的单位为交易日,若freq = 'm'时间间隔为分钟 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 total_money = account.referencePortfolioValue prices = account.referencePrice buylist = [] marketValue = DataFrame() today = account.current_date.strftime('%Y%m%d') for s in range(len(account.universe)/40 + 1): if s == len(account.universe)/40: temp_list = account.universe[s*40:] else : temp_list = account.universe[s*40:(s+1)*40] #MktEqudGet接口一次最多选50个 try: #排除最后一次temp_list为零的可能 marketValue_temp = DataAPI.MktEqudGet(secID = temp_list,tradeDate= today, field=u"secID,marketValue",pandas="1") except : pass marketValue = pd.concat([marketValue,marketValue_temp]) marketValue = marketValue.sort('marketValue',ascending=True).drop_duplicates('secID') marketValue.set_index('secID',inplace=True) marketValue = marketValue.dropna() #排除新股发行日 for s in list(marketValue.index) : if not (np.isnan(prices[s]) or prices[s] == 0) : buylist.append(s) if len(buylist) >= 10 : break sell_list = [x for x in account.valid_secpos if x not in buylist] for stk in sell_list: order_to(stk, 0) #只买最优十只 for stk in buylist: order_to(stk, int(total_money/10/prices[stk]/100)*100) ``` ![](https://box.kancloud.cn/2016-07-30_579cbdb0e832e.jpg) 我会告诉你有庄家(机构持股)的股票更容易飞 ???   小注 :此策略中`DataAPI.JY.EquInstShJYGet`(恒生聚源接口)暂不开放!!!活跃用户自行申请 ```py import pandas as pd import numpy as np from pandas import Series, DataFrame start = '2013-01-01' # 回测起始时间 end = '2015-09-24' # 回测结束时间 benchmark = 'HS300' # 策略参考标准 universe0 = set_universe('A') # 证券池,支持股票和基金 universe1 = set_universe('HS300') universe2 = set_universe('ZZ500') universe = list(set(universe0).difference(set(universe1+universe2))) capital_base = 100000000 # 起始资金 freq = 'd' # 策略类型,'d'表示日间策略使用日线回测,'m'表示日内策略使用分钟线回测 refresh_rate = 20 # 调仓频率,表示执行handle_data的时间间隔,若freq = 'd'时间间隔的单位为交易日,若freq = 'm'时间间隔为分钟 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 total_money = account.referencePortfolioValue prices = account.referencePrice buylist = [] marketValue = DataFrame() today = account.current_date.strftime('%Y%m%d') for s in range(len(account.universe)/40 + 1): if s == len(account.universe)/40: temp_list = account.universe[s*40:] else : temp_list = account.universe[s*40:(s+1)*40] #MktEqudGet接口一次最多选50个 try: #排除最后一次temp_list为零的可能 marketValue_temp = DataAPI.MktEqudGet(secID = temp_list,tradeDate= today, field=u"secID,marketValue",pandas="1") except : pass marketValue = pd.concat([marketValue,marketValue_temp]) marketValue = marketValue.sort('marketValue',ascending=True).drop_duplicates('secID') marketValue.set_index('secID',inplace=True) marketValue = marketValue.dropna() # 机构持股 非第一天上市新股 for s in list(marketValue.index) : try : # 处理巨源的数据接口没有此股 temp = DataAPI.JY.EquInstShJYGet ( secID = s , field = u"instNrfaPct" , pandas = "1" ) except : print account.current_date.strftime('%Y-%m-%d'),' ',s,' ','DataAPI.JY.EquInstShJYGet get wrong' continue #有机构持股 > 30% if temp['instNrfaPct'][0] > 30 and not (np.isnan(prices[s]) or prices[s] == 0) : buylist.append(s) if len(buylist) >= 10 : break sell_list = [x for x in account.valid_secpos if x not in buylist] for stk in sell_list: order_to(stk, 0) for stk in buylist: order_to(stk, int(total_money/10/prices[stk]/100)*100) ``` ![](https://box.kancloud.cn/2016-07-30_579cbdb10f512.jpg) 呵呵,居然看完了,还不赶紧点赞克隆去賺钱!!!内存不够的还不赶紧签到 !!!