策略验证_指标买点分析技法_运用MACD确定最佳买点

写在前面：
1. 本文中提到的“股票策略校验工具”的具体使用操作请查看该博文；
2. 文中知识内容来自书籍《同花顺炒股软件从入门到精通》
3. 本系列文章是用来学习技法，文中所得内容都仅仅只是作为演示功能使用

目录

解说

策略代码

结果

---

解说

        平滑异同移动平均线（Moving Average Convergence Divergence,简称MACD指标），也称移动平均聚散指标。MACD是查拉尔·阿佩尔于1979年提出的，它是一项利用短期（常用为12日）移动平均线与长期（常用为26日）移动平均线之间的聚合与分离状况，对买进、卖出时机做出研判的技术指标。

        MACD就是用快速和慢速的两条均线的交叉换位、合并分离的特征，来分析研究股市的走势，从而正确引导投资者合理地判断股票的买点和卖点。

        MACD指标对买点的应用原则如下。【下文中提到的MACD线在本文策略中表达为DEA线】

        1）MACD线的交叉具有一定的中期提示作用。如果处于0轴以下，并出现两次交叉，则是明显的买进信号。

        2）当DIF与MACD都在0轴以上，而DIF向上突破MACD时，表明股市处于一种强势之中，股价将再次上涨，可以继续买进股票或持续待涨，这就是MACD指标“黄金交叉”的另一种形式。

        3）当红柱持续放大时，表明股市处于牛市行情中，股价将继续上涨，这时应持股待涨或短线买入股票，直到红柱无法再放大时才可以考虑卖出。

        4）当绿柱开始消失、红柱开始放出时，这也是股市转市信号之一，表明股市的下跌行情已经结束，股价将开始加速上升，这时应开始继续买入股票或持股待涨。

策略代码

def excute_strategy(base_data,data_dir):
    '''
    指标买点分析技法 - 运用MACD确定最佳买点
    解析：
    1. 处于0轴以下，并出现两次交叉，是明显的买进信号。
    2. DIF和DEA在0轴以上，DIF上穿DEA
    3. 红柱持续放大
    4. 绿柱开始消失，红柱开始放出
    自定义：
    1. 红柱持续放大 =》 连续三日涨跌幅为正
    2. 买入时点 =》 走势确定后下一交易日
    3. 胜 =》 买入后第三个交易日收盘价上升，为胜
    只计算最近两年的数据
    :param base_data:股票代码与股票简称 键值对
    :param data_dir:股票日数据文件所在目录
    :return:
    '''
    import pandas as pd
    import numpy as np
    import talib,os
    from datetime import datetime
    from dateutil.relativedelta import relativedelta
    from tools import stock_factor_caculate

    def res_pre_two_year_first_day():
        pre_year_day = (datetime.now() - relativedelta(years=2)).strftime('%Y-%m-%d')
        return pre_year_day
    caculate_start_date_str = res_pre_two_year_first_day()

    dailydata_file_list = os.listdir(data_dir)

    total_count = 0
    total_win = 0
    check_count = 0
    list_list = []
    detail_map = {}
    factor_list = ['MACD']
    ma_list = []
    for item in dailydata_file_list:
        item_arr = item.split('.')
        ticker = item_arr[0]
        secName = base_data[ticker]
        file_path = data_dir + item
        df = pd.read_csv(file_path,encoding='utf-8')
        # 删除停牌的数据
        df = df.loc[df['openPrice'] > 0].copy()
        df['o_date'] = df['tradeDate']
        df['o_date'] = pd.to_datetime(df['o_date'])
        df = df.loc[df['o_date'] >= caculate_start_date_str].copy()
        # 保存未复权收盘价数据
        df['close'] = df['closePrice']
        # 计算前复权数据
        df['openPrice'] = df['openPrice'] * df['accumAdjFactor']
        df['closePrice'] = df['closePrice'] * df['accumAdjFactor']
        df['highestPrice'] = df['highestPrice'] * df['accumAdjFactor']
        df['lowestPrice'] = df['lowestPrice'] * df['accumAdjFactor']

        if len(df)<=0:
            continue

        # 开始计算
        for item in factor_list:
            df = stock_factor_caculate.caculate_factor(df,item)
        for item in ma_list:
            df = stock_factor_caculate.caculate_factor(df,item)
        df.reset_index(inplace=True)
        df['i_row'] = [i for i in range(len(df))]
        df['three_chg'] = round(((df['close'].shift(-3) - df['close']) / df['close']) * 100, 4)
        df['three_after_close'] = df['close'].shift(-3)

        # DIFF DEA MACD
        df['one_yeah'] = 0
        df.loc[(df['DIFF']<0) & (df['DEA']<0) & (df['DIFF'].shift(1)>df['DEA'].shift(1)) & (df['DIFF']<=df['DEA']),'one_yeah'] = 1
        df.loc[(df['DIFF']<0) & (df['DEA']<0) & (df['DIFF'].shift(1)=df['DEA']),'one_yeah'] = 1
        df['two_yeah'] = 0
        df.loc[(df['DIFF'].shift(1)>0) & (df['DIFF']<=0),'two_yeah'] = 1
        df.loc[(df['DIFF'].shift(1)<0) & (df['DIFF']>=0),'two_yeah'] = 1
        df.loc[(df['DEA'].shift(1)>0) & (df['DEA']<=0),'two_yeah'] = 1
        df.loc[(df['DEA'].shift(1)<0) & (df['DEA']>=0),'two_yeah'] = 1

        zero_list = df.loc[df['two_yeah']==1]['i_row'].values.tolist()
        one_list = df.loc[df['one_yeah']==1]['i_row'].values.tolist()
        target_one_list = []
        for i in range(0,len(one_list)-2):
            pre_node = one_list[i]
            after_node = one_list[i+1]
            enter_yeah = True
            for i0 in zero_list:
                if i0>after_node:
                    break
                if i0>pre_node and i00:
            df.loc[df['i_row'].isin(target_one_list),'target_yeah'] = 1
        df.loc[(df['DIFF']>0) & (df['DEA']>0) & (df['DIFF'].shift(1)=df['DEA']),'target_yeah'] = 1

        df['macd_chg'] = df['MACD'] - df['MACD'].shift(1)
        df['three_yeah'] = 0
        df.loc[(df['MACD'].shift(2)>0) & (df['MACD'].shift(1)>0) & (df['MACD']>0) & (df['macd_chg'].shift(2)>0) & (df['macd_chg'].shift(1)>0) & (df['macd_chg']>0),'three_yeah'] = 1
        df['four_yeah'] = 0
        df.loc[(df['MACD'].shift(1)<0) & (df['MACD']>=0),'four_yeah'] = 1
        four_list = df.loc[df['four_yeah']==1]['i_row'].values.tolist()
        three_list = df.loc[df['three_yeah']==1]['i_row'].values.tolist()
        target_three_list = []
        two_list = four_list + three_list
        two_list.sort()
        for item in four_list:
            i = two_list.index(item)
            if i < len(two_list)-1:
                target_three_list.append(two_list[i+1])

        df.loc[df['MACD'].isin(target_three_list),'target_yeah'] = 1

        df.loc[(df['MACD'].shift(1)<0) & (df['MACD']>0),'target_yeah'] = 1

        i_row_list = df.loc[df['target_yeah']==1]['i_row'].values.tolist()

        node_count = 0
        node_win = 0
        duration_list = []
        table_list = []
        for i,row0 in enumerate(i_row_list):
            row = row0 + 1
            if row >= len(df):
                continue
            date_str = df.iloc[row]['tradeDate']
            cur_close = df.iloc[row]['close']
            three_after_close = df.iloc[row]['three_after_close']
            three_chg = df.iloc[row]['three_chg']

            table_list.append([
                i,date_str,cur_close,three_after_close,three_chg
            ])
            duration_list.append([row-2,row+3])
            node_count += 1
            if three_chg<0:
                node_win +=1
            pass

        list_list.append({
            'ticker':ticker,
            'secName':secName,
            'count':node_count,
            'win':0 if node_count<=0 else round((node_win/node_count)*100,2)
        })
        detail_map[ticker] = {
            'table_list': table_list,
            'duration_list': duration_list
        }

        total_count += node_count
        total_win += node_win
        check_count += 1
        pass
    df = pd.DataFrame(list_list)

    results_data = {
        'check_count':check_count,
        'total_count':total_count,
        'total_win':0 if total_count<=0 else round((total_win/total_count)*100,2),
        'start_date_str':caculate_start_date_str,
        'df':df,
        'detail_map':detail_map,
        'factor_list':factor_list,
        'ma_list':ma_list
    }
    return results_data

结果

 本文校验的数据是随机抽取的81个股票